Datasets:

Jingmei

/

Pandemic_PMC

like 0

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9497950/

From the Intersection of Food-Borne Zoonoses and EU Green Policies to an In-Embryo One Health Financial Model

The European Union (EU) adopts the One Health (OH) approach, based on the relationships between human, animal, and environmental health. OH concerns a multitude of aspects, some of which are discussed here. OH overlaps the European Green Deal plan and its relaunched Farm to Fork Strategy, which aims at spreading organic farms adopting the circular economy, in order to improve human health through both better environmental conditions and healthier food. Nevertheless, zoonoses cause sanitary cost in terms of infected farm personnel, lower productivity, and lower fertility of infected farm animals. In such scenarios, the decreased breeding yield and the lower income induce higher cost of farm products, meaning that the market price rises, becoming uncompetitive when compared to the prices of industrial products. Consequently, lower revenues can hinder the farm growth expected in the framework of the EU Green Deal. Since zoonosis control is a key element in aligning EU policies aimed at achieving the EU Green Deal goal of "ZERO environmental impact" by 2050, the authors suggest the inclusion of the parameter economic health in the OH approach, in order to individuate EU Member States (MSs) economically unable to conduct eradication programmes and to finance them. Economic health is here considered as a starting point of the new ethical and science-based One Health Financial Model that the authors suggest as an in-embryo model, in which specific rules should regulate public funds, private investments, and trading, which should exclusively concern public services and private enterprises complying with most of the OH parameters. In this way, economic losses due to collateral negative effects deriving from human activities can be progressively decreased, and the entire planet will benefit from the process. Despite the considerable efforts being carried out in the context of the OH approach, war causes tragic and devastating effects on the physical and mental health of human beings, on their lives, on pandemic and zoonotic threats, on animals, on plants and, last but not least, on the environment. War is incompatible with OH. Enormous efforts for peace are therefore urgently needed. 1. Introduction Zoonotic agents are pathogens with an unrestricted host spectrum. In nature, their survival occurs in reservoir animal species, which generally do not present clinical symptoms and, therefore, are difficult to identify. Promiscuity between farmed animals and wildlife increases the risk of transmission of pathogens and their consequent adaptation to new host species, including human beings. Therefore, promiscuity increases the risk of emergence of new zoonoses. According to the World Organisation for Animal Health (OIE), zoonoses represent 60% of human infectious diseases and 75% of the emerging ones; 80% of pathogens of animal origin have strong potential as bioterrorism agents [ 1 ]. Deforestation and destruction of natural areas produce promiscuity, pushing wild species to invade new areas and to arrive in anthropic environments. In high-income countries, domesticated animals are as much a potential reservoir of high-risk zoonoses as the wildlife animals in equatorial rainforests or wet markets. Companion and zoo animals—with limited syndromic monitoring in place—remain an underestimated but potentially high-risk disease reservoir for emerging zoonoses [ 2 ]. Through the global commercialisation of food, food-borne zoonoses (FZs) can also reach individuals who have never been in contact with infected animals or their environment. FZs are transmitted to human beings indirectly, both through food obtained from infected animals—which are contaminated at their origin—and through food previously contaminated in the various steps of production, sale, and domestic use. Once infected, consumers generally become a source of infection for animals and humans, as well as a source of contamination for food and the environment. Human beings, animals, and the environment constitute a cohesive and inextricable system, in which human and animal health are interdependent and linked to the health of the ecosystem in which they live. Therefore, they must be considered under the framework of One Health (OH) [ 1 ]. Since organic farming contributes to environmental and climate protection, long-term soil fertility, high levels of biodiversity, a safe environment, and high animal welfare standards [ 3 ], the European Commission (EC) has set a target of at least 25% of the European Union's (EU's) agricultural land being under organic farming by 2030 [ 4 ]. To achieve this goal and help organic agriculture reach its maximum potential, the EC proposes an action plan for organic production in the EU [ 3 ]. Then, improvement of human health can be achieved through better environmental conditions and healthier food. The EU's organic logo gives a coherent visual identity to organic products produced in the EU. This makes it easier for consumers to identify EU organic products, and helps farmers to market them [ 5 ]. Thus, OH overlaps the European Green Deal plan and its relaunched Farm to Fork Strategy. Nevertheless, zoonoses and animal infectious diseases cause decreased breeding yields and reduced income. Consequently, the cost and market price of farm products become uncompetitive with respect to the price of industrial food. In other words, zoonoses cause lower revenues, hindering the growth of organic farming expected in the framework of the EU Green Deal. In such scenarios, zoonosis control becomes a key element to align EU policies aimed at achieving the goal of "ZERO environmental impact" by 2050. 2. One Health The OH approach is adopted in world policies by: - The World Health Organization (WHO); - The Food and Agriculture Organization of the United Nations (FAO) [ 6 ]; - The World Organisation for Animal Health (OIE) [ 1 ]; - The European Union (EU) [ 7 ]; - The USA, where the Centers for Disease Control and Prevention (CDC—Atlanta, GA) host the National Center for Emerging and Zoonotic Infectious Diseases, which works "to protect people at home and around the world from emerging and zoonotic infections ranging from A to Z—anthrax to Zika—since we are living in an interconnected world where an outbreak of infectious disease is just a plane ride away" [ 8 ]. Recently, the Global Health Summit, held in Rome in May 2021, stated the need to adopt the OH approach in the Rome Declaration , issued at the conclusion of the summit [ 9 ]. On 12/12/2021, the FAO, OIE, WHO, and the United Nations Environment Programme (UNEP) adopted the following definition: "One Health is an integrated, unifying approach that aims to sustainably balance and optimize the health of people, animals and ecosystems. It recognizes the health of humans, domestic and wild animals, plants, and the wider environment (including ecosystems) are closely linked and interdependent. The approach mobilizes multiple sectors, disciplines and communities at varying levels of society to work together to foster well-being and tackle threats to health and ecosystems, while addressing the collective need for clean water, energy and air, safe and nutritious food, taking action on climate change, and contributing to sustainable development". Although health, food, water, the environment, and energy are very broad topics with specific and sectoral concerns, interdisciplinary and cross-sector collaborations could provide more effective tools both to protect us from emerging infectious diseases and antimicrobial resistance, and to promote the biodiversity and integrity of ecosystems. Furthermore, OH can help to address the full spectrum of disease control—from prevention to detection (a lesson learned during the ongoing COVID-19 pandemic), through preparedness, response, and management—and to improve and promote health. The OH approach can be applied at the national, community, regional, and global levels, and relies on shared and effective governance, communication, collaboration, and coordination. Using this approach will make it easier for people to better understand the co-benefits, risks, trade-offs, and opportunities to promote equitable and holistic solutions [ 10 ]. Firstly, the One Health High-Level Expert Panel (OHHLEP) advisory board focuses on preparing a policy-relevant scientific assessment of emerging health crises arising from the human–animal–ecosystem interface, on developing a long-term strategic approach to reduce the risk of zoonotic pandemics—with an associated monitoring and early warning framework—and on the synergies necessary to improve and institutionalise the OH approach, as well as in the areas that drive the risk of pandemics [ 11 ]. The WHO, OIE, and FAO work together to control and prevent health risks arising from the human–animal–ecosystem interface. They are developing global strategies and tools to ensure a coherent and harmonised approach worldwide, and to better coordinate national and international human, animal, and environmental health policies. In October 2017, the second Tripartite Strategy document was published, with a particular focus on the following aspects: strengthening national services for human and animal health and food safety; strengthening and modernisation of early warning and surveillance/monitoring systems; prediction, preparedness, and response to emerging, re-emerging, and neglected infectious diseases; encouraging and promoting coordinated research to achieve a common understanding of the highest-priority zoonotic diseases; and challenges posed by food safety that require a multisectoral approach in the context of improving food security [ 12 ]. The One Health Surveillance Codex (OHS Codex) was established to provide a framework for the OH community to constantly share practical solutions applicable for stakeholders from different One Health Surveillance sectors. The OHS Codex framework includes four high-level action principles, which support, respectively, collaboration, knowledge exchange, data interoperability, and dissemination, which are summarized as follows. (i) The collaboration principle represents the need for positive interaction and communication between actors, and it is considered to be the foundation of OH surveillance; without the willingness and ability to collaborate and communicate, surveillance systems will remain sectorial and fragmented. This principle collects tools and resources that facilitate the understanding of surveillance across sectors and disciplines, whilst also showing that these solutions can be found in environments where collaboration works. (ii) The knowledge principle represents the need for mutual scientific understanding and expertise in the evolution of knowledge of other sectors; the aim of this principle is to provide surveillance professionals and stakeholders with guidance on sources of knowledge. (iii) The data principle represents the ability to understand, reuse, and interpret surveillance data across sectors and disciplines. (iv) The dissemination principle represents the distribution of surveillance results to stakeholders and surveillance actors, including industries and policymakers [ 13 ]. The EC's Green Deal is an integral part of the OH strategy to implement the United Nations' Agenda 2030 and the Sustainable Development Goals (SDGs), which are an urgent and universal call for action by all high- and low-income countries. The EU will work with all partner countries to increase climate and environmental resilience to prevent these challenges from becoming sources of conflict, food insecurity, and forced migration. Some parts of the EU Green Deal, such as the Farm to Fork Strategy and the Biodiversity Strategy, explicitly refer to OH. In recent decades, the OH concept has expanded from the medical and veterinary sciences to include a rapidly growing range of synergistic disciplines, including food safety, food security, public health, health economics, environmental ecosystem health, social sciences, and animal health and welfare. It is now recognised that environmental factors—including chemical contaminants in animals and animal products, veterinary drug residues, and pesticides—play significant roles that call for holistic transdisciplinary approaches to move towards safe and sustainable food systems [ 14 ]. 3. Food-Borne Zoonoses and EU Animal Health Laws Zoonosis control in the EU is regulated by Directive 2003/99/EC [ 15 ], which lists in LIST A the zoonoses to be subjected to mandatory control, including the main FZs, such as brucellosis and tuberculosis in cattle and buffaloes, salmonellosis in poultry and turkeys, and trichinellosis. The individual Member States (MSs) activate National Control Plans (NCPs) in primary production. NCPs are mandatory and possibly co-financed by the European Commission [ 16 ]. They are harmonised in order to make the results comparable, thanks to methods of analysis developed, validated, and disseminated by the European Union Reference Laboratories (EURLs) [ 17 ], which transfer them to individual National Reference Laboratories (NRLs) of each Member State (MS) which, in turn, disseminate them extensively to the laboratories of their own national territory. Other optional plans can be activated, based on the epidemiological situation of specific territories as regards the zoonoses included in LIST B of Directive 2003/99/EC. The NCPs are based on the following aspects: diagnosis (serological diagnosis is adopted if there is no possibility of taking samples useful for direct diagnosis); identification and elimination of infected animals or of the entire herd hosting them; attribution of the sanitary qualification "Officially Free" (OF) to breeding and, progressively, to the entire province or region and to the MS in which the specific zoonosis has been eradicated; the prohibition of vaccination (generally mandatory); biosecurity measures, which must be adopted in a strict way, because the presence of the NCP causes the absence of natural or artificial immunological responses against the specific zoonotic agent, due to the pathogen-free territory and to the prohibition of vaccination, respectively. While vaccination can be authorised in the event of a serious emergency with attribution of the sanitary qualification "Free" (F), antibiotic therapy is not allowed. In case of vaccine authorisation, the OF qualification can be newly acquired when only non-vaccinated animals are present in livestock. NCPs are divided into "eradication plans", if the zoonosis is present in the animal population, and "surveillance plans", which are limited to checking whether the eradicated zoonosis is re-emerging. The surveillance, in turn, is divided into active surveillance, which involves official controls on farms aimed at identifying infected animals, and passive surveillance, which induces the activation of official controls only when suspected cases are signalled. Data relating to the results obtained in primary production under the application of NCPs, reported human cases, test results in food control activities, test results in feed control activities, and antimicrobial resistance converge in The European Union One Health Zoonoses Report , published jointly by the European Food Safety Authority (EFSA) and the European Centre for Disease Prevention and Control (ECDC), in open access and on an annual basis. In the report published in 2021 [ 18 ], relating to data for the year 2020, Salmonella (Le Minor and Popoff, 1987) [ 19 ] remains the most frequently reported etiological agent in episodes of FZs in the EU, and the pathogens considered in relation to the foods of greatest risk were found to be Salmonella in eggs and their derivatives, norovirus in crustaceans and molluscs (including bivalves), and Listeria monocytogenes (Murray, Webb, and Swan, 1926) [ 19 ] in fish and fish products. Correlation between human brucellosis and non-OF territories persists [ 18 ]. Regarding Salmonella infections, on 17 February 2022, the United Kingdom (UK) reported a cluster of cases with monophasic Salmonella Typhimurium sequence type 34 infection. As of 18 May 2022, 324 cases had been reported in 12 EU/EEA countries and the UK, including two distinct strains. As of 3 June 2022, 392 cases of monophasic S . Typhimurium have been identified in the EU/EEA and the UK ( n = 370 confirmed cases and n = 22 probable cases). In addition, cases have been identified in Canada ( n = 4), Switzerland ( n = 48), and the United States ( n = 1), bringing the total number of cases to 445 globally [ 20 , 21 ]. Most cases were detected in persons below 10 years of age, and 41% of all cases were hospitalised. The two strains are multidrug-resistant, and some tested isolates also exhibit resistance to disinfectants based on quaternary ammonium compounds and hydrogen peroxide, but are susceptible to azithromycin, ciprofloxacin, meropenem, and third-generation cephalosporins. Epidemiological investigations have suggested some specific chocolate products, produced in a plant in Belgium, as likely vehicles of infection. Two monophasic Salmonella Typhimurium strains matching the outbreak strains were identified in the buttermilk line at the Belgian plant between December 2021 and January 2022. The buttermilk was provided by an Italian supplier, where S. Typhimurium was not detected. On 8 April 2022, on the basis of official controls, the Belgian food safety authority decided to revoke the production authorisation of the indicated plant due to lack of transparency and insufficient guarantees for safe production. All at-risk products produced at the closed plant have been recalled. National competent authorities in several countries issued public warnings. This outbreak has evolved rapidly, with children most at risk of severe infection. The plant closure and the global recall of all potentially hazardous products have reduced the risk of exposure. However, eight cases cannot be explained by consumption of chocolate products, suggesting that there may also be other sources of infection. The ECDC has published, in open access, the data on antimicrobial resistance [ 22 ] and, jointly with EFSA, The European Union Summary Report on Antimicrobial Resistance in Zoonotic and Indicator Bacteria from Humans, Animals and Food in 2018/2019 [ 23 ]. 3.1. EU Control Programmes EU co-funded veterinary programmes have proven to be a catalyst for achieving improvements in public and animal health, reductions in disease prevalence/incidence, safeguarding of public health (in the case of zoonoses), disease prevention/management in the context of the EU Animal Health Strategy, and economic benefits for the EU as whole by protecting the value of the sector, contributing to market stability, guaranteeing safe trade, increasing extra-EU trade, and reducing human health costs. Significant differences in MSs' veterinary systems and livestock facilities lead to variability in the implementation of programmes, risking jeopardising the results achieved at the EU level—particularly when dealing with transboundary diseases [ 24 ]. Regulation (EU) No 652/2014 laying down provisions for the management of expenditure relating to the food chain, animal health, and animal welfare concerns diseases with impacts on human health, diseases with impacts on animal health (taking into consideration their potential spread and the morbidity and mortality rates in animal populations), diseases and zoonoses that risk being introduced and/or re-introduced into the EU territory from third-party countries, diseases with the potential to generate a crisis situation with serious economic consequences, and diseases with impacts on trade with third-party countries and on intra-EU trade. It also concerns the main FZs, including bovine tuberculosis, bovine brucellosis, echinococcosis, campylobacteriosis, listeriosis, salmonellosis, trichinellosis, and verotoxigenic Escherichia coli (Castellani and Chalmers, 1919) infections [ 19 , 25 ]. Regulation (EU) 2016/429 on transmissible animal diseases and amending and repealing certain acts in the area of animal health ("Animal Health Law") [ 26 ] concerns the main FZs, including infection with Brucella (Meyer and Shaw, 1920)—specifically, B. abortus , B. melitensis , and B. suis ; infection with Mycobacterium bovis (Karlson and Lessel, 1970), M. caprae , and M. tuberculosis , included in the Mycobacterium tuberculosis complex; and infestation with Echinococcus multilocularis (Leuckart, 1863) [ 19 , 27 ]. It was amended and corrected in 2017, 2018, and 2020 [ 28 ]. In 2020, in Northern Europe, 20 MSs were officially brucellosis-free in cattle, and 17 MSs were officially tuberculosis-free in cattle, while these zoonoses persisted in the Mediterranean area. Italy, Portugal, and Spain activated co-funded eradication programmes for bovine brucellosis as well as for bovine tuberculosis (also activated in Ireland and Malta) [ 26 ], while in Greece only the eradication programme concerning ovine and caprine brucellosis ( B. melitensis ) was co-funded [ 20 , 29 ]. Greece reported the highest prevalence of Brucella -positive ruminant herds, and Spain reported the highest prevalence of tuberculosis in cattle. 3.2. Control of Non-Regulated Diseases in Cattle and Buffalo (Cattle Diseases Listed under Category C, D, or E in the EU Animal Health Law) NCPs are in force both for the most important zoonoses and for the most important animal infectious diseases that lack zoonotic potential. For diseases not included in the European Union Animal Health Law Categories A or B under Commission Implementing Regulation (EU) 2020/2002 [ 28 ], approximately one-third of control plans (CPs) are voluntary and can be limited to a well-defined territory of the MS; their funding structure is divided between government and private resources [ 29 ]. Countries that have already eradicated diseases such as enzootic bovine leukosis, bluetongue, infectious bovine rhinotracheitis, and bovine viral diarrhoea have also implemented CPs for other diseases in order to further improve the health status of cattle in their country [ 30 ], increasing the commercial value of animals and animal products. Consequently, the gap in the health status of farmed animals could progressively increase among the EU MSs. 3.1. EU Control Programmes EU co-funded veterinary programmes have proven to be a catalyst for achieving improvements in public and animal health, reductions in disease prevalence/incidence, safeguarding of public health (in the case of zoonoses), disease prevention/management in the context of the EU Animal Health Strategy, and economic benefits for the EU as whole by protecting the value of the sector, contributing to market stability, guaranteeing safe trade, increasing extra-EU trade, and reducing human health costs. Significant differences in MSs' veterinary systems and livestock facilities lead to variability in the implementation of programmes, risking jeopardising the results achieved at the EU level—particularly when dealing with transboundary diseases [ 24 ]. Regulation (EU) No 652/2014 laying down provisions for the management of expenditure relating to the food chain, animal health, and animal welfare concerns diseases with impacts on human health, diseases with impacts on animal health (taking into consideration their potential spread and the morbidity and mortality rates in animal populations), diseases and zoonoses that risk being introduced and/or re-introduced into the EU territory from third-party countries, diseases with the potential to generate a crisis situation with serious economic consequences, and diseases with impacts on trade with third-party countries and on intra-EU trade. It also concerns the main FZs, including bovine tuberculosis, bovine brucellosis, echinococcosis, campylobacteriosis, listeriosis, salmonellosis, trichinellosis, and verotoxigenic Escherichia coli (Castellani and Chalmers, 1919) infections [ 19 , 25 ]. Regulation (EU) 2016/429 on transmissible animal diseases and amending and repealing certain acts in the area of animal health ("Animal Health Law") [ 26 ] concerns the main FZs, including infection with Brucella (Meyer and Shaw, 1920)—specifically, B. abortus , B. melitensis , and B. suis ; infection with Mycobacterium bovis (Karlson and Lessel, 1970), M. caprae , and M. tuberculosis , included in the Mycobacterium tuberculosis complex; and infestation with Echinococcus multilocularis (Leuckart, 1863) [ 19 , 27 ]. It was amended and corrected in 2017, 2018, and 2020 [ 28 ]. In 2020, in Northern Europe, 20 MSs were officially brucellosis-free in cattle, and 17 MSs were officially tuberculosis-free in cattle, while these zoonoses persisted in the Mediterranean area. Italy, Portugal, and Spain activated co-funded eradication programmes for bovine brucellosis as well as for bovine tuberculosis (also activated in Ireland and Malta) [ 26 ], while in Greece only the eradication programme concerning ovine and caprine brucellosis ( B. melitensis ) was co-funded [ 20 , 29 ]. Greece reported the highest prevalence of Brucella -positive ruminant herds, and Spain reported the highest prevalence of tuberculosis in cattle. 3.2. Control of Non-Regulated Diseases in Cattle and Buffalo (Cattle Diseases Listed under Category C, D, or E in the EU Animal Health Law) NCPs are in force both for the most important zoonoses and for the most important animal infectious diseases that lack zoonotic potential. For diseases not included in the European Union Animal Health Law Categories A or B under Commission Implementing Regulation (EU) 2020/2002 [ 28 ], approximately one-third of control plans (CPs) are voluntary and can be limited to a well-defined territory of the MS; their funding structure is divided between government and private resources [ 29 ]. Countries that have already eradicated diseases such as enzootic bovine leukosis, bluetongue, infectious bovine rhinotracheitis, and bovine viral diarrhoea have also implemented CPs for other diseases in order to further improve the health status of cattle in their country [ 30 ], increasing the commercial value of animals and animal products. Consequently, the gap in the health status of farmed animals could progressively increase among the EU MSs. 4. The EU Green Policies concerning the Food System In the European Union, OH overlaps the European Green Deal plan launched by the European Commission (EC Green Deal plan) to achieve the goal of "ZERO environmental impact" by 2050. In this scenario, the devastating impact of war must be considered ( Figure 1 ). The Green Deal relaunches the Farm to Fork Strategy for a healthy and environmentally sustainable food system, providing specific measures to make the economy circular, while concomitantly reducing the use of pesticides, fertilisers, and antibiotics, so as to limit the alarming antimicrobial resistance that is spreading worldwide [ 3 , 4 , 31 ]. 4.1. Public Engagement Public engagement can be stimulated through the improvement of food labelling, which should include information concerning the production environment, aimed at facilitating consumers' choices in the direction of healthy and, at the same time, sustainable diets. A reward mechanism is triggered for farms that adopt the circular economy and that produce in compliance with the objectives of the EU Green Deal. On the other hand, consumers receive beneficial effects on their health, in terms of both food safety and the improvement of environmental conditions [ 32 ]. Labelling is a key factor for food safety in agrifood chains. It is often characterised by asymmetric information. Producers and marketers tend to be better informed than consumers about the potential risks of food. The use of innovative strategies to communicate information on food risks can help reduce the divergence between assessed and perceived risks. In this respect, innovative labels—such as traffic-light labels or the use of nanotechnologies—could be valid alternatives. Furthermore, technologies such as Agri-Food 4.0, Blockchain, and the Internet of Things can be useful tools to inform consumers in real time, while also supporting the supply chain decision-making process and improving the coordination process involving farmers, industries, and consumers [ 33 ]. 4.2. Fighting the Antimicrobial Resistance The Green Deal includes—among the objectives of primary importance—the reduction in the use of antibiotics in livestock production, in order to combat antimicrobial resistance (AMR), which is a global emergency that has increased during the COVID-19 pandemic. This is referred to as the silent, second pandemic [ 34 ]. In addition to the wide use of antimicrobials in animals [ 35 ], plant agriculture frequently uses antibiotics to enhance crop yields. This means that fruits and vegetables have also become potential sources of AMR [ 36 ]. Multidrug resistance (MDR) is continuously expanding worldwide, and poses a challenge in treating infections, necessitating the use of reserve antibiotics, which can have higher cost-to-benefit ratios and a lower safety profile. Among the MDR germs, "the ESKAPE pathogens" have had the greatest impact on healthcare-associated infections—a group of six pathogens with the capacity to elude the bactericidal activity of antibiotics: Enterococcus faecium (Orla-Jensen, 1919), Staphylococcus aureus (Rosenbach, 1884), Klebsiella pneumoniae (Trevisan, 1887), Acinetobacter baumannii (Bouvet and Grimont, 1986), Pseudomonas aeruginosa (Migula, 1900), and Escherichia coli [ 19 ]. The ESKAPE group is characterised by pathogenic and transmission, resistance traits—which are represented by enzymatic inactivation, target changes, and alteration of cell permeability through loss of porins or increased expression of efflux pumps—and mechanical protection through biofilm formation [ 37 ]. In 2019, an estimated 1.27 million deaths were attributable to bacterial AMR. At the regional level, the all-age death rate attributable to resistance was highest in western sub-Saharan Africa, and lowest in Australasia. Lower respiratory tract infections accounted for over 1.5 million resistance-associated deaths in 2019, making this the most burdensome infectious syndrome. The six main pathogens for resistance-associated deaths— E. coli , followed by S. aureus , K. pneumoniae , Streptococcus pneumoniae (Chester, 1901), A. baumannii , and Pseudomonas aeruginosa —were responsible for more than 900,000 deaths attributable to AMR in 2019 [ 19 ]. One pathogen–drug combination—methicillin-resistant S. aureus —caused more than 100,000 deaths attributable to AMR in 2019, while six others each caused between 50,000 and 100,000 deaths: multidrug-resistant (excluding extensively drug-resistant) tuberculosis, E. coli resistant to third-generation cephalosporins and to fluoroquinolones, A. baumannii and K. pneumoniae resistant to carbapenems, and K. pneumoniae resistant to third-generation cephalosporins [ 38 ]. In the EU food system, the AMR monitoring and reporting cover the following food-producing animal populations and foods: broilers; laying hens; fattening turkeys; cattle less than one year old; fattening pigs; fresh meat from broilers; and fresh meat from turkeys, pigs, and cattle. AMR surveillance concerns the monitoring and reporting of antimicrobial resistance of the following bacteria: Salmonella spp., Campylobacter coli , C. jejuni (Doyle, 1948; Véron and Chatelain, 1973), indicator commensal E. coli , Salmonella spp., and E. coli producing extended-spectrum β-lactamases, AmpC β-lactamases, and carbapenemases. In addition, it may cover indicator commensal E. faecalis and E. faecium [ 19 , 39 ]. In the United States of America, Congress issued the Disarm Act of 2021 [ 40 ], in order to develop an innovative strategy to fight the increase in antimicrobial resistance [ 41 ]. 4.1. Public Engagement Public engagement can be stimulated through the improvement of food labelling, which should include information concerning the production environment, aimed at facilitating consumers' choices in the direction of healthy and, at the same time, sustainable diets. A reward mechanism is triggered for farms that adopt the circular economy and that produce in compliance with the objectives of the EU Green Deal. On the other hand, consumers receive beneficial effects on their health, in terms of both food safety and the improvement of environmental conditions [ 32 ]. Labelling is a key factor for food safety in agrifood chains. It is often characterised by asymmetric information. Producers and marketers tend to be better informed than consumers about the potential risks of food. The use of innovative strategies to communicate information on food risks can help reduce the divergence between assessed and perceived risks. In this respect, innovative labels—such as traffic-light labels or the use of nanotechnologies—could be valid alternatives. Furthermore, technologies such as Agri-Food 4.0, Blockchain, and the Internet of Things can be useful tools to inform consumers in real time, while also supporting the supply chain decision-making process and improving the coordination process involving farmers, industries, and consumers [ 33 ]. 4.2. Fighting the Antimicrobial Resistance The Green Deal includes—among the objectives of primary importance—the reduction in the use of antibiotics in livestock production, in order to combat antimicrobial resistance (AMR), which is a global emergency that has increased during the COVID-19 pandemic. This is referred to as the silent, second pandemic [ 34 ]. In addition to the wide use of antimicrobials in animals [ 35 ], plant agriculture frequently uses antibiotics to enhance crop yields. This means that fruits and vegetables have also become potential sources of AMR [ 36 ]. Multidrug resistance (MDR) is continuously expanding worldwide, and poses a challenge in treating infections, necessitating the use of reserve antibiotics, which can have higher cost-to-benefit ratios and a lower safety profile. Among the MDR germs, "the ESKAPE pathogens" have had the greatest impact on healthcare-associated infections—a group of six pathogens with the capacity to elude the bactericidal activity of antibiotics: Enterococcus faecium (Orla-Jensen, 1919), Staphylococcus aureus (Rosenbach, 1884), Klebsiella pneumoniae (Trevisan, 1887), Acinetobacter baumannii (Bouvet and Grimont, 1986), Pseudomonas aeruginosa (Migula, 1900), and Escherichia coli [ 19 ]. The ESKAPE group is characterised by pathogenic and transmission, resistance traits—which are represented by enzymatic inactivation, target changes, and alteration of cell permeability through loss of porins or increased expression of efflux pumps—and mechanical protection through biofilm formation [ 37 ]. In 2019, an estimated 1.27 million deaths were attributable to bacterial AMR. At the regional level, the all-age death rate attributable to resistance was highest in western sub-Saharan Africa, and lowest in Australasia. Lower respiratory tract infections accounted for over 1.5 million resistance-associated deaths in 2019, making this the most burdensome infectious syndrome. The six main pathogens for resistance-associated deaths— E. coli , followed by S. aureus , K. pneumoniae , Streptococcus pneumoniae (Chester, 1901), A. baumannii , and Pseudomonas aeruginosa —were responsible for more than 900,000 deaths attributable to AMR in 2019 [ 19 ]. One pathogen–drug combination—methicillin-resistant S. aureus —caused more than 100,000 deaths attributable to AMR in 2019, while six others each caused between 50,000 and 100,000 deaths: multidrug-resistant (excluding extensively drug-resistant) tuberculosis, E. coli resistant to third-generation cephalosporins and to fluoroquinolones, A. baumannii and K. pneumoniae resistant to carbapenems, and K. pneumoniae resistant to third-generation cephalosporins [ 38 ]. In the EU food system, the AMR monitoring and reporting cover the following food-producing animal populations and foods: broilers; laying hens; fattening turkeys; cattle less than one year old; fattening pigs; fresh meat from broilers; and fresh meat from turkeys, pigs, and cattle. AMR surveillance concerns the monitoring and reporting of antimicrobial resistance of the following bacteria: Salmonella spp., Campylobacter coli , C. jejuni (Doyle, 1948; Véron and Chatelain, 1973), indicator commensal E. coli , Salmonella spp., and E. coli producing extended-spectrum β-lactamases, AmpC β-lactamases, and carbapenemases. In addition, it may cover indicator commensal E. faecalis and E. faecium [ 19 , 39 ]. In the United States of America, Congress issued the Disarm Act of 2021 [ 40 ], in order to develop an innovative strategy to fight the increase in antimicrobial resistance [ 41 ]. 5. Discussion NCPs and CPs for animal diseases provide benefits for animals, farmers, producers, and consumers, because they improve animal health and welfare, reduce the use of antibiotics, and—in the case of zoonotic diseases—improve the safety of animal products. NCPs and CPs reduce direct and indirect losses due to diseases. Their implementation involves associated costs for testing and administrative work; however, the cost is usually considered to be outweighed by the benefits [ 30 ]. Nevertheless, in non-OF territories, zoonoses and animal infectious diseases cause lower productivity and often lower reproductive capacity of infected farm animals. The decreased breeding yield produces increased cost of farm products that, in turn, induce higher and uncompetitive market prices, e.g., in the framework of Regulation (EC) 853/2004, raw milk must come from cows or buffaloes belonging to a herd that is OF or F from brucellosis. In non-OF nor F herds, raw milk may still be used when coming from cows or buffaloes that do not show a positive reaction to tests for brucellosis, nor have any symptoms of the disease. In this situation it is mandatory to have the authorisation of the competent authority in order to make the milk undergo a compulsory heat treatment useful to reveal a negative reaction to the phosphatase test. The cost of additional analyses are generally charged to farmers and, consequently, most of them prefer to discard this milk. Furthermore, infected livestock or infected heads are culled in the frame of zoonosis and animal infectious disease NCPs, and the cost of restocking is not fully covered by indemnities. Thus, zoonoses and animal infectious diseases cause decreased revenues, hindering the growth of organic farms, and hampering the fulfilment of the EU Green Deal objectives [ 42 ]. In 2020, EU data on zoonoses and related zoonotic agents reported a drastic decrease in the numbers of human cases [ 18 ]. This was evidently influenced not only by the UK's exit from the EU, but also by the COVID-19 pandemic and the resulting restriction measures imposed in EU MSs. This worldwide event, which today sees the rise of variants of concern (VOCs) of SARS-CoV-2 in selected countries with unconsolidated economies and in which the administration of vaccines is slow, highlights the importance of the OH approach, which is useful in designing a new health system to be applied homogeneously at a global scale—not only as an ethical necessity, but also as an indispensable safeguard to prevent that those who have been left behind from becoming victims and, at the same time, sources of new emergencies [ 43 ]. The Economic Community of West African States (ECOWAS) has borne a significant burden of zoonotic disease impacts. To address zoonotic disease threats in ECOWAS, a One Health Zoonotic Disease Prioritization (OHZDP) was conducted in December 2018 to prioritise the zoonotic diseases of greatest regional concern and develop the next steps to address these priority zoonoses through a regional, multisectoral, OH approach. ECOWAS was the first region to use the OHZDP process to prioritise zoonotic diseases of greatest concern. With the identification of priority zoonotic diseases for the region, ECOWAS member states can collaborate more effectively to address zoonotic disease threats across the region using a OH approach. Strengthening national- and regional-level multisectoral OH coordination mechanisms allow ECOWAS member states to advance OH, and has a significant impact on improving health outcomes for both people and animals living in a shared environment [ 44 ]. Since it is not possible to stem the spread of pathogens by constructing disjointed barriers, it becomes imperative to act simultaneously at all levels and on a global scale, financing actions that guarantee uniform conditions of protection. This is necessary to counteract the enormous biological plasticity of microorganisms which, together with their rapid reproduction cycles, makes them easily globalised. In the EU, to limit the spread of zoonoses reported as official cases, actions based on the essential protection of the environment are necessary through the concrete implementation of the Farm to Fork Strategy and the pursuit of the Green Deal; in fact, important zoonoses included in Directive 2003/99 EC, although originally linked to animal reservoirs, can be found today in vegetable foods to be consumed raw, as preferential and hazardous vehicles of transmission to humans [ 45 ]. Furthermore, animal diseases threaten global food security. African swine fever (ASF)—a viral haemorrhagic disease characterised by high morbidity and high mortality in domestic and wild swine (but lacking zoonotic potential)—is progressively expanding in Asia and Europe [ 46 ]. In particular, in 2017, ASF cases increased in wild boar in Ukraine, [ 47 ]; in 2021, ASF continued to be reported in wild boar across Europe, and frequent outbreaks in domestic pigs continued to be reported in Romania, with small numbers of outbreaks also reported in Ukraine, which is currently under military attack [ 48 ]. The probable escape of infected wild boars from territories exposed to military attacks or human factors due to the consequences of war (e.g., population displacement) could cause the further spread of outbreaks in neighbouring EU MSs. This is only a minor consequence of the tragedy of the ongoing war, which sees human lives sacrificed and the mass exodus of the exhausted Ukrainian human population, in whom 32,000 new tuberculosis (TB) cases have been estimated in 2020 (with almost 11,000 cases estimated to be drug-resistant TB), and patients have had to stop their required lengthy treatment [ 49 ]. ASF has been present on the island of Sardinia since 1978, without consequences for the Italian mainland, where ASF was later on confirmed on 6 January 2022 in wild boars; the ongoing outbreaks first spread in the regions of Piedmont and Liguria (north-west Italy), and then reached the region of Lazio (Central Italy). Control measures at the event level, including domestic control measures in one domestic outbreak that occurred in a farm hosting nine swine (i.e., disinfection, ante and post mortem inspections; official disposal of carcasses, byproducts, and waste; zoning; traceability; surveillance within the restricted zone; surveillance outside the restricted zone; stamping out; screening; official destruction of animal products; movement control) and wild control measures (i.e., screening; official disposal of carcasses, byproducts, and waste; movement control; surveillance outside the restricted zone; ante and post mortem inspections; official destruction of animal products; surveillance within the restricted zone; zoning) were put in place to avoid the major zooeconomic problems caused by ASF—in particular in the export of fine Italian delicatessen products [ 50 ]. The African swine fever virus (ASFV), in the genus Asfivirus [ 51 ], does not induce neutralising antibodies, making difficult the efforts for the production of vaccines. A new candidate vaccine for ASF has been developed recently. It uses an attenuated whole virus which, by inducing antibodies towards many viral antigens, is able to protect swine against infection, but may present the risk of retromutation to the virulent virus. Therefore, the "reversion to virulence" test is an important milestone as part of a series of safety studies. Recently, the U.S. Department of Agriculture's Agricultural Research Service (ARS) announced that a vaccine candidate for ASF had passed an important safety test required for regulatory approval, moving the vaccine one step closer to commercial availability [ 52 ]. Furthermore, Ukraine remains the only country in Europe where rabies is widespread, with about 1600 rabies cases in animals and sporadic cases in humans [ 53 ]. Rabies is the deadliest of all known zoonoses and it is lethal to mammals. It negatively impacts on food security and livelihood [ 54 ]. Veterinary bodies in countries bordering Ukraine and in other MSs have made exceptions to peacetime restrictions for bringing pets across borders in a bid to aid refugees. Consequently, the risk of introducing rabies in free countries is feared [ 55 ]. In Asian countries (e.g., Cambodia, Indonesia, Vietnam), where dogs and cats are used for human consumption, rabies can be considered a food-related zoonosis in workers at dog and cat slaughterhouses and it can be considered a food-borne zoonosis for the consumers of infected dogs and cats. In Vietnam, workers at dog slaughterhouses are vaccinated in the framework of a national programme for rabies control and prevention [ 56 ]. There is a strong policy impetus for the OH cross-sectoral approach to address the complex challenge of zoonotic diseases—particularly in low/lower- and middle-income countries (LMICs), where there is limited policy visibility on zoonotic diseases—especially high-burden endemic diseases that disproportionately affect marginalised rural populations [ 57 ]. Since 1 January and as of 22 June 2022, 3413 laboratory-confirmed cases and 1 death from monkeypox have been reported to the WHO from 50 countries/territories in 5 WHO regions. In a week, 1310 new cases were reported, and eight new countries reported cases. As of 21 July 2022, 15,848 confirmed cases had been reported. The unexpected appearance of monkeypox and the wide geographic spread of cases indicate that the monkeypox virus might have been circulating below levels detectable by the surveillance systems, and that sustained human-to-human transmission might have been undetected for a period of time [ 58 , 59 ]. In 2020, a total of 4594 suspected cases of monkeypox, including 171 deaths (case fatality rate = 3.7%), have been reported in 127 health zones from 17 out of 26 provinces in the Democratic Republic of the Congo. Communicating monkeypox-related risks and engaging at-risk and affected communities, community leaders, civil society organisations, and healthcare providers—including those at sexual health clinics—in prevention, detection, and care, is essential for preventing further secondary cases and ensuring effective management of the current outbreak [ 60 ]. The WHO has been considering declaring monkeypox a Public Health Emergency of International Concern (PHEIC). In 2019, the U.S. Food and Drug Administration (FDA) approved a smallpox and monkeypox vaccine. This is a good example of preparedness [ 61 ]. The recent "One Health of Peripheries" proposal highlights violence as a cause of morbidity and mortality, and among the approaches to address its complexity is the prevention of violence against animals. Geographical peripheries are heterogeneous, encompassing entire countries, areas circumscribed within countries, cross-border regions, rural areas, indigenous territories, and favelas. In particular, the contextual effects of favelas on health are mediated by imposed risks and the lack of resources (e.g., money, time, infrastructure, knowledge), creating a vicious circle of vulnerability due to the increased burden of diseases that compromise individuals' opportunities for economic and social inclusion. The contextual effects of favelas affect multispecies collectives, and this aspect is even more neglected. Animals are exposed and vulnerable to pollution, humidity, darkness, inadequate ventilation, malnutrition, and high population density. It is necessary to promote the health of animals for their own sake, but also for the sake of the humans living with them [ 62 ]. Collaboration must be tailored to the surveillance objective and context, characterised by a wide range of factors (i.e., epidemiological, ecological, economic, social, and environmental); successful cross-sectoral collaboration is largely rooted in mutual trust and respect between the different actors [ 57 ]. The need for a transition from traditional public health/biosurveillance to a health security intelligence approach to epi/pandemics requires delineating emerging threats from companion, livestock, and wildlife animal communities, which requires funds, the integration of early warning tools, open-source platforms, multisource and multispecies surveillance, proactive diagnostics, field testing technologies, and increased focus on necropsy in captive and wild animals [ 2 ]. Diagnostic tools—able to separate immunoglobulin isotypes in order to specifically detect and quantify them simultaneously—could be useful in serological diagnosis aimed at individuating emerging threats, their epidemiological features, and their evolution during vaccination campaigns [ 63 ]. Furthermore, OH education must be incorporated into scientific, engineering, and humanities curricula, in order to build capacity in OH skills, with the goal of creating networks that will work to improve public health, food safety, food security, and sustainable agriculture. This can be achieved by establishing new perspectives on the interactions between plants, animals, and humans, recognising the threat of disasters and transboundary diseases to food security. The implementation of OH calls for identifying priority areas for added value of joint activities, and for the effective knowledge elicitation of experts from different and relevant disciplines. Consequently, OH may call for updated models for establishing and maintaining effective and timely collaboration and communication across and within disciplines. The establishment of OH approaches and networks can be of high value for countries aiming to establish or improve their OH activities, supporting science-based regulations in the areas of health, food, and the environment [ 64 ]. 6. Conclusions Specific evaluation attributes need to be developed to allow the measurement of the impacts and benefits of collaborative surveillance versus a juxtaposition of isolated sectoral surveillance components [ 65 ]. In the EU, surveillance systems integrate elements of public and animal health systems and food chains to detect, assess, and control multistate food-borne infections. These systems are based on an investigation and notification approach (Early Warning and Response System-EWRS), the European surveillance portal for infectious diseases EpiPulse, and the Rapid Alert System for Food and Feed (RASFF). The exchange of information is facilitated and coordinated by the EFSA and ECDC, and is followed up by the EC [ 66 ]. General requirements for the laboratory and bioinformatic components of whole-genome sequencing (WGS) and associated metadata for food-borne bacteria have been recently published by the International Organization for Standardization. These guidelines provide the basis for harmonisation of bioinformatic analysis and validation of the end-to-end WGS workflows [ 66 , 67 ]. Despite food-borne zoonosis control needs large investments, it can positively affect economies, so the parameter economic health should be included in the OH approach in order to fund eradication programmes in EU MSs facing financial difficulties (heightened by the ongoing COVID-19 pandemic) and, globally, in extra-EU low-income countries, as well as to monitor the effectiveness of interventions and their benefit/cost ratio [ 42 ]. EU actions should include, in the case of economic needs, the financing of NCPs in MSs such as Greece which—despite the persistence of the highest notification rate of human cases for brucellosis in the EU, resulting in the highest number of confirmed domestically acquired cases in an MS, and the highest prevalence of Brucella -positive ruminant herds - does not benefit from the EC co-funded programme for brucellosis in cattle [ 18 , 42 ]. 7. Future Directions In 2012, for the first time, the World Bank highlighted the cumulative societal cost of infectious disease outbreaks. The moment at which a system is able to detect signals of possible threats was correlated with the possible cost of the system. Thus, the generation of high-quality zoonotic situational awareness at the earliest possible stages of an outbreak must be a priority for future health security systems [ 68 ]. Promoting the OH approach for food safety (including control of related zoonotic agents) and food security (including control of related animal and plant infectious diseases with no zoonotic potential) is a way to involve the younger generation in entering employment and training in agriculture and the food system. To this end, the obstacle represented by language barriers must be removed. Establishing a broader concept of OH that incorporates the food system as well as cultural and societal awareness is essential. Developing a holistic view of biomedical, biotechnological, and agricultural sciences can help in transforming traditional academics and researchers into OH practitioners who work in transdisciplinary teams to solve complex problems at the interface of human, animal, plant, and environmental health [ 64 ]. Given the unequal health status of EU farms with respect to both FZs and infectious animal diseases, it would be appropriate to introduce a new OH assessment parameter: economic health. This could be used to allocate European funds that would allow MSs facing economic difficulties to achieve eradication by reaching the same levels as the MSs with the most advanced animal husbandry. Recalling Dr. Bernard Url's conclusive speech in the ONE2022 conference, "We have to change the financial model in which we are locked" [ 69 ]. Since investors bet on innovation only considering its impact/success [ 70 ] and the extent of economic return, we need to build a way to direct investments toward ethical and science-based choices. National and international rules concerning public/private investments and trading could regulate the flow of money, directing it exclusively toward companies that meet a broad range of OH parameters, ranging from environmental impact up to food safety, human health, work conditions, duration of employment, number of layoffs per decade, wage conditions, gender and social equality, etc. A OH conformity certificate for enterprises—listed in the stock market or not—should be awarded considering various aspects of the OH, each of which should be analysed using a proper checklist. Each analysed checklist could then be merged in a final document declaring (or not) the compliance of a specific enterprise with a minimum number of the considered parameters useful to evaluate its OH status. This in-embryo One Health Financial Model is based on money exclusively invested in certified enterprises. In this model, money can be effectively used in the effort to improve human society and global conditions, avoiding economic losses due to negative side effects on the holistically considered health, in the effort to effectively pursue the OH goal ( Figure 2 ). Dedicated agencies and databanks preserving all data concerning certified enterprises are needed. Data—when shared in open access—could efficaciously improve public engagement, addressing consumers' choices toward certified producers who comply with the OH objectives. Despite the considerable efforts being carried out in the context of OH, war causes tragic and devastating effects on the physical and mental health of human beings; on their lives; on epidemic [ 71 ], pandemic, and zoonotic threats; on animals; on plants; and, last but not least, on the environment ( Figure 1 ). War is incompatible with One Health. Enormous efforts for peace are therefore urgently needed.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2655859/

A Recursive Algorithm for Spatial Cluster Detection

Spatial cluster detection involves finding spatial subregions of some larger region where clusters of some event are occurring. For example, in the case of disease outbreak detection, we want to find clusters of disease cases so as to pinpoint where the outbreak is occurring. When doing spatial cluster detection, we must first articulate the subregions of the region being analyzed. A simple approach is to represent the entire region by an n · n grid. Then we let every subset of cells in the grid represent a subregion. With this representation, the number of subregions is equal to 2 n 2 −1. If n is not small, it is intractable to check every subregion. The time complexity of checking all the subregions that are rectangles is θ(n 4 ). Neill et al. 8 performed Bayesian spatial cluster detection by only checking every rectangle. In the current paper, we develop a recursive algorithm which searches a richer set of subregions. We provide results of simulation experiments evaluating the detection power and accuracy of the algorithm.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4352578/

Crystal structure of Bacillus anthracis virulence regulator AtxA and effects of phosphorylated histidines on multimerization and activity

Summary The Bacillus anthracis virulence regulator AtxA controls transcription of the anthrax toxin genes and capsule biosynthesis operon. AtxA activity is elevated during growth in media containing glucose and CO 2 /bicarbonate, and there is a positive correlation between the CO 2 /bicarbonate signal, AtxA activity, and homomultimerization. AtxA activity is also affected by phosphorylation at specific histidines. We show that AtxA crystallizes as a dimer. Distinct folds associated with predicted DNA-binding domains (HTH1 and HTH2) and phosphoenolpyruvate: carbohydrate phosphotransferase system-regulated domains (PRD1 and PRD2) are apparent. We tested AtxA variants containing single and double phosphomimetic (His → Asp) and phosphoablative (His → Ala) amino acid changes for activity in B. anthracis cultures and for protein-protein interactions in cell lysates. Reduced activity of AtxA H199A, lack of multimerization and activity of AtxAH379D variants, and predicted structural changes associated with phosphorylation support a model for control of AtxA function. We propose that (1) in the AtxA dimer, phosphorylation of H199 in PRD1 affects HTH2 positioning, influencing DNA-binding; and (2) phosphorylation of H379 in PRD2 disrupts dimer formation. The AtxA structure is the first reported high-resolution full-length structure of a PRD-containing regulator and can serve as a model for proteins of this family, especially those that link virulence to bacterial metabolism.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4222560/

AhR and Arnt differentially regulate NF-κB signaling and chemokine responses in human bronchial epithelial cells

Background The aryl hydrocarbon receptor (AhR) has gradually emerged as a regulator of inflammation in the lung and other tissues. AhR may interact with the p65-subunit of the nuclear factor (NF)-κB transcription factors, but reported outcomes of AhR/NF-κB-interactions are conflicting. Some studies suggest that AhR possess pro-inflammatory activities while others suggest that AhR may be anti-inflammatory. The present study explored the impact of AhR and its binding partner AhR nuclear translocator (Arnt) on p65-activation and two differentially regulated chemokines, CXCL8 (IL-8) and CCL5 (RANTES), in human bronchial epithelial cells (BEAS-2B). Results Cells were exposed to CXCL8- and CCL5-inducing chemicals, 1-nitropyrene (1-NP) and 1-aminopyrene (1-AP) respectively, or the synthetic double-stranded RNA analogue, polyinosinic-polycytidylic acid (Poly I:C) which induced both chemokines. Only CXCL8, and not CCL5, appeared to be p65-dependent. Yet, constitutively active unligated AhR suppressed both CXCL8 and CCL5, as shown by siRNA knock-down and the AhR antagonist α-naphthoflavone. Moreover, AhR suppressed activation of p65 by TNF-α and Poly I:C as assessed by luciferase-assay and p65-phosphorylation at serine 536, without affecting basal p65-activity. In contrast, Arnt suppressed only CXCL8, but did not prevent the p65-activation directly. However, Arnt suppressed expression of the NF-κB-subunit RelB which is under transcriptional regulation by p65. Furthermore, AhR-ligands alone at high concentrations induced a moderate CXCL8-response, without affecting CCL5, but suppressed both CXCL8 and CCL5-responses by Poly I:C. Conclusion AhR and Arnt may differentially and independently regulate chemokine-responses induced by both inhaled pollutants and pulmonary infections. Constitutively active, unligated AhR suppressed the activation of p65, while Arnt may possibly interfere with the action of activated p65. Moreover, ligand-activated AhR suppressed CXCL8 and CCL5 responses by other agents, but AhR ligands alone induced CXCL8 responses when given at sufficiently high concentrations, thus underscoring the duality of AhR in regulation of inflammation. We propose that AhR-signaling may be a weak activator of p65-signaling that suppresses p65-activity induced by strong activators of NF-κB, but that its anti-inflammatory properties also are due to interference with additional pathways. Background The aryl hydrocarbon receptor (AhR) has gradually emerged as a regulator of inflammation in the lung and other tissues. AhR may interact with the p65-subunit of the nuclear factor (NF)-κB transcription factors, but reported outcomes of AhR/NF-κB-interactions are conflicting. Some studies suggest that AhR possess pro-inflammatory activities while others suggest that AhR may be anti-inflammatory. The present study explored the impact of AhR and its binding partner AhR nuclear translocator (Arnt) on p65-activation and two differentially regulated chemokines, CXCL8 (IL-8) and CCL5 (RANTES), in human bronchial epithelial cells (BEAS-2B). Results Cells were exposed to CXCL8- and CCL5-inducing chemicals, 1-nitropyrene (1-NP) and 1-aminopyrene (1-AP) respectively, or the synthetic double-stranded RNA analogue, polyinosinic-polycytidylic acid (Poly I:C) which induced both chemokines. Only CXCL8, and not CCL5, appeared to be p65-dependent. Yet, constitutively active unligated AhR suppressed both CXCL8 and CCL5, as shown by siRNA knock-down and the AhR antagonist α-naphthoflavone. Moreover, AhR suppressed activation of p65 by TNF-α and Poly I:C as assessed by luciferase-assay and p65-phosphorylation at serine 536, without affecting basal p65-activity. In contrast, Arnt suppressed only CXCL8, but did not prevent the p65-activation directly. However, Arnt suppressed expression of the NF-κB-subunit RelB which is under transcriptional regulation by p65. Furthermore, AhR-ligands alone at high concentrations induced a moderate CXCL8-response, without affecting CCL5, but suppressed both CXCL8 and CCL5-responses by Poly I:C. Conclusion AhR and Arnt may differentially and independently regulate chemokine-responses induced by both inhaled pollutants and pulmonary infections. Constitutively active, unligated AhR suppressed the activation of p65, while Arnt may possibly interfere with the action of activated p65. Moreover, ligand-activated AhR suppressed CXCL8 and CCL5 responses by other agents, but AhR ligands alone induced CXCL8 responses when given at sufficiently high concentrations, thus underscoring the duality of AhR in regulation of inflammation. We propose that AhR-signaling may be a weak activator of p65-signaling that suppresses p65-activity induced by strong activators of NF-κB, but that its anti-inflammatory properties also are due to interference with additional pathways. Introduction The aryl hydrocarbon receptor (AhR) is a ligand-activated transcription factor originally discovered as a receptor for 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) and other environmental pollutants, including polycyclic aromatic hydrocarbons (PAHs). In its classical mode of action, cytosolic AhR activated by hydrocarbons rapidly translocates to the nucleus where it dimerizes with the AhR nuclear translocator (Arnt). The AhR/Arnt complex binds to dioxin or xenobiotic response elements (DREs or XREs) in the promoter region of target genes, such as the cytochrome P450 enzymes CYP1A1 and -1B1 [ 1 ]. It has now become evident that the AhR is also involved in regulation of inflammation as well as a variety of other endogenous processes. Chronic low-grade inflammation is a common feature of a variety of multi-factorial diseases including various pulmonary disorders [ 2 ],[ 3 ]. It has been suggested that increases in such chronic disorders at least partly may be related to inhalation of environmental pollutants, contributing to development or exacerbation of inflammation [ 4 ]–[ 6 ]. That one of the main cellular sensors of hydrocarbons is now emerging as a regulator of innate immune responses may therefore be of particular importance. In addition, the role of AhR in inflammation does not seem to be restricted to effects of inhaled man-made chemicals. There seems to be a considerable cross-talk between AhR-signaling and the nuclear factor-κB (NF-κB) family of transcription factors [ 7 ],[ 8 ]. The classical NF-κB-pathway typically characterized by the p65/p50 dimer is central in regulation of inflammatory responses through binding to κB-sites in the promoter region of a variety of pro-inflammatory genes, including several cytokines and chemokines [ 7 ],[ 8 ]. In lung cells, TCDD exposure increased NF-κB activity and IL-6 expression [ 9 ]. TCDD also induced IL-1β, IL-6 and CXCL8 (IL-8) through AhR-mediated activation of NF-κB in rheumathoid arthritis [ 10 ]. Furthermore, Vogel and colleagues [ 11 ],[ 12 ] showed that TCDD induced dimerization of AhR and RelB of the alternative NF-κB pathway and up-regulation of CXCL8 through a novel RelB/AhR response element (RelBAHRE) in macrophages and breast cancer cells [ 12 ]. Unligated AhR has been found to dimerize with p65 and activate κB-sites in the IL-6 and c- myc promoters [ 9 ],[ 13 ], and overexpression of constitutively activated AhR was associated with inflammatory skin lesions in mice [ 14 ]. In addition, the archetypical PAH benzo[ a ]pyrene (B[ a ]P) appeared to induce CXCL8 expression through binding of AhR to consensus XRE sites in the CXCL8 promoter [ 15 ]. In stark contrast, AhR-deficient mice displayed increased NF-κB activity and inflammation in the lungs after inhalation of lipopolysaccharide (LPS), cigarette smoke as well as instillation of crystalline silica [ 16 ],[ 17 ]. AhR knock-down also seems to cause hypersensitivity towards systemic inflammation in mice exposed to LPS by intraperitoneal injection [ 18 ],[ 19 ], and AhR deprivation increases LPS-induced cytokine responses in various cell types [ 18 ],[ 20 ],[ 21 ]. In addition, AhR-agonists such as TCDD and β-naphthoflavone (BNF) suppressed NF-κB signaling and cytokine responses induced by TNF-α, LPS and crystalline silica [ 17 ],[ 21 ],[ 22 ]. Thus, AhR seems to elicit both pro- and anti-inflammatory functions and both enhance and suppress NF-κB activity in the lung and other tissues. We have previously found that 1-nitropyrene (1-NP), a prominent PAH in diesel exhaust, induced a cytokine/chemokine expression profile in human bronchial epithelial cells (BEAS-2B) characterized by high levels of CXCL8, whereas its amine metabolite 1-aminopyrene (1-AP) induced a different response characterized by increased CCL5 (RANTES) [ 23 ],[ 24 ]. CXCL8 is a potent neutrophil attractant associated with innate immune responses [ 25 ],[ 26 ], while CCL5 activates and attracts eosinophils commonly involved in allergic reactions [ 27 ]. CXCL8 may also contribute to tumor growth [ 28 ]–[ 30 ], while CCL5 may induce anti-tumor responses [ 31 ]–[ 33 ]. In addition to their differential biological properties, expression of CXCL8 and CCL5 also appear to be regulated differentially in airway epithelial cells [ 34 ],[ 35 ]. The effects of 1-NP on CXCL8 seems to involve ROS-formation and signaling through the TACE/TGF-α/EGFR-cascade [ 36 ]–[ 38 ], while the mechanisms of 1-AP induced CCL5 are less clear. However, the chemokine responses induced by both compounds seems at least partly to involve activation of β2-adrenergic receptors [ 39 ] (as well as unpublished results from our lab). Due to their differential effects, the 1-NP-induced CXCL8 versus 1-AP-induced CCL5 represented an intriguing model for studying the impact of AhR and Arnt on pro-inflammatory responses. In the present work we have explored the roles of AhR and Arnt in the regulation of CXCL8 and CCL5 with emphasis on the involvement of the classical and alternative NF-κB pathway in BEAS-2B cells. Cells were exposed to 1-NP and 1-AP as well as polyinosinic-polycytidylic acid (Poly I:C), a synthetic double-stranded RNA analogue and Toll-like receptor 3 (TLR3) agonist known to induce NF-κB signaling and both CXCL8 and CCL5 responses in BEAS-2B cells [ 38 ]. The results suggest that the constitutive (endogenous) activity of unligated AhR and Arnt differentially suppress CXCL8 and CCL5 responses, and we propose an explanation for how AhR may both induce and suppress NF-κB signaling, even within the same cell type. The results also suggest that anti-inflammatory effects of AhR necessarily extend beyond interactions with the NF-kB pathway. Results Role of p65 in CXCL8 and CCL5 regulation in 1-NP- or 1-AP-exposed BEAS-2B cells We have previously shown that 1-NP induces CXCL8, while 1-AP induces CCL5 in BEAS-2B cells [ 23 ],[ 24 ]. As a first step, we explored the involvement of the classical NF-κB pathways in the 1-NP- and 1-AP-induced chemokine responses by transfecting the cells with siRNA against p65 or non-targeting control siRNA. The classical NF-κB pathway is generally considered necessary for transcription of CXCL8 [ 40 ]. In line with this, we found that p65 silencing completely blocked both basal and induced CXCL8 responses (Figure 1 A and C). In contrast, p65 siRNA had little impact on, or rather increased, the 1-AP-induced CCL5 response (Figure 1 B and D). The protein level of p65 was markedly down-regulated in cells treated with the p65 siRNAs, confirming the efficiency of the transfection (Figure 1 E). The above results suggest that the classical NF-κB pathway is needed for the CXCL8 response, but not for CCL5 in our cell model. Figure 1 P65 is required for CXCL8 but not CCL5 responses in 1- NP- or 1- AP- exposed BEAS- 2B cells. Cells were transfected with siRNA against p65 (siP65) or non-targeting control siRNA (siNT), and exposed to 20 μM 1-NP, 1-AP or vehicle (DMSO) alone. CXCL8 (A) and CCL5 (B) gene expression were measured after 6 h by real-time PCR. CXCL8 (C) and CCL5 (D) protein levels in the medium were measured by ELISA after 18 h exposure as described under "Materials and methods". Efficiency of transfection was assessed by Western blotting (E) . The results are expressed as mean ± SEM (A/B: n = 1 (triplicate determinations); C/D: n ≥ 3; E: representative blot, n = 3). *Significantly different from unexposed controls; †Significantly different from cells transfected with non-targeting siRNA. Role of AhR and Arnt in CXCL8 and CCL5 regulation in 1-NP- or 1-AP-exposed BEAS-2B cells As a next step we explored the roles of AhR and Arnt in the regulation of 1-NP-induced CXCL8 and 1-AP-induced CCL5. Thus, BEAS-2B cells were transfected with siRNA against AhR, Arnt or non-targeting control siRNA prior to exposure. Silencing of both AhR and Arnt increased 1-NP-induced CXCL8 mRNA expression after 6 h exposure and protein release after 18 h exposure (Figure 2 A and C), with a markedly higher response in cells transfected with Arnt siRNA. However, only silencing of AhR induced a strong increase in both basal and induced CCL5 (Figure 2 B and D). In comparison, targeting Arnt resulted in a more moderate increase in CCL5 that seemed restricted to basal and not induced levels, as 1-AP failed to increase CCL5 release in Arnt-depleted cells (Figure 2 D). Notably, AhR silencing increased basal levels of CCL5 but not CXCL8, indicating that AhR regulated the two chemokines differentially. In comparison Arnt silencing seemed to induce an increase in basal levels of both CXCL8 and CCL5. This was confirmed by a more detailed examination of the effects on basal chemokine levels (Additional file 1 : Figure S1, online supplementary materials). The efficiency of the siRNA transfection was verified by Western blotting, showing depletion of AhR and Arnt by their respective siRNAs (Figure 2 E). In addition, AhR and Arnt silencing also reduced basal CYP1A1 expression and blocked B[ a ]P-dependent induction of CYP1A1 (Figure 2 F), which further confirmed the efficiency of the siRNA transfections. Figure 2 AhR and Arnt differentially regulate CXCL8 and CCL5 responses in 1- NP- or 1- AP- exposed BEAS- 2B cells. Cells were transfected with siRNA against AhR (siAHR) and Arnt (siARNT) or non-targeting control siRNA (siNT), and exposed to 20 μM 1-NP, 1-AP or vehicle (DMSO) alone. CXCL8 (A) and CCL5 (B) gene expression were measured after 6 h by real-time PCR. CXCL8 (C) and CCL5 (D) protein levels in the medium were measured by ELISA after 18 h exposure as described under "Materials and methods". Efficiency of transfection was assessed by Western blotting (E) as well as expression of CYP1A1 after 6 h exposure to 20 μM B[a]P, 1-NP and 1-AP, by real-time PCR (F) . The results are expressed as mean ± SEM (A/B/F: n = 1 (triplicate determinations); C/D: n ≥ 3; E: representative blots, n ≥ 3). *Significantly different from unexposed controls; †Significantly different from cells transfected with non-targeting siRNA. Having shown that AhR and Arnt depletion enhanced CXCL8 and CCL5 responses, we next explored whether blocking the PAH-binding site of AhR would result in similar effects. Indeed, both 1-NP and 1-AP seem to have the ability to bind and activate AhR in other cells types [ 41 ],[ 42 ], although they did not induce CYP1A1 in the BEAS-2B cells (Figure 2 F). Thus, to investigate whether the suppressive effects of AhR and Arnt on CXCL8 and CCL5 were related to ligand-mediated activation, BEAS-2B cells were treated with the AhR antagonist α-naphthoflavone (ANF), prior to PAH exposure. Pretreatment with 0.5 μM of ANF reduced B[ a ]P-induced CYP1A1 induction by more than 50%, confirming the efficiency of the antagonist (Figure 3 A). However, in contrast to the enhanced chemokine responses observed by siRNA transfection (Figure 2 ), blocking the ligand-binding site of AhR with ANF did not affect CXCL8 levels (Figure 3 B), and rather decreased 1-AP-induced CCL5 (Figure 3 C). Figure 3 Blocking of AhR ligand binding does not increase CXCL8 and CCL5 responses in BEAS- 2B cells. Cells were incubated with 0.5 μM of the AhR antagonist α-naphtoflavone (ANF) for 30 min prior to exposure for 18 h to 20 μM B[ a ]P, 1-NP, 1-AP or vehicle (DMSO) alone. CYP1A1 (A) levels were measured by real-time PCR, while CXCL8 (B) and CCL5 (C) levels in the medium were measured by ELISA as described under "Materials and methods". The results are expressed as mean ± SEM (A: n = 1 (triplicate determinations); B/C: n = 5). *Significantly different from unexposed controls; †Significant effect of ANF. The above results suggest that AhR acts as a suppressor of both 1-NP-induced CXCL8 and 1-AP-induced CCL5, while Arnt more selectively suppresses 1-NP-induced CXCL8 responses in BEAS-2B cells. These effects seemed to be independent of ligand-mediated activation of AhR, and may therefore rather be due to basal or constitutive receptor activity. Moreover, since CCL5 seemed to be regulated independently of p65 (Figure 1 ), suppression of CCL5 by AhR was most likely mediated through NF-κB-independent mechanisms. Induction of NF-κB activity by 1-NP and TNF-α, and the roles of AhR and Arnt Since basal and 1-NP-induced CXCL8 seemed dependent on p65 expression, we next investigated the effect of 1-NP on activation of NF-κB using a luciferase assay for p65 activity. In an initial screen, the pro-inflammatory cytokine TNF-α used as positive control appeared to induce a strong increase in NF-κB-driven luciferase activity after both 6 and 16 h exposure (Figure 4 A). In comparison, 1-NP did not affect luciferase activity at the early time point and only induced a moderate increase after 16 h (Figure 4 A). However, this effect was not statistically significant in the next set of experiments (Figure 4 B). Moreover, no effect on IκB-degradation or phosphorylation of p-p65 were observed after 2 or 4 h exposure to 1-NP or 1-AP (Additional file 2 : Figure S2, online supplementary material). Considering that 1-NP induced a marked increase in CXCL8 mRNA expression already after 4 h exposure (Figures 1 A and 2 A), it does not seem that 1-NP mediated the CXCL8 response through activation of the classical NF-κB pathway. Thus, p65 most likely played a strictly permissive role in 1-NP-induced CXCL8 (Figure 1 ), and that the observed effect of p65 silencing was due to attenuation of the basal transcription factor activity. In line with this, p65 depletion also reduced CXCL8 levels in controls and 1-AP-exposed cells (Figure 1 ). Figure 4 AhR silencing increased TNF- α-, but not 1- NP- induced p50/ p65 reporter gene expression in BEAS- 2B cells. Cells were transfected with a NF-κB-luciferase (p50/p65) reporter gene, and exposed to 1-NP or TNF-α for 6 or 16 h (A) , and assayed for luciferase activity as described under "Materials and methods". Cells transfected siRNAs against AhR (siAHR), Arnt (siARNT) or non-targeting control siRNA (siNT) were further transfected with NF-κB-luciferase promoter and assayed for luciferase activity after a 16 h exposure to 1-NP (20 μM) or TNF-α (50 ng/ml) (B) . The results are expressed as mean ± SEM (A: n = 2; B: n = 5). *Significantly different from unexposed controls; †Significantly different from cells transfected with non-targeting siRNA. To assess any possible influence of AhR and Arnt on p65 activity, we measured p65-induced luciferase activity in AhR- or Arnt-depleted BEAS-2B cells after 16 h exposure to 1-NP or TNF-α. AhR silencing by siRNA did not affect basal or 1-NP-induced p65 activity significantly (Figure 4 B). If anything, the low-level induction of p65 activity by 1-NP was rather reduced. Thus, it seems that not only 1-AP-induced CCL5, but also the 1-NP-induced CXCL8 response was suppressed by AhR through mechanisms other than interference with the classical NF-κB pathway. However, AhR silencing resulted in a moderate, but statistically significant increase in TNF-α-induced p65 activity (Figure 4 B), suggesting that the constitutive activity of AhR modestly suppresses the p65 signaling induced by stronger activators of the classical NF-κB pathway. Cells transfected with Arnt siRNA did not display any changes in luciferase activity as compared to non-targeting control siRNA (Figure 4 B). Role of AhR and Arnt in p65 and CXCL8/CCL5 regulation in Poly I:C-exposed BEAS-2B cells To examine if the chemokine regulation by AhR and Arnt were of a more general nature, we continued by exploring their effects in cells exposed to the Toll-like receptor 3 (TLR3)-ligand Poly I:C. In agreement with previous reports [ 43 ],[ 44 ], Poly I:C induced strong increases in both CXCL8 and CCL5 along with activation of the classical NF-κB pathway, as evidenced by the degradation of IκB and increased phosphorylation of p65 at serine 536 (Additional file 2 : Figure S2, online supplementary materials). This allowed for assessing the effects of AhR and Arnt on CXCL8 and CCL5 responses induced by a single compound without known affinity for the AhR or CYP-enzymes, that unlike 1-NP and 1-AP, also activated the classical NF-κB pathway. We thus examined the effects of targeting p65, AhR and Arnt by siRNA in Poly I:C-exposed cells. In line with the observed increase in TNF-α-induced p65 activity by AhR-silencing (Figure 4 B), we found that depletion of AhR resulted in increased Poly I:C-induced phosphorylation of p65 at serine 536 (Ser536) in BEAS-2B cells at 2 h, while Arnt depletion did not cause any significant effects (Figure 5 A and B). This strengthens the notion that AhR suppresses activation of p65-signaling induced by strong inducers of the classical NF-κB pathway. As in PAH-exposed cells (Figure 1 ), we again observed that CXCL8, but not CCL5, was dependent on p65 (Figure 5 C and D). In further agreement with the data obtained by 1-NP and 1-AP exposure (Figure 2 ), silencing of both AhR and Arnt increased CXCL8 responses in Poly I:C-exposed BEAS-2B cells (Figure 5 E), while only AhR silencing augmented CCL5 (Figure 5 F). The effects obtained by the custom made siRNA against AhR were confirmed by use of commercially available siRNA (Additional file 3 : Figure S3, online supplementary materials). This fortifies the observations made by PAH treatment and suggests that AhR and Arnt may play a general role in the regulation of CXCL8 and CCL5 in BEAS-2B cells, independently of exposure. The results further suggest that AhR may interfere with phosphorylation of p65 at Ser536 which is considered important for transactivation of p65 and transcription of CXCL8 [ 45 ],[ 46 ]. Moreover, as CCL5 was not affected by p65 silencing, AhR necessarily also regulated Poly I:C-induced CCL5 expression through interactions with other pathways. Figure 5 AhR and Arnt differentially regulate CXCL8 and CCL5 responses as well as p65 phosphorylation in Poly I: C- exposed BEAS- 2B cells. Cells were transfected with siRNA against p65 (siP65), AhR (siAHR), Arnt (siARNT) or non-targeting control siRNA (siNT), and exposed to 10 μg/ml Poly I:C. Intracellular protein levels of total and phospho-p65 (Ser536) as well as AhR and Arnt were detected by Western blotting after 2 and 4 h exposure, as described under "Materials and methods". The figure displays representative blots (A) , as well as relative changes in phospho-p65 compared to total p65 quantified by densitometric analysis of the Western blots (B) . CXCL8 (C and E) and CCL5 (D and F) protein levels in the medium were measured by ELISA after 18 h exposure, as described under "Materials and methods". The results are expressed as mean ± SEM (n ≥ 3). *Significantly different from unexposed controls; †Significantly different from cells transfected with non-targeting siRNA. Interaction with CXCL8 and CCL5 promoters and subcellular localization of p65, AhR and Arnt in Poly I:C-exposed BEAS-2B cells For a more complete understanding of the roles of p65, AhR and Arnt in CXCL8 and CCL5 regulation, we also assessed to what extent they interacted directly with the NF-κB-response element containing regions of the CXCL8 and CCL5 promoters when cells were exposed to ligands. In line with the above results, ChIP assay revealed that Poly I:C treatment induced a strong increase in p65 binding to the CXCL8 promoter, but not to the CCL5 promoter (Figure 6 ). Moreover, we were unable to detect binding of AhR and Arnt to the CXCL8 and CCL5 promoters by ChIP (data not shown). Thus, neither AhR nor Arnt appeared to suppress chemokine responses by interfering directly the NF-κB responsive regions of the CXCL8 and CCL5 promoters. Furthermore, we also assessed the subcellular localization of p65, AhR and Arnt by immunocytochemistry. In control cells, p65 was primarily located in the cytosol, and translocated into the nucleus upon exposure to Poly I:C and partly also when exposed to 1-NP (Additional file 4 : Figure S4, online supplementary materials). AhR and Arnt on the other hand, appeared to be located both in the cytosol and in nucleus of resting cells, and no apparent change were observed in the sub-cellular localization upon exposure (Additional file 4 : Figure S4, online supplementary materials). This is in line with the observation that AhR signaling was not activated by the exposure. Figure 6 Interaction between p65 and the CXCL8 and CCL5 promoters in Poly I: C exposed BEAS -2B cells. Cells were exposed to 10 μg/ml Poly I:C for 3 h. Interaction between p65 and the NF-κB responsive regions of the CXCL8 and CCL5 promoters were assessed by ChIP assay as described under Materials and methods". Results are expressed as fold increase compared to p65 binding to the CXCL8 promoter after adjustment for the input control (GAPDH). The results are expressed as mean ± SEM ( n = 6). The insert figure shows the same data set on a log-scale. *Significantly different from unexposed controls. Role of RelB in AhR and Arnt-mediated regulation of CXCL8 and CCL5 RelB of the alternative (non-canonical) NF-κB pathway is a well-known suppressor of the classical NF-κB pathway and inflammation that itself is under transcriptional regulation by p65 [ 47 ]–[ 49 ]. Therefore, if AhR or Arnt silencing enhances p65-signaling, RelB should be increased in a similar manner as CXCL8. Conversely, it has been proposed that the anti-inflammatory role of AhR is related to stabilization of RelB, resulting in a more rapid degradation of RelB in AhR-depleted cells or tissues exposed to cigarette smoke [ 16 ],[ 20 ]. We first examined the impact of AhR and Arnt silencing on RelB expression. The results show that basal RelB levels were markedly increased in cells transfected with Arnt, but not AhR siRNA (Figure 7 A). RelB levels were also increased in Arnt-depleted cells exposed to Poly I:C for 2 and 4 h (Figure 7 B). In contrast, AhR silencing first increased RelB levels after 4 h exposure. Therefore, as observed with CXCL8, Arnt seemed to suppress both basal and induced RelB while AhR more selectively suppressed only induced RelB levels. The latter is consistent with the observation that AhR only suppressed induced and not basal p65 activity. However, the Arnt-mediated suppression of RelB suggests that Arnt may interfere with p65-signaling despite its lack of effect on p65 activation (Figures 4 and 5 ). Furthermore, the results suggest that neither AhR nor Arnt suppressed chemokine responses by stabilizing RelB in Poly I:C-exposed BEAS-2B cells. Figure 7 Depletion of AhR and Arnt increases RelB levels in Poly I: C - exposed BEAS- 2B cells. Cells were transfected with siRNA against AhR (siAHR), Arnt (siARNT) or non-targeting control siRNA (siNT), and exposed to 10 μg/ml Poly I:C for 2 and 4 h. Intracellular protein levels of RelB and β-actin were detected by Western blotting as described under "Materials and methods". The figure displays representative blots of RelB levels in unexposed cells (A) and Poly I:C-exposed cells (B) . The graph depicts relative changes in RelB (B) compared to β-actin quantified by densitometric analysis of the Western blots. The results are expressed as mean ± SEM (n = 3). *Significantly different from unexposed controls; †Significantly different from cells transfected with non-targeting siRNA. Notably, RelB has also been implicated in AhR- or Arnt-mediated transcription regulation in various ways [ 11 ],[ 12 ],[ 16 ],[ 20 ],[ 49 ]. So, to further explore whether RelB could be involved in AhR- or Arnt-mediated chemokine suppression, we then examined the effect of RelB silencing on chemokine responses in 1-NP-, 1-AP- or Poly I:C-exposed BEAS-2B cells. RelB depletion by siRNA increased CXCL8 responses (Figure 8 A,C and E) but reduced CCL5 responses (Figure 8 B,D and F). Thus, the effects of RelB partly resembled that of Arnt (Figures 2 and 5 ), by suppressing both basal and induced CXCL8 and not CCL5. However, in contrast to Arnt, RelB seemed to be required for CCL5 responses. The observation that AhR in addition to suppressing CCL5 selectively suppressed induced, but not basal CXCL8 levels (Figures 2 and 5 ), further suggests that AhR did not mediate its anti-inflammatory effects through RelB. Figure 8 RelB suppresses CXCL8, but not CCL5, in BEAS- 2B cells. Cells were transfected with siRNA against RelB (siRELB) or non-targeting control siRNA (siNT). Transfected cells were exposed to 20 μM 1-NP, 1-AP, or vehicle (DMSO) alone (A-D) or 10 μg/ml Poly I:C or vehicle (water) alone (E and F) . CXCL8 (A) and CCL5 (B) gene expression were measured after 6 h by real-time PCR. CXCL8 (C and E) and CCL5 (D and F) protein levels in the medium were measured by ELISA after 18 h exposure as described under "Materials and methods". Efficiency of transfection was assessed by Western blotting (G) . The results are expressed as mean ± SEM (A/B: n = 1 (triplicate determinations); C-F : n ≥ 3; G: representative blot, n = 3) *Significantly different from unexposed controls; †Significantly different from cells transfected with non-targeting siRNA. In contrast to the effects obtained by RelB siRNA (Figure 8 ), silencing its binding-partner in the alternative NF-κB pathways, p100/p52, attenuated both CXCL8 and CCL5 responses (Additional file 5 : Figure S5A, online supplementary materials). Moreover, neither 1-NP, 1-AP or Poly I:C induced degradation of p100 to p52 (Additional file 5 : Figure S5B, online supplementary materials), the hallmark of alternative NF-κB activation [ 50 ]. This suggests that the RelB-mediated suppression of CXCL8 (Figure 8 ) was independent of the alternative NF-κB pathway. However, it is possible that the RelB-dependency of CCL5 (Figure 8 ) may be related to some permissive role of the alternative NF-κB pathway. Activation of AhR by BNF suppresses Poly I:C-induced CXCL8 and CCL5 Our present results suggest that the constitutive activity of unligated AhR and Arnt regulate chemokine responses through differential mechanisms. As a final step, we also explored how ligand-mediated activation of AhR-signaling would affect CXCL8 and CCL5 regulation. It has been reported that the non-toxic AhR agonist β-naphthoflavone (BNF) may suppress TNF-α-induced NF-κB activation and LPS-induced cytokine responses [ 21 ],[ 22 ]. Thus we examined the effect of BNF on Poly I:C-induced CXCL8 and CCL5 in BEAS-2B cells. Pre-incubation with 1 μM BNF for 30 min prior to Poly I:C exposure for 18 h, attenuated both Poly I:C-induced CXCL8 and CCL5 (Figure 9 A and B). This effect seem to extend to other cell types, as BNF suppressed Poly I:C-induced chemokine responses in A549 human alveolar type-II-like epithelial carcinoma cells and THP-1 human leukemia monocytes (Additional file 6 : Figure S6, online supplementary materials). Of notice, Poly I:C appeared unable to induce CXCL8-responses in the A549 cells. We also observed that pre-incubation with another well-known AhR-agonist, B[ a ]P, suppressed both CXCL8 and CCL5 responses in BEAS-2B cells exposed to Poly I:C or LPS (Additional file 7 : Figure S7, online supplementary materials). In contrast, treatment with high concentrations of BNF and ANF (10 and 25 μM, respectively) alone, induced moderate two-fold increases in CXCL8 but not CCL5 (Figure 9 C and D). Therefore it seems that not only the constitutive activity of unligated AhR, but also ligand-mediated activation of the AhR may suppress pro-inflammatory responses in BEAS-2B cells. These results also show that activation of AhR by external ligands increase the suppression of cytokine/chemokine responses compared to the effects of unligated, constitutively active AhR. However, in absence of other, stronger pro-inflammatory stimuli, AhR-ligands elicited a moderat pro-inflammatory effect in BEAS-2B cells when given at sufficiently high concentrations. Figure 9 AhR- activation suppresses Poly I: C- induced CXCL8 and CCL5 responses , but stimulate CXCL8 responses alone in BEAS- 2B cells. Cells were incubated with 1 μM of the AhR agonist β-naphtoflavone (BNF) for 30 min prior to exposure to 10 μg/ml Poly I:C for 18 h (A and B) , or exposed to high concentrations of ANF or BNF alone for 18 h (C and D) . CXCL8 (A and C) and CCL5 (B and D) levels in the medium were measured by ELISA as described under "Materials and methods". The results are expressed as mean ± SEM (n = 3). *Significant increase induced by Poly I:C; †Significant reduction induced by BNF. Role of p65 in CXCL8 and CCL5 regulation in 1-NP- or 1-AP-exposed BEAS-2B cells We have previously shown that 1-NP induces CXCL8, while 1-AP induces CCL5 in BEAS-2B cells [ 23 ],[ 24 ]. As a first step, we explored the involvement of the classical NF-κB pathways in the 1-NP- and 1-AP-induced chemokine responses by transfecting the cells with siRNA against p65 or non-targeting control siRNA. The classical NF-κB pathway is generally considered necessary for transcription of CXCL8 [ 40 ]. In line with this, we found that p65 silencing completely blocked both basal and induced CXCL8 responses (Figure 1 A and C). In contrast, p65 siRNA had little impact on, or rather increased, the 1-AP-induced CCL5 response (Figure 1 B and D). The protein level of p65 was markedly down-regulated in cells treated with the p65 siRNAs, confirming the efficiency of the transfection (Figure 1 E). The above results suggest that the classical NF-κB pathway is needed for the CXCL8 response, but not for CCL5 in our cell model. Figure 1 P65 is required for CXCL8 but not CCL5 responses in 1- NP- or 1- AP- exposed BEAS- 2B cells. Cells were transfected with siRNA against p65 (siP65) or non-targeting control siRNA (siNT), and exposed to 20 μM 1-NP, 1-AP or vehicle (DMSO) alone. CXCL8 (A) and CCL5 (B) gene expression were measured after 6 h by real-time PCR. CXCL8 (C) and CCL5 (D) protein levels in the medium were measured by ELISA after 18 h exposure as described under "Materials and methods". Efficiency of transfection was assessed by Western blotting (E) . The results are expressed as mean ± SEM (A/B: n = 1 (triplicate determinations); C/D: n ≥ 3; E: representative blot, n = 3). *Significantly different from unexposed controls; †Significantly different from cells transfected with non-targeting siRNA. Role of AhR and Arnt in CXCL8 and CCL5 regulation in 1-NP- or 1-AP-exposed BEAS-2B cells As a next step we explored the roles of AhR and Arnt in the regulation of 1-NP-induced CXCL8 and 1-AP-induced CCL5. Thus, BEAS-2B cells were transfected with siRNA against AhR, Arnt or non-targeting control siRNA prior to exposure. Silencing of both AhR and Arnt increased 1-NP-induced CXCL8 mRNA expression after 6 h exposure and protein release after 18 h exposure (Figure 2 A and C), with a markedly higher response in cells transfected with Arnt siRNA. However, only silencing of AhR induced a strong increase in both basal and induced CCL5 (Figure 2 B and D). In comparison, targeting Arnt resulted in a more moderate increase in CCL5 that seemed restricted to basal and not induced levels, as 1-AP failed to increase CCL5 release in Arnt-depleted cells (Figure 2 D). Notably, AhR silencing increased basal levels of CCL5 but not CXCL8, indicating that AhR regulated the two chemokines differentially. In comparison Arnt silencing seemed to induce an increase in basal levels of both CXCL8 and CCL5. This was confirmed by a more detailed examination of the effects on basal chemokine levels (Additional file 1 : Figure S1, online supplementary materials). The efficiency of the siRNA transfection was verified by Western blotting, showing depletion of AhR and Arnt by their respective siRNAs (Figure 2 E). In addition, AhR and Arnt silencing also reduced basal CYP1A1 expression and blocked B[ a ]P-dependent induction of CYP1A1 (Figure 2 F), which further confirmed the efficiency of the siRNA transfections. Figure 2 AhR and Arnt differentially regulate CXCL8 and CCL5 responses in 1- NP- or 1- AP- exposed BEAS- 2B cells. Cells were transfected with siRNA against AhR (siAHR) and Arnt (siARNT) or non-targeting control siRNA (siNT), and exposed to 20 μM 1-NP, 1-AP or vehicle (DMSO) alone. CXCL8 (A) and CCL5 (B) gene expression were measured after 6 h by real-time PCR. CXCL8 (C) and CCL5 (D) protein levels in the medium were measured by ELISA after 18 h exposure as described under "Materials and methods". Efficiency of transfection was assessed by Western blotting (E) as well as expression of CYP1A1 after 6 h exposure to 20 μM B[a]P, 1-NP and 1-AP, by real-time PCR (F) . The results are expressed as mean ± SEM (A/B/F: n = 1 (triplicate determinations); C/D: n ≥ 3; E: representative blots, n ≥ 3). *Significantly different from unexposed controls; †Significantly different from cells transfected with non-targeting siRNA. Having shown that AhR and Arnt depletion enhanced CXCL8 and CCL5 responses, we next explored whether blocking the PAH-binding site of AhR would result in similar effects. Indeed, both 1-NP and 1-AP seem to have the ability to bind and activate AhR in other cells types [ 41 ],[ 42 ], although they did not induce CYP1A1 in the BEAS-2B cells (Figure 2 F). Thus, to investigate whether the suppressive effects of AhR and Arnt on CXCL8 and CCL5 were related to ligand-mediated activation, BEAS-2B cells were treated with the AhR antagonist α-naphthoflavone (ANF), prior to PAH exposure. Pretreatment with 0.5 μM of ANF reduced B[ a ]P-induced CYP1A1 induction by more than 50%, confirming the efficiency of the antagonist (Figure 3 A). However, in contrast to the enhanced chemokine responses observed by siRNA transfection (Figure 2 ), blocking the ligand-binding site of AhR with ANF did not affect CXCL8 levels (Figure 3 B), and rather decreased 1-AP-induced CCL5 (Figure 3 C). Figure 3 Blocking of AhR ligand binding does not increase CXCL8 and CCL5 responses in BEAS- 2B cells. Cells were incubated with 0.5 μM of the AhR antagonist α-naphtoflavone (ANF) for 30 min prior to exposure for 18 h to 20 μM B[ a ]P, 1-NP, 1-AP or vehicle (DMSO) alone. CYP1A1 (A) levels were measured by real-time PCR, while CXCL8 (B) and CCL5 (C) levels in the medium were measured by ELISA as described under "Materials and methods". The results are expressed as mean ± SEM (A: n = 1 (triplicate determinations); B/C: n = 5). *Significantly different from unexposed controls; †Significant effect of ANF. The above results suggest that AhR acts as a suppressor of both 1-NP-induced CXCL8 and 1-AP-induced CCL5, while Arnt more selectively suppresses 1-NP-induced CXCL8 responses in BEAS-2B cells. These effects seemed to be independent of ligand-mediated activation of AhR, and may therefore rather be due to basal or constitutive receptor activity. Moreover, since CCL5 seemed to be regulated independently of p65 (Figure 1 ), suppression of CCL5 by AhR was most likely mediated through NF-κB-independent mechanisms. Induction of NF-κB activity by 1-NP and TNF-α, and the roles of AhR and Arnt Since basal and 1-NP-induced CXCL8 seemed dependent on p65 expression, we next investigated the effect of 1-NP on activation of NF-κB using a luciferase assay for p65 activity. In an initial screen, the pro-inflammatory cytokine TNF-α used as positive control appeared to induce a strong increase in NF-κB-driven luciferase activity after both 6 and 16 h exposure (Figure 4 A). In comparison, 1-NP did not affect luciferase activity at the early time point and only induced a moderate increase after 16 h (Figure 4 A). However, this effect was not statistically significant in the next set of experiments (Figure 4 B). Moreover, no effect on IκB-degradation or phosphorylation of p-p65 were observed after 2 or 4 h exposure to 1-NP or 1-AP (Additional file 2 : Figure S2, online supplementary material). Considering that 1-NP induced a marked increase in CXCL8 mRNA expression already after 4 h exposure (Figures 1 A and 2 A), it does not seem that 1-NP mediated the CXCL8 response through activation of the classical NF-κB pathway. Thus, p65 most likely played a strictly permissive role in 1-NP-induced CXCL8 (Figure 1 ), and that the observed effect of p65 silencing was due to attenuation of the basal transcription factor activity. In line with this, p65 depletion also reduced CXCL8 levels in controls and 1-AP-exposed cells (Figure 1 ). Figure 4 AhR silencing increased TNF- α-, but not 1- NP- induced p50/ p65 reporter gene expression in BEAS- 2B cells. Cells were transfected with a NF-κB-luciferase (p50/p65) reporter gene, and exposed to 1-NP or TNF-α for 6 or 16 h (A) , and assayed for luciferase activity as described under "Materials and methods". Cells transfected siRNAs against AhR (siAHR), Arnt (siARNT) or non-targeting control siRNA (siNT) were further transfected with NF-κB-luciferase promoter and assayed for luciferase activity after a 16 h exposure to 1-NP (20 μM) or TNF-α (50 ng/ml) (B) . The results are expressed as mean ± SEM (A: n = 2; B: n = 5). *Significantly different from unexposed controls; †Significantly different from cells transfected with non-targeting siRNA. To assess any possible influence of AhR and Arnt on p65 activity, we measured p65-induced luciferase activity in AhR- or Arnt-depleted BEAS-2B cells after 16 h exposure to 1-NP or TNF-α. AhR silencing by siRNA did not affect basal or 1-NP-induced p65 activity significantly (Figure 4 B). If anything, the low-level induction of p65 activity by 1-NP was rather reduced. Thus, it seems that not only 1-AP-induced CCL5, but also the 1-NP-induced CXCL8 response was suppressed by AhR through mechanisms other than interference with the classical NF-κB pathway. However, AhR silencing resulted in a moderate, but statistically significant increase in TNF-α-induced p65 activity (Figure 4 B), suggesting that the constitutive activity of AhR modestly suppresses the p65 signaling induced by stronger activators of the classical NF-κB pathway. Cells transfected with Arnt siRNA did not display any changes in luciferase activity as compared to non-targeting control siRNA (Figure 4 B). Role of AhR and Arnt in p65 and CXCL8/CCL5 regulation in Poly I:C-exposed BEAS-2B cells To examine if the chemokine regulation by AhR and Arnt were of a more general nature, we continued by exploring their effects in cells exposed to the Toll-like receptor 3 (TLR3)-ligand Poly I:C. In agreement with previous reports [ 43 ],[ 44 ], Poly I:C induced strong increases in both CXCL8 and CCL5 along with activation of the classical NF-κB pathway, as evidenced by the degradation of IκB and increased phosphorylation of p65 at serine 536 (Additional file 2 : Figure S2, online supplementary materials). This allowed for assessing the effects of AhR and Arnt on CXCL8 and CCL5 responses induced by a single compound without known affinity for the AhR or CYP-enzymes, that unlike 1-NP and 1-AP, also activated the classical NF-κB pathway. We thus examined the effects of targeting p65, AhR and Arnt by siRNA in Poly I:C-exposed cells. In line with the observed increase in TNF-α-induced p65 activity by AhR-silencing (Figure 4 B), we found that depletion of AhR resulted in increased Poly I:C-induced phosphorylation of p65 at serine 536 (Ser536) in BEAS-2B cells at 2 h, while Arnt depletion did not cause any significant effects (Figure 5 A and B). This strengthens the notion that AhR suppresses activation of p65-signaling induced by strong inducers of the classical NF-κB pathway. As in PAH-exposed cells (Figure 1 ), we again observed that CXCL8, but not CCL5, was dependent on p65 (Figure 5 C and D). In further agreement with the data obtained by 1-NP and 1-AP exposure (Figure 2 ), silencing of both AhR and Arnt increased CXCL8 responses in Poly I:C-exposed BEAS-2B cells (Figure 5 E), while only AhR silencing augmented CCL5 (Figure 5 F). The effects obtained by the custom made siRNA against AhR were confirmed by use of commercially available siRNA (Additional file 3 : Figure S3, online supplementary materials). This fortifies the observations made by PAH treatment and suggests that AhR and Arnt may play a general role in the regulation of CXCL8 and CCL5 in BEAS-2B cells, independently of exposure. The results further suggest that AhR may interfere with phosphorylation of p65 at Ser536 which is considered important for transactivation of p65 and transcription of CXCL8 [ 45 ],[ 46 ]. Moreover, as CCL5 was not affected by p65 silencing, AhR necessarily also regulated Poly I:C-induced CCL5 expression through interactions with other pathways. Figure 5 AhR and Arnt differentially regulate CXCL8 and CCL5 responses as well as p65 phosphorylation in Poly I: C- exposed BEAS- 2B cells. Cells were transfected with siRNA against p65 (siP65), AhR (siAHR), Arnt (siARNT) or non-targeting control siRNA (siNT), and exposed to 10 μg/ml Poly I:C. Intracellular protein levels of total and phospho-p65 (Ser536) as well as AhR and Arnt were detected by Western blotting after 2 and 4 h exposure, as described under "Materials and methods". The figure displays representative blots (A) , as well as relative changes in phospho-p65 compared to total p65 quantified by densitometric analysis of the Western blots (B) . CXCL8 (C and E) and CCL5 (D and F) protein levels in the medium were measured by ELISA after 18 h exposure, as described under "Materials and methods". The results are expressed as mean ± SEM (n ≥ 3). *Significantly different from unexposed controls; †Significantly different from cells transfected with non-targeting siRNA. Interaction with CXCL8 and CCL5 promoters and subcellular localization of p65, AhR and Arnt in Poly I:C-exposed BEAS-2B cells For a more complete understanding of the roles of p65, AhR and Arnt in CXCL8 and CCL5 regulation, we also assessed to what extent they interacted directly with the NF-κB-response element containing regions of the CXCL8 and CCL5 promoters when cells were exposed to ligands. In line with the above results, ChIP assay revealed that Poly I:C treatment induced a strong increase in p65 binding to the CXCL8 promoter, but not to the CCL5 promoter (Figure 6 ). Moreover, we were unable to detect binding of AhR and Arnt to the CXCL8 and CCL5 promoters by ChIP (data not shown). Thus, neither AhR nor Arnt appeared to suppress chemokine responses by interfering directly the NF-κB responsive regions of the CXCL8 and CCL5 promoters. Furthermore, we also assessed the subcellular localization of p65, AhR and Arnt by immunocytochemistry. In control cells, p65 was primarily located in the cytosol, and translocated into the nucleus upon exposure to Poly I:C and partly also when exposed to 1-NP (Additional file 4 : Figure S4, online supplementary materials). AhR and Arnt on the other hand, appeared to be located both in the cytosol and in nucleus of resting cells, and no apparent change were observed in the sub-cellular localization upon exposure (Additional file 4 : Figure S4, online supplementary materials). This is in line with the observation that AhR signaling was not activated by the exposure. Figure 6 Interaction between p65 and the CXCL8 and CCL5 promoters in Poly I: C exposed BEAS -2B cells. Cells were exposed to 10 μg/ml Poly I:C for 3 h. Interaction between p65 and the NF-κB responsive regions of the CXCL8 and CCL5 promoters were assessed by ChIP assay as described under Materials and methods". Results are expressed as fold increase compared to p65 binding to the CXCL8 promoter after adjustment for the input control (GAPDH). The results are expressed as mean ± SEM ( n = 6). The insert figure shows the same data set on a log-scale. *Significantly different from unexposed controls. Role of RelB in AhR and Arnt-mediated regulation of CXCL8 and CCL5 RelB of the alternative (non-canonical) NF-κB pathway is a well-known suppressor of the classical NF-κB pathway and inflammation that itself is under transcriptional regulation by p65 [ 47 ]–[ 49 ]. Therefore, if AhR or Arnt silencing enhances p65-signaling, RelB should be increased in a similar manner as CXCL8. Conversely, it has been proposed that the anti-inflammatory role of AhR is related to stabilization of RelB, resulting in a more rapid degradation of RelB in AhR-depleted cells or tissues exposed to cigarette smoke [ 16 ],[ 20 ]. We first examined the impact of AhR and Arnt silencing on RelB expression. The results show that basal RelB levels were markedly increased in cells transfected with Arnt, but not AhR siRNA (Figure 7 A). RelB levels were also increased in Arnt-depleted cells exposed to Poly I:C for 2 and 4 h (Figure 7 B). In contrast, AhR silencing first increased RelB levels after 4 h exposure. Therefore, as observed with CXCL8, Arnt seemed to suppress both basal and induced RelB while AhR more selectively suppressed only induced RelB levels. The latter is consistent with the observation that AhR only suppressed induced and not basal p65 activity. However, the Arnt-mediated suppression of RelB suggests that Arnt may interfere with p65-signaling despite its lack of effect on p65 activation (Figures 4 and 5 ). Furthermore, the results suggest that neither AhR nor Arnt suppressed chemokine responses by stabilizing RelB in Poly I:C-exposed BEAS-2B cells. Figure 7 Depletion of AhR and Arnt increases RelB levels in Poly I: C - exposed BEAS- 2B cells. Cells were transfected with siRNA against AhR (siAHR), Arnt (siARNT) or non-targeting control siRNA (siNT), and exposed to 10 μg/ml Poly I:C for 2 and 4 h. Intracellular protein levels of RelB and β-actin were detected by Western blotting as described under "Materials and methods". The figure displays representative blots of RelB levels in unexposed cells (A) and Poly I:C-exposed cells (B) . The graph depicts relative changes in RelB (B) compared to β-actin quantified by densitometric analysis of the Western blots. The results are expressed as mean ± SEM (n = 3). *Significantly different from unexposed controls; †Significantly different from cells transfected with non-targeting siRNA. Notably, RelB has also been implicated in AhR- or Arnt-mediated transcription regulation in various ways [ 11 ],[ 12 ],[ 16 ],[ 20 ],[ 49 ]. So, to further explore whether RelB could be involved in AhR- or Arnt-mediated chemokine suppression, we then examined the effect of RelB silencing on chemokine responses in 1-NP-, 1-AP- or Poly I:C-exposed BEAS-2B cells. RelB depletion by siRNA increased CXCL8 responses (Figure 8 A,C and E) but reduced CCL5 responses (Figure 8 B,D and F). Thus, the effects of RelB partly resembled that of Arnt (Figures 2 and 5 ), by suppressing both basal and induced CXCL8 and not CCL5. However, in contrast to Arnt, RelB seemed to be required for CCL5 responses. The observation that AhR in addition to suppressing CCL5 selectively suppressed induced, but not basal CXCL8 levels (Figures 2 and 5 ), further suggests that AhR did not mediate its anti-inflammatory effects through RelB. Figure 8 RelB suppresses CXCL8, but not CCL5, in BEAS- 2B cells. Cells were transfected with siRNA against RelB (siRELB) or non-targeting control siRNA (siNT). Transfected cells were exposed to 20 μM 1-NP, 1-AP, or vehicle (DMSO) alone (A-D) or 10 μg/ml Poly I:C or vehicle (water) alone (E and F) . CXCL8 (A) and CCL5 (B) gene expression were measured after 6 h by real-time PCR. CXCL8 (C and E) and CCL5 (D and F) protein levels in the medium were measured by ELISA after 18 h exposure as described under "Materials and methods". Efficiency of transfection was assessed by Western blotting (G) . The results are expressed as mean ± SEM (A/B: n = 1 (triplicate determinations); C-F : n ≥ 3; G: representative blot, n = 3) *Significantly different from unexposed controls; †Significantly different from cells transfected with non-targeting siRNA. In contrast to the effects obtained by RelB siRNA (Figure 8 ), silencing its binding-partner in the alternative NF-κB pathways, p100/p52, attenuated both CXCL8 and CCL5 responses (Additional file 5 : Figure S5A, online supplementary materials). Moreover, neither 1-NP, 1-AP or Poly I:C induced degradation of p100 to p52 (Additional file 5 : Figure S5B, online supplementary materials), the hallmark of alternative NF-κB activation [ 50 ]. This suggests that the RelB-mediated suppression of CXCL8 (Figure 8 ) was independent of the alternative NF-κB pathway. However, it is possible that the RelB-dependency of CCL5 (Figure 8 ) may be related to some permissive role of the alternative NF-κB pathway. Activation of AhR by BNF suppresses Poly I:C-induced CXCL8 and CCL5 Our present results suggest that the constitutive activity of unligated AhR and Arnt regulate chemokine responses through differential mechanisms. As a final step, we also explored how ligand-mediated activation of AhR-signaling would affect CXCL8 and CCL5 regulation. It has been reported that the non-toxic AhR agonist β-naphthoflavone (BNF) may suppress TNF-α-induced NF-κB activation and LPS-induced cytokine responses [ 21 ],[ 22 ]. Thus we examined the effect of BNF on Poly I:C-induced CXCL8 and CCL5 in BEAS-2B cells. Pre-incubation with 1 μM BNF for 30 min prior to Poly I:C exposure for 18 h, attenuated both Poly I:C-induced CXCL8 and CCL5 (Figure 9 A and B). This effect seem to extend to other cell types, as BNF suppressed Poly I:C-induced chemokine responses in A549 human alveolar type-II-like epithelial carcinoma cells and THP-1 human leukemia monocytes (Additional file 6 : Figure S6, online supplementary materials). Of notice, Poly I:C appeared unable to induce CXCL8-responses in the A549 cells. We also observed that pre-incubation with another well-known AhR-agonist, B[ a ]P, suppressed both CXCL8 and CCL5 responses in BEAS-2B cells exposed to Poly I:C or LPS (Additional file 7 : Figure S7, online supplementary materials). In contrast, treatment with high concentrations of BNF and ANF (10 and 25 μM, respectively) alone, induced moderate two-fold increases in CXCL8 but not CCL5 (Figure 9 C and D). Therefore it seems that not only the constitutive activity of unligated AhR, but also ligand-mediated activation of the AhR may suppress pro-inflammatory responses in BEAS-2B cells. These results also show that activation of AhR by external ligands increase the suppression of cytokine/chemokine responses compared to the effects of unligated, constitutively active AhR. However, in absence of other, stronger pro-inflammatory stimuli, AhR-ligands elicited a moderat pro-inflammatory effect in BEAS-2B cells when given at sufficiently high concentrations. Figure 9 AhR- activation suppresses Poly I: C- induced CXCL8 and CCL5 responses , but stimulate CXCL8 responses alone in BEAS- 2B cells. Cells were incubated with 1 μM of the AhR agonist β-naphtoflavone (BNF) for 30 min prior to exposure to 10 μg/ml Poly I:C for 18 h (A and B) , or exposed to high concentrations of ANF or BNF alone for 18 h (C and D) . CXCL8 (A and C) and CCL5 (B and D) levels in the medium were measured by ELISA as described under "Materials and methods". The results are expressed as mean ± SEM (n = 3). *Significant increase induced by Poly I:C; †Significant reduction induced by BNF. Discussion In the present study we have compared the effects of AhR and Arnt depletion on the classical NF-κB pathway and expression of two differentially regulated chemokines, CXCL8 and CCL5, in bronchial epithelial BEAS-2B cells exposed to PAHs or the TLR3 agonist Poly I:C. The results showed that regulation of CXCL8, but not CCL5, was dependent on basal and/or induced p65 activity. AhR suppressed both CXCL8 and CCL5 responses in BEAS-2B cells by widely differing stimuli, while Arnt only suppressed CXCL8. Moreover, the AhR-antagonist ANF failed to reproduce the effects of AhR knock-down on CXCL8/CCL5 regulation. This suggests that AhR suppressed chemokine responses independently of Arnt, and the effects were most likely due to the constitutive activity of unligated AhR. AhR suppressed p65 activation induced by strong activators of the classical NF-κB pathway, such as TNF-α and Poly I:C, but did not affect basal p65 activity. In addition, since AhR also suppressed CCL5, as well as the p65-independent increase in CXCL8 by 1-NP (which were unable to enhance p65 activity), the anti-inflammatory effects of AhR were necessarily not restricted to interference with NF-κB-signaling. In contrast to AhR, Arnt seemed unable to suppress the onset of p65 activation. Nevertheless, Arnt may still interfere with p65 signaling through other mechanisms, possibly in collaboration with RelB, which is a well-known suppressor of p65 activity. A schematic presentation of the possible roles of AhR and Arnt in suppression of CXCL8 and CCL5 in BEAS-2B cells, as discussed below, is presented in Figure 10 . In addition to the above, we also observed that potent-AhR agonists could suppress both CXCL8 and CCL5-responses by TLR-ligands, but given alone at sufficiently high concentrations, they elicited a moderate activation of CXCL8. Thus ligand-activated AhR appear to possess both pro- and anti-inflammatory effects in BEAS-2B cells. Figure 10 The constitutive activity of unligated AhR and Arnt differentially regulate CXCL8 and CCL5 in BEAS- 2B cells. The model summarizes the possible pathways discussed for AhR- and Arnt-mediated regulation of CXCL8 and CCL5 in BEAS-2B cells exposed to 1-NP and 1-AP (A) or Poly I:C (B) . "x/X" and "y/Y" represents hitherto unidentified signaling pathways/response elements involved in regulation of CXCL8 and CCL5, respectively. The positioning of the response elements in the CXCL8 or CCL5 promoters is not meant to be representative. Moreover, "y/Y" induced by 1-AP (A) and Poly I:C (B) are not necessarily identical and AhR may also affect other pathways involved in Poly I:C-induced CXCL8 in addition to NF-κB (B) . The suggested position and association of Arnt and RelB is partly based on findings reported by Wright and Duckett [ 49 ]. The role of AhR and Arnt in chemokine regulation may differ in other cell types. In an elegant series of experiments Chen and colleagues [ 9 ] showed that AhR could induce IL-6 expression and NF-κB activity in BEAS-2B cells and H1355 human lung adenocarcinoma cells. Overexpression of AhR resulted in increased formation of AhR/p65-complexes that activated κB-sites in the IL-6 promoter, while TCDD-exposure activated the classical p65/p50-heterodimer [ 9 ]. In line with this, we observed that high concentrations of the AhR-ligands ANF and BNF induced a moderate increase in CXCL8, but not the p65-independent CCL5. However, in extension of the findings by Chen et al., [ 9 ] we further observed that AhR in the same cell type may also possess anti-inflammatory properties, and we propose that whether AhR act as suppressor or inducer of inflammation largely depends on the type and combination of stimuli. Importantly, our present results show that AhR depletion did not suppress basal p65 activity or CXCL8 levels in unexposed cells. However, NF-κB activation induced by TNF-α and Poly I:C was significantly enhanced. This may at least partly, be due to AhR-mediated suppression of p65-phosphorylation at Ser536, which is important to CXCL8 transcription by promoting coupling of p65 to the basal transcriptional machinery [ 45 ],[ 46 ],[ 51 ]. Phosphorylation of p65 at Ser536 is mediated by Inhibitor of κB Kinases (IKKs) [ 45 ],[ 51 ]. Recent findings suggest that AhR may interact directly with IKKs [ 52 ], providing a possible explanation for the observed effects of AhR knock-down on Poly I:C-induced Ser536 phosphorylation of p65. In addition, the study by Chen and co-workers [ 9 ] indicates that AhR/p65 may be a less efficient activator of κB-sites than p65/p50 (TCDD-exposure induced more than twice the NF-κB activity elicited by AhR-overexpression, despite almost identical binding of p65 to the IL-6 promoter). It is therefore conceivable that AhR/p65-dimerization may restrict the availability of free p65 levels, thus limiting formation of the more potent p65/p50-dimers when cells are exposed to strong, AhR-independent, activators of the classical NF-κB pathway such as Poly I:C, TNF-α or LPS. A recent publication by Vogel and colleagues [ 53 ] offers an interesting perspective to our current findings. In their study, they found that AhR was under transcriptional regulation by p65, and that LPS exposure induced AhR expression in human dendritic cells [ 53 ]. Thus, the suppressive effects of AhR on p65 activity may possibly be ascribed to a negative feedback mechanism. However, in contrast to these observations Poly I:C exposure did not induced AhR expression in BEAS-2B cells, but rather appeared to reduce the level of AhR. In any case, the scenario of AhR as a partial NF-κB agonist acting antagonistic in competition with stronger NF-kB activators may provide an explanation for how different studies have observed that AhR either induces or suppresses NF-κB signaling. In contrast to AhR, Arnt did not seem to affect activation of the classical NF-κB pathway as assessed by p65-luciferace activity or p65 phosphorylation. Yet, Arnt depletion enhanced the expression of the NF-κB target gene RelB [ 47 ],[ 49 ] in both un-stimulated and Poly I:C-exposed BEAS-2B cells. In addition, we observed that Arnt silencing selectively enhanced the p65-dependent chemokine CXCL8 and not the p65-independent CCL5. It seems that although Arnt did not appear to suppress the activation of p65, it may possibly still interfere with p65 signaling through some other mechanisms. Of notice, we further observed that RelB itself suppressed CXCL8 and not CCL5, in parallel with Arnt. In addition to being transcriptionally regulated by p65/p50, RelB is also a well-known suppressor of the classical NF-κB pathway and inflammation that may represent a negative feed-back loop for NF-κB regulation [ 48 ],[ 49 ]. Interestingly, Wright and Duckett [ 49 ] showed that Arnt facilitated RelB binding to NF-κB responsive promoters preventing p65/p50-binding and suppressing NF-κB-mediated transcription in CD30-exposed T-cells (Karpas 299 cells). In coherence with our present findings, they also observed that Arnt silencing increased RelB levels in CD30-stimulated cells [ 49 ]. Whether Arnt facilitates RelB-mediated suppression of p65 signaling also in BEAS-2B cells is an intriguing scenario that requires further investigation. If this is the case, the reason Arnt depletion did not affect the NF-κB luciferase-assay is likely that the Arnt/RelB-binding domain of wild-type NF-κB responsive promoters was lacking in the luciferase reporter-gene promoter. Others have shown that increased inflammation in lungs of LPS and cigarette smoke (CS)-exposed AhR −/− mice were associated with a rapid loss of RelB [ 16 ]. Thus, the anti-inflammatory role of AhR has been suggested to be due to RelB-stabilization [ 20 ]. Our present data does not suggest that this is the case in BEAS-2B cells. Silencing of AhR enhanced RelB levels in BEAS-2B cells, although first after 4 h exposure to Poly I:C. This is consistent with the observation that AhR suppressed activation of p65 by Poly I:C and TNF-α, but not basal p65 activity. In addition, we also observed that AhR and RelB were differentially involved in regulation of CXCL8 and CCL5, which further suggest that AhR does not mediate its effects through RelB. Notably, although transfection with AhR siRNA did not affect p65 activity in 1-NP-exposed cells, the AhR silencing increased 1-NP-induced CXCL8 without affecting basal CXCL8 levels. In addition, AhR (in contrast to Arnt) also suppressed the p65-independent CCL5 responses in the BEAS-2B cells. This suggests that suppression of the classical NF-κB pathway cannot be the only mechanism for the anti-inflammatory effects of the AhR, but that at least two additional pathways must be affected (Figure 9 ). In line with this, it has been reported that the activity of other pro-inflammatory transcription factors such as activator protein-1 (AP-1) and CCAAT-enhancer binding protein (C/EBP) may be affected by AhR [ 54 ],[ 55 ]. Therefore it is tempting to speculate that the differential suppression of CXCL8 and CCL5 by AhR and Arnt in BEAS-2B cells could be due suppression of a broader range of pro-inflammatory pathways by AhR, or through effects on post-transcriptional or translational processes. The majority of the present work focused on the anti-inflammatory role of constitutive (unligated) AhR and Arnt activity. However, we also observed that the AhR-agonist BNF as well as B[ a ]P suppressed Poly I:C-induced CXCL8 or CCL5 responses. Thus, ligand-mediated activation increased chemokine suppression by AhR, beyond the effects of the constitutive activity of the unligated receptor. Similarly BNF and other AhR ligands have been reported to suppress LPS-induced cytokine expression and inflammation, in vitro and in vivo [ 18 ],[ 21 ]. Moreover, in a recent study TCDD was found to suppress silica-induced inflammation [ 17 ]. Considering that TCDD has been reported to induce p65/p50 through AhR [ 9 ] and that our present CCL5-responses were unaffected by p65 depletion (Figures 4 and 7 ), it seems unlikely that BNF and B[ a ]P attenuated Poly I:C-induced CXCL8 and CCL5 through suppression of NF-κB signaling in the BEAS-2B cells. However, whether unligated and ligand-activated (i.e. endogenously and exogenously activated) AhR suppress inflammation through similar pathways remains to be clarified. The present study was performed in BEAS-2B cells, a common model of human bronchial epithelial cells which has been reported to exhibit the highest homology in gene expression pattern with primary lung cells as compared to other established lung cell lines [ 56 ]. As any result obtained with immortalized cell lines, generalizations should always be interpreted with some caution. However, several in vivo studies with AhR knock-out mice or in primary mouse cells support the notion that AhR suppresses induction of NF-κB activity and inflammatory reactions in different cells and tissues [ 16 ],[ 17 ],[ 19 ]–[ 21 ]. Moreover, in control experiments with A549 and THP-1 cells, we observed comparable inhibition of Poly I:C-induced chemokine responses by the AhR agonist BNF (Additional file 5 : Figure S5, online supplementary materials), further suggesting that the present findings are not restricted to the BEAS-2B cell line. However, in a recent paper, Vogel and colleagues [ 57 ] observed that combined exposure with TCDD enhanced the expression of IL-6, IL-10, IL-22, IL-23, CXCL2 and CXCL3 in LPS-stimulated dendritic cells. Only LPS-induced DC-CK1 was suppressed by TCDD exposure, while IL-12 and TNFα responses were unaffected [ 57 ]. Whether these discrepancies with our present observations are related to cell-specific effects or variations in effects of AhR in different target genes remains to be clarified. Conclusion Understanding the role of AhR and Arnt in the regulation of inflammation may be crucial to elucidate the pro-inflammatory effects of complex chemical mixtures, such as air pollution which may contain a variety of organic chemicals with varying affinity for the AhR [ 58 ]. Moreover, individual variation in AhR and Arnt expression levels appear to be considerable [ 59 ],[ 60 ] and may possibly affect susceptibility towards inflammatory disease. Though it remains to unravel the precise mechanisms by which AhR and its partners modulate inflammation, our present results suggest that depending on the type and combination of stimuli AhR may possess both pro- and anti-inflammatory functions, even within a single cell type. Both constitutive and ligand-activated AhR may elicit a weak to moderate pro-inflammatory signal, but seems to restrict activation of p65 upon encounters with strong activators of the classical NF-κB pathway, such as Poly I:C or TNF-α. The AhR binding partner Arnt displays a separate anti-inflammatory role. While AhR appears to attenuate the onset of p65 activation, including phosphorylation at Ser536, Arnt may possibly block the access of activated p65 to kB-sites in NF-κB-regulated promoters (Figure 10 ). This would fit with the traditional perception of a differential intracellular localization of AhR and Arnt in cytosol and nucleus of resting cells, respectively [ 9 ]. However, Arnt may also be present in the cytosol [ 49 ], and our current findings suggest that AhR and Arnt were relatively evenly distributed across the two compartments in unstimulated BEAS-2B cells. Finally, our results suggest that both unligated and ligand-activated AhR also suppress pro-inflammatory responses through pathways other than NF-κB. Thus, AhR and Arnt appear to play complex many-faceted roles in the regulation of pro-inflammation reactions in lung cells. Materials and methods Reagents Culture medium, Nutrition Mixture F12 HAM Kaigin's modification (F12K) was obtained from Sigma–Aldrich, LHC-9 medium was from Invitrogen, while RPMI 1640 and fetal bovine serum (FBS) were from Gibco BRL. Ampicillin and fungizone were from Bristol-Myer Squibb, and penicillin/streptomycin was from Bio Whittaker. B[ a ]P, 1-NP, 1-AP, α-naphthoflavone (ANF), β-naphthoflavone (BNF), dimethyl sulphoxide (DMSO), Poly I:C-potassium salt, antibodies against β-actin, short interfering RNA (siRNA) against AhR (sense: GGACAAACUUUCAGUUCUU, anti-sense: AAGAACUGAAAGUUUGUCC) and RelB (sense: GACAAGAAAUCCACAAACA, anti-sense: UGUUUGUGGAUUUCUUGUC) and non-targeting control siRNA (sense: UAGCGACUAAACACAUCAA, anti-sense: UUGAUGUGUUUAGUCGCUA) were from Sigma-Aldrich. SiRNA against Arnt (ON-TARGET plus Smart Pool L-007207-00-0005) was from Dharmacon RNAi Technologies (Thermo Fischer Scientific). Antibodies against AhR, Arnt, RelB and p52, and siRNA against p52 (sc-29409) with corresponding non-targeting control siRNA (sc-37007) were from Santa Cruz Biotechnology. Antibodies against p65, phospho-p65 (Ser536), IκB, histone H1 and GAPDH, ChIP grade antibody against p65 (D14E12), and SiRNA against p65 (SignalSilence® NF-κB p65 siRNA I #6261) with corresponding non-targeting control siRNA (SignalSilence® Control siRNA #6568) were from Cell Signaling Technology. ChIP grade antibodies against AhR (ChIP grade ab2) and Arnt (ChIP grade ab2770) were from Abcam plc. All other chemicals used were purchased from commercial sources at the highest purity available. Cell cultures BEAS-2B cells, a SV40 hybrid (Ad12SV40) transformed human bronchial epithelial cell line, were from European Collection of Cell Cultures. Cells were maintained in serum-free LHC-9 medium in collagen-coated (PureColTM, Inamed Biomaterials) flasks. The human alveolar type-II-like carcinoma cell line A549 and the human leukemia monocyte cell line THP-1 were obtained from the American Type Culture Collection (Manhasset, VA, USA). A549 cells were cultured in F12K medium, supplemented with ampicillin (100 μg/ml), penicillin/streptomycin (100 μg/ml), fungizone (0.25 μg/ml) and 10% heat-inactivated FBS, while THP-1 cells were cultured in RPMI 1640 medium supplemented with 100 μg/ml gentamicin and 10% heat-inactivated FBS fetal calf serum. Cells were cultured in a humidified atmosphere at 37°C with 5% CO 2 , and passaged twice per week. Prior to exposure, cells were plated in 12-well culture dishes, grown to near confluence and treated as described elsewhere. When used, DMSO concentrations in all samples were below 0.5%. Gene expression Total RNA was isolated using Absolutely RNA Miniprep Kit (Stratagene, La Jolla, CA, USA) and reverse transcribed to cDNA on a PCR System 2400 (PerkinElmer) using a High Capacity cDNA Archive Kit (Applied Biosystems). Real-time PCR was performed using pre-designed TaqMan Gene Expression Assays and TaqMan Universal PCR Master Mix and run on ABI 7500 fast (Applied Biosystems). Gene expression of CXCL8 (Hs00174103_m1), CCL5 (Hs00174575_m1) and CYP1A1 (Hs00153120_m1) were normalized against 18S rRNA (Hs99999901_s1), and expressed as fold change compared to untreated control as calculated by the ΔΔCt method (ΔCt = Ct [Gene of Interest] – Ct [18S]; ΔΔCt = ΔCt [Treated] – ΔCt [Control]; Fold change = 2 [−ΔΔCt] ). Chemokine release Chemokine protein levels in cell-supernatants were determined by ELISA assays for CXCL8 (Human IL-8 Cytoset) and CCL5 (Human RANTES Cytoset) from Biosource International as described elsewhere [ 24 ]. Absorbance was measured using a plate reader (TECAN Sunrise, Phoenix Research Products) complete with software (Magellan V 1.10). Gene silencing by siRNA BEAS-2B cells were reverse-transfected with the respective siRNAs, using HiPerFect transfection reagent according to the Fast-Forward protocol for adherent cells recommended by Qiagen adapted for 12-well plate format, to give a final siRNA concentration of 10 nM and 2.75 μl of HiPerFect in a total of 1 ml growth medium. The effectiveness of gene silencing was monitored at 48, 72 and 96 h by measuring protein levels by Western blotting. NF-κB luciferase assay Cells seeded at 55x10 3 cells/cm 2 were transfected 24 h later with the respective siRNAs using DharmaFECT 1 (Thermo Fischer Scientific) according to protocol. The next day, reporter gene constructs were transiently transfected into cells. For this, the media were changed to opti-MEM (Gibco) containing 2.5 μl/ml of Fugene 6 (Roche), 375 ng of a NF-kB-driven luciferase reporter vector (Stratagene ref: 219058–51) and 40 ng of a renilla reporter vector, used as an internal control, according to the manufacturer's instructions. 24 h after this second transfection, cells were exposed for 6 or 16 h to DMSO or 1-NP or TNFα. Cell extracts were assayed for luciferase activity with a Dual-Luciferase® Reporter assay system from Promega according to the manufacturer's instructions. Immunoblotting Total and phosphorylated protein levels were analyzed by Western blotting as described elsewhere [ 61 ]. The Compartment Protein Extraction Kit (BioChain Institute, Inc.) was used for or extraction/isolation of nuclei and cytosol according to producer's recommendations. The blots were developed and quantified using a ChemiDox™ XRS + molecular imager with Image Lab™ software, Bio-Rad Laboratories Inc. (Hercules, CA, USA). Immunocytochemistry Cells were grown in X-well Tissue Culture Chambers (Sarstedt). Effect of treatment were stopped by washing with ice-cold PBS, and cells were fixed by incubation inn methanol for 4 min, before incubation overnight with antibodies against p65, AhR or Arnt (working dilution 1: 200 in PBS with 1% BSA). After washing and incubating for 3 h with secondary antibodies conjugated with Alexa Fluor®488 or 594, the preparations were visualized using a Zeiss AxioObserved.Z1 equiped with an AxioCam ERc 5 s digital camera (Zeiss). Chromatin immunoprecipitation (ChIP) assay ChIP assay was performed according to the protocol specified by the manufacturer (HighCell# ChIP kit, Diagenode Inc.), In brief, treated cells were collected by trypsinisation, and DNA-protein interactions were cross-linked with 1% formaldehyde. Chromatin shearing was performed by sonication at 4°C with a Bioruptor® Plus equipped with a Minichiller (Diagenode Inc.). Lysates were centrifuged, and aliquots (1%) of the supernatants were collected for input control, while the remaining supernatants were incubated overnight (at 4°C) with magnetic beads coated with ChIP grade antibodies against p65, AhR or Arnt. DNA was isolated from the collected precipitates after reversal of the cross-linking by incubation for 15 min at 55°C. Real-time PCR were performed using primer sets for human CXCL8 and CCL5 promoter regions containing NF-κB response elements [ 62 ],[ 63 ]. The sequences of primers used for ChIP assay were as follows: human CXCL8 sense 5'-AGTGTGATGACTCAGGTTTGCCCT-3' and anti-sense 5'-AAGCTTGTGTGCTCTGCTGTCTCT-3'; human CCL5 promoter sense 5'-GGGAAGAAGATTGCCTAAAC-3' and antisense 5'-TGTGGAAATCAAAGGGACAG-3'. Statistical analysis Statistical significance was evaluated by GraphPad Prism software (GraphPad Software Inc., San Diego, CA, USA), using analysis of variance (ANOVA) with Bonferroni post-test. Reagents Culture medium, Nutrition Mixture F12 HAM Kaigin's modification (F12K) was obtained from Sigma–Aldrich, LHC-9 medium was from Invitrogen, while RPMI 1640 and fetal bovine serum (FBS) were from Gibco BRL. Ampicillin and fungizone were from Bristol-Myer Squibb, and penicillin/streptomycin was from Bio Whittaker. B[ a ]P, 1-NP, 1-AP, α-naphthoflavone (ANF), β-naphthoflavone (BNF), dimethyl sulphoxide (DMSO), Poly I:C-potassium salt, antibodies against β-actin, short interfering RNA (siRNA) against AhR (sense: GGACAAACUUUCAGUUCUU, anti-sense: AAGAACUGAAAGUUUGUCC) and RelB (sense: GACAAGAAAUCCACAAACA, anti-sense: UGUUUGUGGAUUUCUUGUC) and non-targeting control siRNA (sense: UAGCGACUAAACACAUCAA, anti-sense: UUGAUGUGUUUAGUCGCUA) were from Sigma-Aldrich. SiRNA against Arnt (ON-TARGET plus Smart Pool L-007207-00-0005) was from Dharmacon RNAi Technologies (Thermo Fischer Scientific). Antibodies against AhR, Arnt, RelB and p52, and siRNA against p52 (sc-29409) with corresponding non-targeting control siRNA (sc-37007) were from Santa Cruz Biotechnology. Antibodies against p65, phospho-p65 (Ser536), IκB, histone H1 and GAPDH, ChIP grade antibody against p65 (D14E12), and SiRNA against p65 (SignalSilence® NF-κB p65 siRNA I #6261) with corresponding non-targeting control siRNA (SignalSilence® Control siRNA #6568) were from Cell Signaling Technology. ChIP grade antibodies against AhR (ChIP grade ab2) and Arnt (ChIP grade ab2770) were from Abcam plc. All other chemicals used were purchased from commercial sources at the highest purity available. Cell cultures BEAS-2B cells, a SV40 hybrid (Ad12SV40) transformed human bronchial epithelial cell line, were from European Collection of Cell Cultures. Cells were maintained in serum-free LHC-9 medium in collagen-coated (PureColTM, Inamed Biomaterials) flasks. The human alveolar type-II-like carcinoma cell line A549 and the human leukemia monocyte cell line THP-1 were obtained from the American Type Culture Collection (Manhasset, VA, USA). A549 cells were cultured in F12K medium, supplemented with ampicillin (100 μg/ml), penicillin/streptomycin (100 μg/ml), fungizone (0.25 μg/ml) and 10% heat-inactivated FBS, while THP-1 cells were cultured in RPMI 1640 medium supplemented with 100 μg/ml gentamicin and 10% heat-inactivated FBS fetal calf serum. Cells were cultured in a humidified atmosphere at 37°C with 5% CO 2 , and passaged twice per week. Prior to exposure, cells were plated in 12-well culture dishes, grown to near confluence and treated as described elsewhere. When used, DMSO concentrations in all samples were below 0.5%. Gene expression Total RNA was isolated using Absolutely RNA Miniprep Kit (Stratagene, La Jolla, CA, USA) and reverse transcribed to cDNA on a PCR System 2400 (PerkinElmer) using a High Capacity cDNA Archive Kit (Applied Biosystems). Real-time PCR was performed using pre-designed TaqMan Gene Expression Assays and TaqMan Universal PCR Master Mix and run on ABI 7500 fast (Applied Biosystems). Gene expression of CXCL8 (Hs00174103_m1), CCL5 (Hs00174575_m1) and CYP1A1 (Hs00153120_m1) were normalized against 18S rRNA (Hs99999901_s1), and expressed as fold change compared to untreated control as calculated by the ΔΔCt method (ΔCt = Ct [Gene of Interest] – Ct [18S]; ΔΔCt = ΔCt [Treated] – ΔCt [Control]; Fold change = 2 [−ΔΔCt] ). Chemokine release Chemokine protein levels in cell-supernatants were determined by ELISA assays for CXCL8 (Human IL-8 Cytoset) and CCL5 (Human RANTES Cytoset) from Biosource International as described elsewhere [ 24 ]. Absorbance was measured using a plate reader (TECAN Sunrise, Phoenix Research Products) complete with software (Magellan V 1.10). Gene silencing by siRNA BEAS-2B cells were reverse-transfected with the respective siRNAs, using HiPerFect transfection reagent according to the Fast-Forward protocol for adherent cells recommended by Qiagen adapted for 12-well plate format, to give a final siRNA concentration of 10 nM and 2.75 μl of HiPerFect in a total of 1 ml growth medium. The effectiveness of gene silencing was monitored at 48, 72 and 96 h by measuring protein levels by Western blotting. NF-κB luciferase assay Cells seeded at 55x10 3 cells/cm 2 were transfected 24 h later with the respective siRNAs using DharmaFECT 1 (Thermo Fischer Scientific) according to protocol. The next day, reporter gene constructs were transiently transfected into cells. For this, the media were changed to opti-MEM (Gibco) containing 2.5 μl/ml of Fugene 6 (Roche), 375 ng of a NF-kB-driven luciferase reporter vector (Stratagene ref: 219058–51) and 40 ng of a renilla reporter vector, used as an internal control, according to the manufacturer's instructions. 24 h after this second transfection, cells were exposed for 6 or 16 h to DMSO or 1-NP or TNFα. Cell extracts were assayed for luciferase activity with a Dual-Luciferase® Reporter assay system from Promega according to the manufacturer's instructions. Immunoblotting Total and phosphorylated protein levels were analyzed by Western blotting as described elsewhere [ 61 ]. The Compartment Protein Extraction Kit (BioChain Institute, Inc.) was used for or extraction/isolation of nuclei and cytosol according to producer's recommendations. The blots were developed and quantified using a ChemiDox™ XRS + molecular imager with Image Lab™ software, Bio-Rad Laboratories Inc. (Hercules, CA, USA). Immunocytochemistry Cells were grown in X-well Tissue Culture Chambers (Sarstedt). Effect of treatment were stopped by washing with ice-cold PBS, and cells were fixed by incubation inn methanol for 4 min, before incubation overnight with antibodies against p65, AhR or Arnt (working dilution 1: 200 in PBS with 1% BSA). After washing and incubating for 3 h with secondary antibodies conjugated with Alexa Fluor®488 or 594, the preparations were visualized using a Zeiss AxioObserved.Z1 equiped with an AxioCam ERc 5 s digital camera (Zeiss). Chromatin immunoprecipitation (ChIP) assay ChIP assay was performed according to the protocol specified by the manufacturer (HighCell# ChIP kit, Diagenode Inc.), In brief, treated cells were collected by trypsinisation, and DNA-protein interactions were cross-linked with 1% formaldehyde. Chromatin shearing was performed by sonication at 4°C with a Bioruptor® Plus equipped with a Minichiller (Diagenode Inc.). Lysates were centrifuged, and aliquots (1%) of the supernatants were collected for input control, while the remaining supernatants were incubated overnight (at 4°C) with magnetic beads coated with ChIP grade antibodies against p65, AhR or Arnt. DNA was isolated from the collected precipitates after reversal of the cross-linking by incubation for 15 min at 55°C. Real-time PCR were performed using primer sets for human CXCL8 and CCL5 promoter regions containing NF-κB response elements [ 62 ],[ 63 ]. The sequences of primers used for ChIP assay were as follows: human CXCL8 sense 5'-AGTGTGATGACTCAGGTTTGCCCT-3' and anti-sense 5'-AAGCTTGTGTGCTCTGCTGTCTCT-3'; human CCL5 promoter sense 5'-GGGAAGAAGATTGCCTAAAC-3' and antisense 5'-TGTGGAAATCAAAGGGACAG-3'. Statistical analysis Statistical significance was evaluated by GraphPad Prism software (GraphPad Software Inc., San Diego, CA, USA), using analysis of variance (ANOVA) with Bonferroni post-test. Competing interests The authors declare that they have no competing interests. Authors' contributions JØ participated in conceiving and designing the study, carried out the majority of experiments, performed the statistical analysis and drafted the manuscript. VL and DG participated in designing the siRNA experiments and drafting of the manuscript. VL also carried out the luciferase assay. TS carried out the studies in A549 cells and participated in drafting the manuscript. RB performed the immunocytochemistry and participated in drafting the revised manuscript. ML, MR, PES and JAH participated in the conceiving and design of the study and helped to draft the manuscript. DLG participated in the design of the study, helped to draft the manuscript, and together with JAH coordinated the collaboration between NIPH and the Inserm institute/University of Rennes. All authors read and approved the final manuscript. Additional files Supplementary Material Additional file 1: Figure S1. AhR, Arnt, p65 and RelB differentially regulate basal CXCL8 andCCL5 levels in BEAS-2B cells. Cells were transfected with siRNA against AhR (siAHR), Arnt (siARNT), p65 (siP65), RelB (siRELB) or non-targeting control siRNA (siNT). Basal levels of CXCL8 (A, C and E) and CCL5 (B, D and F) protein levels in the medium (produced over a period of 18 h) were measured by ELISA as described under "Materials and methods". The results are expressed as mean ± SEM (n ≥6). *Significantly different from cell transfected with non-targeting siRNA. Click here for file Additional file 2: Figure S2. Poly I:C induce both CXCL8 and CCL5 responses and mediates IκB degradation and p65 phosphorylation in BEAS-2B cells. Cells were exposed different concentrations of Poly I:C for 18 h, and CXCL8 (A) and CCL5 (B) levels in the medium were measured by ELISA as described under "Materials and methods". Subsequently, cells were exposed to 20 μM 1-NP, 1-AP or 10 μg/ml of Poly I:C for 2 and 4 h (n = 2). Intracellular protein levels of IκB, β-actin, p65 and phosphor-p65 (Ser536) were detected by Western blotting. The figure displays representative blots as well as changes in IκB and phospho-p65 (p-p65) relative to β-actin or total p65, respectively, as quantified by densitometric analysis of the Western blots. The results are expressed as mean ± SEM (n = 2). Click here for file Additional file 3: Figure S3 AhR suppress CXCL8 and CCL5 responses in Poly I:C-exposed BEAS-2B cells. Cells were transfected with commercially available siRNA (Dharmacon RNAi Technologies, Thermo Fischer Scientific) against AhR (siAHR; ON-TARGET plus Smart Pool L-004990-00) or non-targeting control siRNA (siNT; ON-TARGET Control Pool D-001810-10-05), and exposed to 10 μg/ml Poly I:C. CXCL8 (A) and CCL5 (B) protein levels in the medium were measured by ELISA after 18 h exposure, as described under "Materials and methods", and expressed as fold increase compared to untreated controls transfected with siNT. The results are expressed as mean ± SEM (n = 2-3). Click here for file Additional file 4: Figure S4. Sub-cellular localization of p65, AhR and Arnt in 1-NP and Poly I:C exposed BEAS-2B cells. Cells were exposed to 1-NP (20 μM) or Poly I:C (10 μg/ml) for 2.5 h. Sub-cellular localization of p65, AhR and Arnt were detected by immunocytochemistry with specific antibodies, as described under "Materials and methods". Click here for file Additional file 5: Figure S5. Characterization of the role of alternative NF-κB signaling in CXCL8 and CCL5 responses in BEAS-2B cells. Cells were transfected with siRNA against p100/p52 (siP52) or non-targeting control siRNA (siNT), and exposed to 20 μM 1-NP, 1-AP or 10 μg/ml Poly I:C for 18 h. CXCL8 (A) and CCL5 (B) levels in the medium were measured by ELISA as described under "Materials and methods". Efficiency of transfection with siP52 was assessed by Western blotting (C). Subsequently, non-transfected cells were exposed to 20 μM 1-NP, 1-AP or 10 μg/ml of Poly I:C for 2 and 4 h. Intracellular protein levels of p100, p52 and β-actin were detected by Western blotting (D). The figure presents results obtained from 1 experiment. Click here for file Additional file 6: Figure S6. AhR-activation suppresses Poly I:C-induced chemokine responses in A549 and THP-1 cells. A549 (A and C) and THP-1 (B and D) cells were incubated with 1 μM of the AhR agonist β-naphtoflavone (BNF) for 30 min prior to exposure to 10 μg/ml Poly I:C for 18 h. CXCL8 (A and B) and CCL5 (C and D) levels in the medium were measured by ELISA as described under "Materials and methods". *Significant increase induced by Poly I:C; †Significant reduction induced by BNF. The results are expressed as mean ± SEM (1 experiment performed in triplicates). Click here for file Additional file 7: Figure S7. Pre-treatment with B[ a ]P suppress LPS and Poly I:C induced chemokine responses in BEAS-2B cells. Cells were incubated with 25 μM B[ a ]P for 2 h or 24 h prior to exposure to 10 μg/ml LPS or Poly I:C for 18 h. CXCL8 (A) and CCL5 (B) levels in the medium were measured by ELISA as described under "Materials and methods". The results are expressed as mean ± SEM (n = 2). Click here for file Acknowledgement We thank E. Lilleaas (Norwegian Inst. of Public Health, Oslo Norway) for technical assistance throughout the study and Professor O. Fardel (UMR INSERM 1085, Rennes, France) for his helpful advice. The work was supported by the Research Council of Norway, through the Environmental Exposures and Health Outcomes-program (MILPAAHEL; grants no. 185620 and 228143).

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3064645/

Identification of a mutated BHK-21 cell line that became less susceptible to Japanese encephalitis virus infection

The pathogenesis of Japanese encephalitis virus (JEV) is not definitely elucidated as the initial interaction between virus and host cell receptors required for JEV infection is not clearly defined yet. Here, in order to discover those membrane proteins that may be involved in JEV attachment to or entry into virus permissive BHK-21 cells, a chemically mutated cell line (designated 3A10-3F) that became less susceptible to JEV infection was preliminarily established and selected by repeated low moi JEV challenges and RT-PCR detection for viral RNA E gene fragment. The susceptibility to JEV of 3A10-3F cells was significantly weakened compared with parental BHK-21 cells, verified by indirect immunofluorescence assay, virus plague formation assay, and flow cytometry. Finally, two-dimensional electrophoresis (2-DE) coupled with LC-MS/MS was utilized to recognize the most differentially expressed proteins from membrane protein extracts of 3A10-3F and BHK-21 cells respectively. The noted discrepancy of membrane proteins included calcium binding proteins (annexin A1, annexin A2), and voltage-dependent anion channels proteins (VDAC 1, VDAC 2), suggesting that these molecules may affect JEV attachment to and/or entry into BHK-21 cells and worthy of further investigation. Findings Japanese encephalitis virus (JEV), a member of genus Flavivirus in the family Flaviviridae , is the causative agent of Japanese encephalitis (JE), the mosquito-borne viral encephalitis epidemic in eastern, southeastern and southern Asia, leading to an estimated ~50,000 infections annually, of which ~15,000 will die and up to 50% of survivors are left with severe residual neurological complications [ 1 , 2 ]. During the past decades, JE is spreading beyond its traditional boundaries and has reported from previously unaffected areas such as Saipan islands, Pakistan and northern Australia [ 3 - 5 ]. Coupled with a high rate of mortality and residual neurological complications in survivors, it makes JE a serious public health problem in tropical and subtropical areas in the world. The first step of virus infection requires the interaction between virus attachment proteins (VAPs) and cellular receptors, which is known to contribute to host range, tissue tropism and viral pathogenesis. In the cases of flaviviruses including JEV, envelope glycoprotein E, protruding as spikes on the surface of virions, is considered to be the dominant antigen in mediating receptor binding and membrane fusion, hemagglutination, neutralization and virulence [ 6 - 8 ]. As of enzootic nature, JEV maintains a natural cycle among birds, pigs, and other vertebrate hosts by mosquitoes without serious sickness, and thus the cells from above species, such as African green monkey kidney cells (Vero), baby hamster kidney cells (BHK-21), and Aedes albopictus cells (C6/36) [ 9 ], are frequently applied in studies associated with JEV pathogenicity due to their ability to permit JEV entry and replication within them. Such broad tropism of JEV rationally suggests that most possibly there exists more than one cellular receptor responsible for virus binding and entry into susceptible cells above mentioned. To date, little is known about JEV cellular receptors. Much earlier, a report stated that a 74 kDa protein on Vero cells was found to be capable of binding JEV and might be involved in virus uptake process [ 10 ]. A recent paper also indicated that several proteins on the surface of C6/36 cells with masses ranging from 35-80 kDa and 150-200 kDa may bind to JEV, but failed to identify specific proteins by mass spectroscopic fingerprint analysis [ 11 ]. In 2009, a paper reported that heat shock protein 70 is a putative receptor for JEV on mouse neuroblastoma (Neuro2a) cells [ 12 ]. Hence, the detailed interaction between JEV and its putative receptor(s) is not exclusively defined yet. One of the most convincing methods to verify a putative virus receptor is to transfer the receptor gene into a cell line that cannot bind virus and later demonstrate that the receptor-negative cell acquires the ability to bind virus and permit virus to replicate within it after the receptor gene is regained [ 13 ]. For this reason, the availability of a specific virus receptor-negative/-defective cell line is usually a prerequisite for virus receptor confirmation. Unfortunately, such JEV receptor-negative/-defective cell line is not currently available after our extensive searching for animal and human cell lines, and thus it has to be established artificially. The successful identification of the cellular receptor for anthrax toxin provided us a practical strategy to create any specific virus receptor-negative/-defective cells [ 14 ]. Here, BHK-21 cells, permissive to JEV entry and replication, were subject to co-culture with a DNA alkylating mutagen ICR-191 to introduce random small DNA deletions and frame shift mutations in the genes of normally cultured BHK-21 cells under conditions that led to ~90% cell death. The survived cells were grown up but underwent a slow die-off for 4 weeks. These viable but mutated BHK-21 cells were successively subject to several rounds of infectious JEV challenge at a low moi of 0.1, subcloned by limiting dilution, and then each individual subclones was detected by RT-PCR for negative JEV E gene RNA fragment. RT-PCR primers were synthesized according to the RNA sequence of JEV SA-14 strain [ 15 ] from 1528 to 2234 of E gene conserved fragment, which were sense-strand primer, 5'-CGGAATTCGAGAAGTCACACTGGACTGTGAGCC-3', and antisense-strand primer, 5'-CGCTGCCAGTCTTTGAGCTCCCTTCAAAGT-3'. Finally, such a JEV RNA-negative mutated cell line, designated 3A10-3F, was picked out after two cycles of limiting dilution cloning from eight subclones of mutated BHK-21 cells (Figure 1 ). Further, the susceptibility of 3A10-3F to JEV infection was tested by indirect immunofluorescence assay (IFA), viral plaque formation assay, and flow cytometry. In IFA, JEV E protein synthesized within the cells was detected by monoclonal antibody against E protein. The results clearly displayed that fluorescent signals from 3A10-3F were much weaker (Figure 2b ) than BHK-21 cells (Figure 2d ) with the same dosage of JEV, while both negative controls of the two cell lines appeared the same (Figure 2a and Figure 2c ). Figure 1 Selection of mutated cell lines negative of JEV E gene RNA fragments by RT-PCR . Eight single-cell clones of mutated cells (lanes 4-11) were detected in triplicate. The arrow indicates 707 bp fragments containing the conserved JEV E gene sequence. Only one single-cell clone, designated 3A10-3F, showed JEV RNA negative (lane 8). BHK-21 cells (lane 1) and BHK-21 cells infected with JEV (lane 2) were negative and positive controls respectively. Lane 3 was DL-2000 DNA marker. Figure 2 JEV binding to 3A10-3F and BHK-21 cells measured by indirect immunofluorescence assay (IFA) . JEV antigen were detected with anti-JEV mAb and FITC-conjugated goat anti-mouse IgG at 24 hr post-infection, and uninfected 3A10-3F cells (a) and BHK-21 cells (c) were negative controls. The cytoplasm and nuclei were stained red by Evans blue, and green fluorescent signals of viral antigen were significantly weaker in 3A10-3F cells (b) than in BHK-21 cells (d) (Magnification: ×400). To reckon the susceptibility of 3A10-3F to JEV, virus plaque formation assay and flow cytometry were applied. It appeared that JEV was able to replicate in the same kinetic mode in either 3A10-3F or BHK-21 cells at 0-48 h post-infection (Figure 3 ), but the numbers of infectious virions detected by plaque formation assay in 3A10-3F cells were remarkably declined. At moi of 1, the highest JEV titer at 48 h post-infection from 3A10-3F cells was decreased by 2 orders of logarithm compared with that of titer from BHK-21 cells (Figure 3a ). At moi of 10, the highest titer from 3A10-3F cells was lowered by 1 order of logarithm compared with that of from BHK-21 cells. Conspicuously there was a similar decline mode of virus replication in 3A10-3F cells compared with its parental cells between the two moi of JEV. Figure 3 JEV replication in 3A10-3F (black trianle) and BHK-21 (black square) cells measured by plaque formation assay . Cells were infected with JEV at moi of 1 (a) and 10 (b). The viral titers were determined by plaque formation assay for culture supernatant samples harvested at 0, 12, 24, 36, and 48 hr post-infection. The numbers of virions detected in 3A10-3F cells was greatly reduced. The results displayed were the means of independent experiments performed in duplicate. In flow cytometry, the binding between 3A10-3F cells and JEV was 2.10% (Figure 4a ), much lower than the binding between BHK-21 cells and JEV (48.84%, Figure 4b ), clearly indicating that 3A10-3F cells are defective of JEV binding because of the compromise of normal surface protein functioning for JEV attachment and/or entry. Besides, 3A10-3F cells appear morphologically similar (Figure 5a ) to normal BHK-21 cells (Figure 5c ), and still susceptible to HSV-1 (SM64 strain) infection as they manifested the similar CPEs (Figure 5b ) as rapidly as BHK-21 cells did (Figure 5d ). The reason to choose HSV-1 is that HSV binds to its surface protein molecules definitely different from that of JEV on BHK-21 cells, thus excluding the possibility that several mosquito-borne flaviviruses (such as JEV, Dengue virus, West Nile virus, etc) may attach the same protein molecules when virus-host interaction occurs [ 16 ]. Therefore, the HSV-1 infection test reinforced that resistance of 3A10-3F cells to JEV infection was specific for JEV (machinery for gene expression appeared intact), and was much less probably resulted from the decreased virus replication within the cells but from less virus binding to cell surface molecules necessary for JEV attachment and/or entry. Figure 4 JEV binding to 3A10-3F and BHK-21 cells measured by flow cytometry . Cells were incubated with JEV for 1 hr and were stained with rabbit anti-JEV antibodies followed by FITC-conjugated goat anti-rabbit IgG. Limited binding of JEV to 3A10-3F cells was shown (2.10%, a), whereas JEV significantly bound BHK-21 cells (48.84%, b). Figure 5 Cell morphology and CPEs of 3A10-3F and BHK-21 cells under a phase-contrast microscope . 3A10-3F cells (a) appeared morphologically similar to parental BHK-21 cells (c). After infected with HSV-1, 3A10-3F cells (b) appeared similar CPEs to normal BHK-21 cells infected with HSV-1(d) (Magnification: ×200). All data above implied that (i) JEV was still permissive to replicate within 3A10-3F cells but at a low level; (ii) JEV might enter 3A10-3F cells through different molecule routes, possibly high affinity and low affinity receptors [ 10 ]; and (iii) 3A10-3F cells became resistant to JEV infection most likely due to the altered surface protein expression after chemical mutagenesis rather than the mutations of certain intracellular factors essential for JEV replication. Finally, two-dimensional electrophoresis (2-DE) coupled with mass spectrometry was used to determine those differentially expressed proteins between 3A10-3F and BHK-21 cell membrane protein extracts following Mirza's method [ 17 ]. Altogether 23 spots of differentially expressed proteins on SDS-PAGE gels were picked out (Figure 6 ) upon computerized sifting criteria (the smooth parameter was set to a value of 2; the minimum area was above 8 pixels; the saliency parameter was experimentally adjusted to 1), including six up-regulated spots (11, 12, 14, 17, 19, 21) and six down-regulated spots (1, 4-7, 9) in 3A10-3F cells compared with BHK-21 cells, seven spots (13, 15, 16, 18, 20, 22, 23) solely expressed in 3A10-3F cells and four spots (2, 3, 8, 10) in BHK-21 cells respectively. These spots were subject to mass spectrometry analysis in combination with a mouse peptide database (NCBInr) searching, and the results indicated that 14 spots among the 23 samples (spots 1, 3-5, 7, 10-12, 14-16, 20, 22, and 23) were determined with good peptide coverage and significant scoring (Additional file 1 table S1). Not surprisingly, not all proteins extracted were membrane proteins, but all proteins with most altered abundance must have been disclosed. With such certitude, four membrane proteins from 3A10-3F cells recognized by mass spectrometry, annexin 1 and annexin 2, voltage-dependent anion channel (VDACs) 1 and 2, were of our particular interests. Figure 6 Representative 2-DE maps of 3A10-3F and BHK-21 cell membrane protein extracts . Membrane proteins from 3A10-3F (a) and BHK-21 cells (b) were separated by 2-DE and stained with sliver. 23 most differentially expressed protein spots were marked with arrows, and these spots were analyzed by LC-MS/MS. Three separate gels were prepared for each cell line. Annexins are a family of structurally related proteins whose common properties are to bind both phospholipids and cellular membranes in a calcium-dependent manner [ 18 ]. Annexin 1 involves in diverse cellular roles, including membrane fusion, exocytosis, differentiation, apoptosis, calcium channels, inflammatory reactions, and interaction with cytoskeletal proteins [ 19 ]. To date, few but one report [ 20 ] referred the role of annexin 1 in viral infection, stating that infectious pancreatic necrosis virus (IPNV) infection to fish cell increased the expression of salmon annexin 1, and the increased expression of salmon annexin 1 inhibited the apoptosis of IPNV-infected cells and supported the growth of IPNV in cells. In addition, annexin 2 had been identified as a receptor for cytomegalovirus (CMV) and respiratory syncytial virus (RSV) [ 21 , 22 ], and also promoted entry of HIV-1 into cells and proper assembly of HIV-1 in cells [ 23 , 24 ]. In our study, the expression of annexin 1 and annexin 2 were found to be significantly reduced on 3A10-3F cells, possibly suggesting that they may be involved in JEV attachment and/or entry. VDACs is a multigene family of evolutionarily conserved and well characterized porins found in outer mitochondrial membranes of all eukaryotes [ 25 ], where they control homeostasis by transport of ATP and ADP [ 26 ]. Numerous research groups also reported the presence of VDAC proteins in the plasma membrane of various cell types [ 27 , 28 ]. VDAC 1 contributed to ATP transport across the plasma membrane of murine cells [ 29 ], and could act as NADH-ferricyanide reductase to inhibit the release of synthesized ATP and greatly decrease the activity of exogenous NADH/cytochrome-c system of intact mitochondria [ 30 ], however the significance of its total absence from 3A10-3F on JEV infection is not reported. Mammalian VDAC 2 exhibited other more biological activities, such as interaction with Bcl-2 family proteins, critical regulators of apoptosis [ 31 ]. In our study, VDAC 2 should not be a candidate for JEV binding molecule as it displayed an increased expression (16-fold) on 3A10-3F cells. Nevertheless, some membrane molecules other than annexins and VDACs may not be excluded to mediate JEV entry into cells. It is not unusual that a virus particle utilizes multiple surface proteins during cell entry [ 32 - 34 ]. Several viruses utilize at least two different molecules to interact with their host cells via: (i) the binding receptors, merely serve as attachment factors that concentrate or recruit viruses on cell surface; (ii) co-receptors, that are used by the virus after binding to the cells, not only bind viruses but are also responsible for directing the bound viruses into endocytic pathways and for transmitting followed signals to the cytoplasm [ 35 ]. Taken together, the mutated cell line 3F10-3F from parental BHK-21 cells became less susceptible to JEV infection, and the discrepancy of membrane proteins between daughter and parental cells thus provides important clues to further investigate the individual role of the membrane proteins in JEV infection on BHK cells. List of abbreviations BHK: baby hamster kidney; CPE: cytopathic effect; FITC: fluorescein isothiocyanate; HSV: herpes simplex virus; IFA: indirect immunofluorescence assay; ICR-191: (6-chloro-9-[3-(2-chloroethylamino)propylamino]-2-methoxyacridine; JEV: Japanese encephalitis virus; LC-MS: liquid chromatography-mass spectrometry; moi: multiplicity of infection; VDAC: voltage-dependent anion channels protein Competing interests The authors declare that they have no competing interests. Authors' contributions TD and JR conceived the study and drafted the manuscript. TD carried out the chemical mutation on BHK-21 cells. WZ did RT-PCR. JR carried out IFA, plaque formation assay, flow cytometry, and 2-DE. TD and JR did bioinformatic analysis. WM and JR proofread the manuscript. All authors read and approved the final manuscript. Supplementary Material Additional file 1 Table 1: Comparison of LC-MS/MS recognized proteins between 3A10-3F and BHK-21 cells . A wider table describing recognized protein properties in MS Word .doc format. Click here for file Acknowledgements The authors would especially like to thank Dr. Yusong Ruan, State Key Laboratory of Biomembrane and Membrane Biotechnology, Institute of Zoology, Chinese Academy of Sciences, for technical support for 2-DE and LC-MS/MS analysis. This work is supported by Natural Science Foundation of China (30600526, 30470091).

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7428724/

Emerging Pandemic Diseases: How We Got to COVID-19

Infectious diseases prevalent in humans and animals are caused by pathogens that once emerged from other animal hosts. In addition to these established infections, new infectious diseases periodically emerge. In extreme cases they may cause pandemics such as COVID-19; in other cases, dead-end infections or smaller epidemics result. Established diseases may also re-emerge, for example by extending geographically or by becoming more transmissible or more pathogenic. Disease emergence reflects dynamic balances and imbalances, within complex globally distributed ecosystems comprising humans, animals, pathogens, and the environment. Understanding these variables is a necessary step in controlling future devastating disease emergences. Introduction Unimagined just a few short months ago, the ongoing COVID-19 pandemic has upended our entire planet, quickly challenging past assumptions and future certainties. It possesses simultaneously three characteristics that have allowed it to render an historic assault on the human species, triggering a virtual global "lockdown" as the only weapon against uncontrolled spread. It combines the characteristics of being a virus that to our knowledge has never before infected humans in a sustained manner, together with its extraordinary efficiency in transmitting from person to person and its relatively high level of morbidity and mortality, especially among seniors and those with underlying co-morbidities. It indeed is the perfect storm of an emerging infectious disease. Yet, pandemics such as COVID-19 are not entirely new phenomena. Newly emerging (and re-emerging ) infectious diseases have been threatening humans since the neolithic revolution, 12,000 years ago, when human hunter-gatherers settled into villages to domesticate animals and cultivate crops ( Dobson and Carper, 1996 ; Morens et al., 2020b ; Morens et al., 2008a ). These beginnings of domestication were the earliest steps in man's systematic, widespread manipulation of nature. Ancient emerging zoonotic diseases (see Box 1 ) with deadly consequences include smallpox, falciparum malaria, measles, and bubonic/pneumonic plague. Some, e.g., the Justinian plague (541 AD) and the Black Death (1348 AD), killed substantial proportions of humans in the "known" world, i.e., the world known to those whose recordings of it survive, predominantly in Asia, the Middle East, and Europe. Box 1 Terms Related to Emerging Infectious Diseases Antigenic immunodominance: Ability of a protein epitope to elicit an immune response greater than the response to one or more adjacent epitopes Cell tropism: Ability of a pathogen to infect a particular cell type Endemic: Noun and adjective denoting prevalence of human infection Enzootic: Noun and adjective denoting prevalence of animal infection Epidemic, Pandemic: Noun and adjective denoting highly incident disease (epidemic) or spread that is global or covers very large geographic areas (pandemic) Epizootic, Panzootic: Noun and adjective analogous to epidemic and pandemic, but with respect to animal diseases Fomite: An inanimate object that transmits infection, e.g., a towel or doorknob Host-Switching, Spillover: Process by which a pathogen adapted to one host species becomes adapted to another host species Disease emergence: Appearance of a disease in a new host Zoonosis: A human infection caused by an animal pathogen that may be either a dead-end infection or that may initiate person-to-person spread Only a century ago, the 1918 influenza pandemic killed 50 million or more people, apparently the deadliest event in recorded human history ( Morens and Taubenberger, 2020 ). The HIV/AIDS pandemic, recognized in 1981, has so far killed at least 37 million. And the past decade has witnessed unprecedented pandemic explosions: H1N1 "swine" influenza (2009), chikungunya (2014), and Zika (2015), as well as pandemic-like emergences of Ebola fever over large parts of Africa (2014 to the present). Since there are four endemic coronaviruses that circulate globally in humans, coronaviruses must have emerged and spread pandemically in the era prior to the recognition of viruses as human pathogens. The severe acute respiratory syndrome (SARS) coronavirus (SARS-CoV) emerged from an animal host, likely a civet cat, in 2002–2003, to cause a near-pandemic before disappearing in response to public health control measures. The related Middle East respiratory syndrome (MERS) coronavirus (MERS-CoV) emerged into humans from dromedary camels in 2012, but has since been transmitted inefficiently among humans ( Cui et al., 2019 ). COVID-19, recognized in late 2019, is but the latest example of an unexpected, novel, and devastating pandemic disease. One can conclude from this recent experience that we have entered a pandemic era ( Morens et al., 2020a ; Morens et al., 2020b ). The causes of this new and dangerous situation are multifaceted, complex, and deserving of serious examination. Infectious Diseases that Have Emerged in the Past In thinking about these recent infectious disease emergences, it is necessary to first consider currently existing infectious diseases that newly emerged in the past and then over time became endemic (prevalent in humans) or enzootic (prevalent in animals) ( Fauci and Morens, 2012 ; Morens et al., 2020b ; Morens and Fauci, 2012 ; Morens et al., 2004 ). Such existing diseases may provide important clues about the mechanisms of disease emergence and persistence and why thus far we have been largely unable to prevent and control many of them. The fact that many past emerging infectious microbes and viruses (hereafter grouped together as "microbes") have adapted to stable co-existence with humans is evidenced by the presence of endogenous retroviruses in human DNA ( Johnson, 2019 ) and by latently infecting herpesviruses such as herpes simplex (HSV), cytomegalovirus (CMV), Epstein-Barr virus (EBV), and varicella-zoster virus (VZV). VZV, for example, is a highly cytolytic, highly contagious, and potentially fatal virus that has adapted to long-term survival in human populations via a complex survival mechanism. Unlike other highly contagious human-adapted respiratory viruses such as measles—whose survival requires very large populations in order to avoid exhausting susceptible persons—VZV establishes latent non-cytolytic infections in human ganglia, periodically reactivating into an infectious/cytolytic form (zoster) that can be transmitted—even in populations without circulating varicella ("chicken pox")—to new birth cohorts of susceptible persons to be manifested as highly contagious varicella. Human infectious agents such as retroviruses, herpesviruses, and many others tell us that long-ago emergences of certain diseases can result in long-term microbial survival by co-opting certain of our genetic, cellular, and immune mechanisms to ensure their continuing transmission. In the terminology of British biologist Richard Dawkins, evolution occurs at the level of gene competition and we, phenotypic humans, are merely genetic "survival machines" in the competition between microbes and humans ( Dawkins, 1976 ). It may be a matter of perspective who is in the evolutionary driver's seat. This perspective has implications for how we think about and react to emerging infectious disease threats. From the human point of view, the fact that modern endemic diseases emerged and became established, at some unobserved time in the past ( Table 1 ) ( Morens et al., 2004 ), and that some of these diseases survived by adopting complicated long-term survival strategies, provides a compelling rationale for dual strategies for immediate and long-term control. First, in the immediate sense, it is important to mitigate spread of infection, illness, and death. Second, it is critical to prevent the persistence of microbes that may lead to additional emergences that are cumulatively as deadly, or more so, than the original emergences ( Dobson and Carper, 1996 ). That viral genetic descendants of the 1918 influenza pandemic virus are still causing seasonal outbreaks throughout the world, and still killing cumulatively millions of people a century later ( Morens and Taubenberger, 2020 ), is a powerful reminder that single disease emergences can have consequences beyond immediate morbidity and mortality. In the ancient ongoing struggle between microbes and man, genetically more adaptable microbes have the upper hand in consistently surprising us and often catching us unprepared. Table 1 Emerging Infectious Diseases in History Year Name Deaths Comments 430 BCE "Plague of Athens" ∼100,000 First identified trans-regional pandemic 541 Justinian plague ( Yersinia pestis ) 30–50 million Pandemic; killed half of world population 1340s "Black Death" ( Yersinia pestis ) ∼50 million Pandemic; killed at least a quarter of world population 1494 Syphilis ( Treponema pallidum ) >50,000 Pandemic brought to Europe from the Americas c. 1500 Tuberculosis High millions Ancient disease; became pandemic in Middle Ages 1520 Hueyzahuatl ( Variola major ) 3.5 million Pandemic brought to New World by Europeans 1793–1798 "The American plague" ∼25,000 Yellow fever terrorized colonial America 1832 2nd cholera pandemic (Paris) 18,402 Spread from India to Europe/Western Hemisphere 1918 "Spanish" influenza ∼50 million Led to additional pandemics in 1957, 1968, 2009 1976–2020 Ebola 15,258 First recognized in 1976; 29 regional epidemics to 2020 1981 Acute hemorrhagic conjunctivitis rare deaths First recognized in 1969; pandemic in 1981 1981 HIV/AIDS ∼37 million First recognized 1981; ongoing pandemic 2002 SARS 813 Near-pandemic 2009 H1N1 "swine flu" 284,000 5th influenza pandemic of century 2014 Chikungunya uncommon Pandemic, mosquito-borne 2015 Zika ∼1,000? ∗ Pandemic, mosquito-borne Selected important emerging and re-emerging infectious diseases of the past and present, 430 BCE–2020 CE. Mortality estimates are in most cases imprecise; see text. ∗ Zika mortality has not been fully established. Most deaths are fetal or related to outcomes of severe congenital infections. The latest example of this, the COVID-19 pandemic, which emerged in December 2019, is still exploding globally ( Figure 1 ). At time of writing, over 22 million cases have been detected, with over 800,000 deaths recorded ( World Health Organization ); however, these are undoubtedly significant undercounts, reflecting early and still problematic access to diagnostic testing coupled with incomplete diagnoses of fatal cases. As COVID-19 is caused by a novel virus (SARS-CoV-2) producing a spectrum of disease whose clinical, pathologic, and epidemiologic patterns have never before been observed, we are gaining insights only incrementally. At some time in the future we will be better able to compare and contrast COVID-19 to other important emerging diseases; however, at this time we are still just entering a steep learning curve that will surely keep surprising us as we struggle to control what is already among the deadliest pandemics of the past century. Figure 1 Global Daily Incident Cases of COVID-19 by World Health Organization Region as of August 18, 2020 The data ( World Health Organization ) show that beginning in March 2020, the pandemic exploded in Europe and the Americas, particularly in the United States, was blunted in these two regions between March and May 2020, and then began to explode anew in the Americas and to a lesser extent in Europe beginning in late May. Since May 2020, the pandemic has been increasing significantly in the SEARO as well as the AFRO regions. WRPO, Western Pacific; AFRO, Africa; EMRO, Eastern Mediterranean; SEARO, Southeast Asia; EURO, Europe; AMRO, Americas. Definitions of Emerging Infectious Diseases The once-emerging/now prevalent diseases mentioned above, e.g., many viruses causing upper respiratory, enteric, or dermal/mucosal infections, are not considered to be truly emerging even when they vary seasonally or geographically; however, upon this background of existing diseases, new diseases still continue to emerge. Emerging diseases have been categorized as newly emerging, re-emerging, or "deliberately emerging," that is, associated with bioterrorism ( Table 2 ; Figure 2 ) ( Morens and Fauci, 2012 ; Morens et al., 2004 , 2008a ). To these we add "accidentally emerging" human-generated diseases, such as repeated emergences of vaccine-derived polioviruses (VDPVs) resulting from naturally occurring back-mutations of live virus vaccines, as well as a live human-engineered vaccine that escaped to cause a new epizootic disease: naturally transmitted vaccinia ( Lum et al., 1967 ). Although these four categories are distinct, they are also interrelated: newly emerging diseases can persist and then re-emerge and can also become agents of deliberate or accidental release. An example crossing the latter two categories is the 1979 Sverdlovsk (now Yekaterinaberg) anthrax accident, in which an unintentional explosion in a Russian bioweapons factory released anthrax into the air, resulting in at least 100 human deaths ( Meselson et al., 1994 ). Such deliberately emerging and accidentally emerging diseases represent a special case in which emergence/epidemicity is best prevented by global biosafety cooperation. Table 2 Major Categories of Emerging Infectious Diseases Newly emerging infectious diseases Diseases recognized in humans for the first time, e.g., HIV/AIDS (1981), Nipah virus (1999), SARS (2002), MERS (2012), COVID-19 (2019) Re-emerging infectious diseases Diseases that have historically infected humans but continue to re-appear either in new locations (e.g., West Nile in the United States and Russia in 1999) or in resistant forms (e.g., methicillin-resistant Staphylococcus aureus ) Deliberately emerging infectious diseases Diseases associated with intent to harm, including mass bioterrorism Accidentally emerging infectious diseases Diseases created by humans that are released unintentionally, e.g., epizootic vaccinia and transmissible vaccine-derived polioviruses Not included are currently established endemic diseases that are presumed to have been newly emerging at some time in the past and then went on to develop long-term persistence in human or animal populations (see text). Figure 2 Recent Emerging Infectious Diseases The global extent of newly emerging, re-emerging, and "deliberately emerging" infectious disease from 1981 to the present (2020). Among possible ways to achieve such cooperation are by strengthening the United Nations and its agencies, particularly the World Health Organization and the Office International des Épizooties (OIE; World Organisation for Animal Health); by supporting collaborative multinational research in prevention of disease emergence; by studying high-consequence pathogens under appropriate safety and containment conditions; by demanding renewed international intergovernmental efforts at the global level to collaborate on research related to the risks of global pathogen emergence risks and how to prevent them; and by preventing bioweapons development. A fifth related category, significant because it emphasizes the importance of responding to disease emergence threats with countermeasures, is that of diseases that are "de-emerging," i.e., those that have been eliminated or even eradicated or that are in the process of elimination and/or eradication ( Table 3 ) ( Dowdle and Cochi, 2011 ; Hopkins, 2013 ; Tomori, 2011 ). Smallpox and the veterinary disease rinderpest were declared eradicated in 1980 and 2011, respectively. SARS, which emerged in 2002–2003 and spread globally to 29 countries, infecting 8,096 people and killing 813, was controlled and ultimately eliminated from human spread by effective public health efforts ( Cui et al., 2019 ). By some definitions, SARS was thereby eradicated, although it presumably remains in enzootic circulation and could re-emerge from nature, as Ebola viruses have been doing for the past 44 years ( Baseler et al., 2017 ). Table 3 Selected Human Infectious Diseases Variables that relate to their potential for eradication. (Top) Selected infectious diseases that have either been eradicated, are now being targeted for eradication, or are being significantly controlled by public health and medical or veterinary actions. (Bottom) Human infectious diseases that are currently considered non-eradicable but for which some important disease aspects could potentially be eliminated with existing tools (e.g., eliminating human rabies without or before eradicating rabies in wild animals). The information is based on published data reflecting 2008 determinations ( https://www.cartercenter.org ), supplemented by additional widely available publications. Columns show disease features that favor eradicability (blue circles), are of uncertain relevance to eradicability (yellow), or are expected to make eradication more difficult or impossible (red). Some of these features (e.g., ease of detecting disease and immunity, seasonality in tropical versus temperate climates) are subjective and situationally variable; see text and references for the individual diseases. ∗ Sterile immunity for life refers to the ability of a natural infection or vaccine to induce a type and degree of immunity that prevents infection/reinfection and eliminates carriage and transmissibility to others and to animal and environmental reservoirs. Other diseases nearing eradication include dracunculiasis, lymphatic filariasis, measles, polio, and rubella ( Table 3 ). Such successes in eradication/control reflect the availability of improved tools and strategies for prevention and control, as well as international public and private efforts to reduce their substantial mortality and morbidity. Successes in eradication and control of infectious diseases remind us that we are not helpless in the face of emerging diseases. Eradicating/controlling existing diseases and preventing/controlling newly emerging diseases are related efforts demanding the same scientific, public health, and civic/political focus that will be required to successfully address this formidable challenge. Variables in Disease Emergence: The Agent, Host, and Environment Microbes that cause human diseases by definition have existed in some other environmental niche before emerging to infect humans and other animals. While some such organisms have long been human pathogens that mutated into new forms—e.g., re-emergences of antibiotic-resistant bacteria such as methicillin-resistant Staphylococcus aureus (MRSA)—most are, and historically have been, zoonotic ( Woolhouse and Gowtage-Sequeria, 2005 ; Woolhouse et al., 2005 ). Such zoonotic microbial emergences are often associated with mutational mechanisms allowing host-switching from animals to humans, as discussed below. The triad of causations of emerging and other diseases, as conceptualized for over a century, represents interactions between infectious agents, their hosts, and the environment ( Figure 3 ). This conceptualization acknowledges the reality that, while infectious diseases themselves are necessarily "caused" by microbial agents, emergences that produce epidemics and pandemics are also significantly determined by co-factors related to the host and to host-environmental interactions ( Centers for Disease Control and Prevention, 2011 ). Figure 3 Infectious Agents, Hosts, and the Environment: Determinants of Disease Emergence and Persistence Diseases, including emerging diseases, result from interactions between infectious agents, hosts, and the environment. Adapted from Fauci and Morens (2012) ; Morens and Fauci (2012) ; Morens et al. (2004) , (2008a) . The Role of the Infectious Agent in the Emergence of Infectious Diseases Considerations of the emergences of infectious diseases begin with the infectious agent itself. Although many established diseases, such as tuberculosis, malaria, and cholera, are bacterial or protozoal, and yet others are caused by fungi (e.g ., cryptococci) or agents such as Rickettsia or prions, the majority of important newly emerging and re-emerging diseases in the past century have been viral. This review therefore emphasizes viruses, including SARS-CoV, and SARS-CoV-2, influenza, arboviruses, and hemorrhagic fever viruses among others. Genetic instability of microorganisms is an inherent property allowing rapid microbial evolution to adapt to ever-changing ecologic niches. This is particularly true of RNA viruses such as influenza viruses, flaviviruses, enteroviruses, and coronaviruses, which have inherently deficient or absent polymerase error-correction mechanisms and are transmitted as quasispecies or swarms of many, often hundreds or thousands of, genetic variants. Emergences of viral diseases begin with the genetic plasticity of the infectious agent, which may repeatedly encounter ecologic niches into which it can evolve and adapt under facilitative circumstances, e.g., those provided by the hosts in the context of the host environment. For viruses transmitted by person-to-person mechanisms, transmission by quasispecies may increase the likelihood that one or more viral variants within the quasispecies will be infectious for cells of a new host, leading to infection, viral amplification, and expansion of a new and different quasispecies, facilitating onward transmission (see below). Other determinants of emerging infectious agents include cell tropism, ability to circumvent innate immune responses, and antigenic immunodominance, among others. Many viruses enter cells via one or more cell receptors ( Figure 4 ) ( Dai et al., 2020 ; Jayawardena et al., 2020 ); some infect different cells via different receptors, while some cell receptors may be entry points for multiple different types of viruses. The situation is extraordinarily complex from the point of view of both virus and host, with a bewildering array of receptors, alternate receptors, and co-receptors and of countless viruses able to utilize them, reflecting that "viruses have deep evolutionary roots in the cellular world" ( Baranowski et al., 2001 ). This is exemplified by the SARS-like bat β-coronaviruses, or sarbecoviruses, whose receptor binding domains appear to be hyper-evolving by sampling a variety of mammalian receptors ( Hu et al., 2017 ). Figure 4 Variable Mechanisms of Viral Entry into Host Cells Examples of cell receptors for various DNA and RNA viruses. The cartoon image shows a spherical cell with different receptors for different categories of selected viruses. Viruses and cells are not reflective of relative sizes. The figure is suggested by the text and images of Jayawardena et al. (2020) . Many viruses enter cells via phagocytic or pinocytic endocytosis, the latter including clathrin-mediated or claveolin-mediated endocytosis, yet other viruses enter cells via fusion or direct penetration ( Dai et al., 2020 ). SARS-CoV and SARS-CoV-2 are β-coronaviruses that enter human cells via angiotensin-converting enzyme-2 (ACE-2) receptors, whose non-human counterpart receptors are ubiquitous on cells of other species ( Cui et al., 2019 ; Hasan et al., 2020 ). This means that coronaviruses of many other mammalian species may essentially be pre-adapted to human infectivity. Evidence suggests that there are many bat coronaviruses pre-adapted to emerge, and possibly to emerge pandemically ( Andersen et al., 2020 ; Hasan et al., 2020 ; Hu et al., 2017 ; Menachery et al., 2016 ; Wang et al., 2018 ; Zhou et al., 2020a ). Influenza A viruses (IAVs) infect cells via binding to terminal sialic acids found on lumenal respiratory epithelial cells of avian as well as mammalian and human hosts and contain a neuraminidase that cleaves these same receptors to allow viral release, which facilitates onward viral transmission ( Morens and Taubenberger, 2020 ; Taubenberger et al., 2019 ). Since sialic acids are found on a wide variety of mammalian and non-mammalian cells, it is not surprising that many viruses attach to these receptors, although viral affinities to receptors are complicated. For example, in vitro lectin assays suggest variable affinities of influenza for different types of terminal sialic acids, e.g., those with α-2,3 (ostensibly avian) and those with α-2,6 (ostensibly mammalian) linkages. However, autopsy studies have confirmed fatal human infections caused by IAVs that prefer either receptor, reflecting the complexity of virus-host interactions across the animal kingdom. Yet another aspect of infecting viruses is that some (e.g., HIV, influenza) express immunodominant epitopes on external proteins that dominate the elicitation of immune responses, resulting in less robust immune responses to other, often adjacent, epitopes. This may have the effect of blunting, or limiting the breadth of, an otherwise optimal host immune response. Almost all viruses have proteins that inhibit innate protective host responses, such as the host interferon response ( Blanco-Melo et al., 2020 ; Mesev et al., 2019 ). Some viruses are able to infect FcR bearing cells via antibody-dependent infection enhancement or ADE—a mechanism by which virus-IgG complexes are "tied down" by FcRs to the cell surface, facilitating cell entry via another receptor ( Morens, 1994 ; Sullivan, 2001 ). During the 2002–2003 SARS epidemic, it was found that both post-infectious and vaccine-elicited antibody caused in vitro ADE with SARS-CoV ( Jaume et al., 2012 ; Wang et al., 2016 , 2014 ; Yip et al., 2014 ). It is not known whether this phenomenon, studied mostly in vitro , has implications for natural human coronavirus infections or vaccinations ( Wan et al., 2020 ); however, it does represent a potential safety concern associated with vaccine development for SARS-CoV and SARS-CoV-2. In considering SARS-CoV-2 vaccine development and antibody treatment/prophylaxis, it is also of some concern that natural infection with the feline coronavirus (FECV) initiates a non-fatal infection that sometimes leads to development of viral variants (known as feline infectious peritonitis viruses [FIPVs]) that are macrophage-tropic and that can bind to host anti-spike protein antibody (antibody against the external viral protein that attaches to ACE-2 receptors and elicits protective immunity) to allow viruses to enter macrophages via FcRs (ADE), leading to a distinct and universally fatal disease known as feline infectious peritonitis, or FIP ( Vennema et al., 1990 ; Weiss and Scott, 1981 ). Post-infectious gene editing by the host has also been proposed as a mechanism for development of subacute sclerosing pan-encephalitis (SSPE), a fatal human disease associated with persistent natural measles virus infection complicated by generation of host-edited mutant viruses ( Cattaneo et al., 1986 ). This represents yet another variation on mechanisms of viral emergence ( Baranowski et al., 2001 ; Cattaneo et al., 1986 ; Novella et al., 2011 ). Fortunately, viruses such as those that cause SSPE, derived from in-host gene editing, are not necessarily transmissible. In experimental studies, early FIPV-associated feline deaths result from both FIPV spike protein vaccination and passive transfusion with anti-FIPV antibody ( Vennema et al., 1990 ; Weiss and Scott, 1981 ). Neither FECV nor FIPV are phylogenetically close to SARS-CoV or SARS-CoV-2, both being α-coronaviruses utilizing aminopeptidase N or other protein or glycan receptors rather than the ACE-2 receptors that bind SARS β-coronaviruses. But evidence for ADE with multiple different α- and β-coronaviruses suggests that as we proceed to develop SARS-CoV-2 vaccines and therapeutic antibodies, much remains to be learned about this complex viral family. Also of importance to the infectivity of newly emerging infectious diseases are viral genetic properties associated with pathogenicity and co-pathogenicity , exemplified most clearly with pandemic IAVs. The 1918 H1N1 pandemic virus, which killed an estimated 50 million people (equivalent to 200 million when adjusted to the 2020 population) was particularly lethal because of at least two inherent properties: (1) an avian-descended H1 hemagglutinin (HA) that is unusually cytopathic and immunopathogenic compared to the HAs of most other IAVs and (2) a marked co-pathogenic ability—the viral genetic basis of which remains poorly understood—to precipitate fatal bacterial bronchopneumonias in association with pneumopathogenic bacteria carried silently in the human upper respiratory tract ( Morens et al., 2008b ; Morens and Taubenberger, 2020 ; Taubenberger et al., 2019 ). We now know that not only avian H1s but also 4 of the other 15 avian HAs found within the wild waterfowl and shore birds ( Anseriformes and Charidriiformes ) reservoir have similar pathogenic properties, and thus they represent future threats for highly fatal pandemic emergences ( Morens and Taubenberger, 2020 ; Taubenberger et al., 2019 ). A perhaps even more shocking example of pandemic emergence associated with enhanced pathogenicity is that of Zika, a flavivirus known for decades. Zika had never caused a human epidemic; however, in 2015, it suddenly spread pandemically around the global tropical belt, causing millions of infections and severe fetal losses and birth defects ( Fauci and Morens, 2016 ). The apparent cause of the pandemic was a mutation resulting in the change of a single amino acid in the external viral glycoprotein ( Shan et al., 2020 ). Pathogenic variability undoubtedly applies to many other virus types, exemplified, for example, by comparing the highly pathogenic Ebola Zaire strain to the closely related but lowly pathogenic Ebola Reston strain ( Baseler et al., 2017 ). Although not yet adequately studied, pathogenic variability might also be a property of animal coronaviruses. That all three recently emerging human coronaviruses (the agents of SARS, MERS, and COVID-19) exact a high degree of human morbidity and mortality suggests that enzootic coronaviruses as a whole may be inherently pathogenic for humans. On the one hand, preliminary data suggest that SARS-CoV-2 may elicit an unbalanced innate immune response associated with decreased expression of interferons I and III and with increased inflammatory cytokine production ( Mesev et al., 2019 ), consistent with preliminary COVID-19 findings ( Vanderheiden et al., 2020 ). On the other hand, preliminary data suggest that viral-bacterial co-pathogenesis may be of somewhat less concern with SARS-CoV-2 than with influenza, measles, and other pathogenic respiratory viruses, although comprehensive clinical/autopsy series have not yet been published. With regard to coronavirus disease severity, it is worth considering the conventional wisdom (not always correct) that viruses that kill their hosts limit their own ability to be transmitted and that, if transmitted person to person, they would be expected to be selected for attenuation of pathogenicity over time. It is conceivable that the four endemic coronaviruses of humans—the β-coronaviruses OC43 and HKU1 and the α-coronaviruses 229E and NL63—emerged long ago as zoonotic and perhaps highly pathogenic viruses that evolved into attenuated forms over time ( Cui et al., 2019 ). Such natural attenuation of pathogenicity has not yet been observed with the agents of SARS, MERS, or COVID-19, although with a SARS-CoV-2 case-fatality in the range of 1%, and with evidence for significant asymptomatic and presymptomatic transmission in a largely susceptible population, evidence for selection pressures for attenuation may not be detectable in the short term. Moreover, viral evolution toward lower pathogenicity does not apply to all infectious diseases. For many other organisms such as cholera (expressing a bacteriophage toxin causing diarrhea) or rotavirus infection (with an NSP4 diarrhea-causing toxin), or for cough-inducing tuberculosis and many respiratory viruses, damage to the gastrointestinal tract and lungs, respectively, facilitates transmission, since diarrhea and coughing expel infectious pathogens into the environment, increasing the chance of infecting additional hosts. Disease severity thus reflects a balancing act between killing or incapacitating hosts, on the one hand, and optimizing microbial transmission, and therefore survival, on the other. Similar principles may also apply to non-viral diseases with environmental modes of transmission: for example, with enzootic anthrax, host-killing may be an important transmission mechanism, as rotting carcasses leave anthrax spores in the ground to reignite future infections ( Turner et al., 2014 ). Agent-host interaction variables are often exceedingly complex. The Role of the Host in the Emergence of Infectious Diseases Host variables that underlie the emergence of infectious diseases include those variables specific to individuals within the host population and those variables that relate to the host population as a whole ( Morens and Fauci, 2012 ; Morens et al., 2004 , 2008a ). Since a virus replicates within the cells of the host, and since viruses usually infect hosts via specific receptors on the cells of various tissues and organs, the new host must express cellular receptors or other cell-surface properties to which the virus can bind and initiate viral internalization. Major portals of host entry for infectious agents include those that are visibly external to the environment such as the skin or that can be reached directly from the environment such as the respiratory and gastrointestinal tracts, as well as organs reached systemically such as the liver, heart, and other internal organs. Human beings have many different organ systems, each with many different cell types, and with each cell having arrays of different receptors; therefore, it is not surprising that switching of a pathogen from an animal host to humans results in very different clinical and epidemiologic outcomes, including different disease manifestations and transmission mechanisms. These factors ultimately relate to the potential for establishment of infection in the new host as well as the likelihood of sustained transmission within the new host population and, as such, have a bearing on whether host-switching succeeds or fails. SARS-CoV and SARS-CoV-2 enter cells via ACE-2 receptors ( Wang et al., 2020 ), found on lung alveolar epithelial cells, gastrointestinal enterocytes, arterial and venous endothelial cells, and arterial smooth muscle cells, among other cell types ( Hamming et al., 2004 ; Wang et al., 2020 ), which explains the excretion of SARS-CoV-2 and potential transmission via the respiratory and enteric routes. With regard to the latter, although SARS-CoV-2 infects cells of the gastrointestinal tract, fecal transmission has not to date been implicated in significant person-to-person viral spread. Different viruses utilize widely different strategies for binding to, penetrating, and entering cells, e.g., polioviruses, HIV, influenza viruses, coronaviruses, and others ( Figure 4 ) ( Bowers et al., 2017 ; Cicala et al., 2011 ; Jayawardena et al., 2020 ; Laureti et al., 2018 ). Some viruses enter cells via binding to two different proximate receptors, a primary and secondary receptor, e.g., receptors for binding and for fusion (for example, numerous flaviviruses [ Laureti et al., 2018 ]). As noted, mechanisms of viral entry into cells are exceedingly variable and complex ( Jayawardena et al., 2020 ). Viruses may also infect macrophages and macrophage-like cells, as is the case with mosquito-borne and tick-borne flaviviruses. These viruses are injected into perivascular dermal tissue by their respective vectors and are taken up by dendritic cells and carried to regional lymph nodes, where they initiate systemic infection. ADE, discussed above, and other ADE-like phenomena, may also facilitate cellular infection ( Morens, 1994 ; Sullivan, 2001 ). Tissue/cell tropism also has a bearing on the types of immune responses that are elicited. For example, in a systemic infection like measles, high-level viremia is associated with infection of multiple organs, tissues, and cell types; the resulting broad systemic B and T cell responses lead to lifelong protection from reinfection. In contrast, influenza A viruses and respiratory syncytial virus (RSV), among many other respiratory viruses that infect surface epithelial cells, do not cause viremia and infect only surface respiratory epithelial cells. As a result, infectious virions do not have intimate interactions with the systemic immune system. The major site of influenza virus-immune system interaction is in the semi-organized tear duct-, nasal-, and mucosal-associated lymphoid tissues (TALT, NALT, and MALT), as well as the post-natally generated inducible bronchus-associated iBALT ( Moyron-Quiroz et al., 2007 ), leading to tissue compartmentalization of the immune response, perhaps in part explaining the weakly protective immune responses of naturally acquired or vaccine-associated influenza and RSV. Preliminary evidence from clinical and pathological studies of both SARS-CoV and SARS-CoV-2, which indicate viral infection of multiple tissues, is consistent with elicitation of robust and hopefully long-lasting protective immunity, providing a potential for control of COVID-19 with vaccines. More ominously, expression of ACE-2 receptors on endothelial and numerous other cells, and autopsy evidence of significant SARS-C0V-2 endothelial infection ( Fox et al., 2020 ), are consistent with systemic viral infection causing both pulmonary and extra-pulmonary pathology, including widespread microthrombus formation, among other outcomes. Some emerging viruses encounter pre-existing partial population immunity , e.g., pandemic influenza viruses. Most notably, in the influenza pandemics of 1968 and 2009, caused by an H3N2 and an H1N1 virus, respectively, segments of the population had pre-existing immunity that interfered with early viral spread and possibly with viral evolution. Although insufficient to prevent emergence, such population immunity did protect certain segments of the population ( Morens and Taubenberger, 2020 ; Taubenberger et al., 2019 ). In fact, in most influenza pandemics, influenza-specific case-fatality in the elderly, which increases regularly over about age 60 with seasonal influenza, is nevertheless apparently blunted by poorly characterized immune effects of prior influenza exposures, emphasizing the complexity of viral infectivity and host resistance factors ( Morens and Taubenberger, 2011 ). In contrast, there are suggestions that decreases in natural infections with, or vaccinations against, pathogens may facilitate the emergence of related organisms, e.g., increased incidence of human monkeypox after the cessation of smallpox vaccination that followed eradication of the monkeypox-related smallpox virus ( Lloyd-Smith, 2013 ), or the long-held theory that existing or newly evolving enteroviruses will emerge, or are already emerging, to fill an "ecologic niche" created by the near-eradication of the three polioviruses ( Rieder et al., 2001 ). In the case of COVID-19, some evidence suggests the absence of pre-existing population immunity, afforded by exposure to the four endemic coronaviruses, sufficient to prevent infection ( Corman et al., 2018 ). Although these endemic viruses share few epitopes that cross-react significantly with SARS-CoV-2 in serologic studies, it has been speculated that endemic coronavirus cross-protection may nevertheless prevent or at least limit the severity of disease in some, especially in young persons ( Nickbakhsh et al., 2020 ). There is preliminary but growing evidence that infectious disease severity or even susceptibility may in some cases be related to host genetic variables associated with the innate immune response, as is the case with epidemiologic information concerning severe disease caused by H5N1 poultry-associated influenza ( Morens and Taubenberger, 2015 ). Several host genes have been provisionally linked to susceptibility to such severe viral disease outcomes ( Nguyen et al., 2020 ; Tang et al., 2008 ), including the interferon-related transmembrane protein 3 (IFITM3) SNP rs22522-C allele ( Everitt et al., 2012 ; Zheng et al., 2017 ), which has been proposed (based on very preliminary data) to be involved in severity of SARS-CoV-2 disease ( Zhang et al., 2020 ). In addition, differential interactions with IFITMs such as IFITM3 include inhibiting the human endemic α-coronavirus while enhancing entry of SARS-CoV and MERS-CoV β-coronaviruses ( Huang et al., 2011 ; Zhao et al., 2018 ). With respect to COVID-19, several studies have associated blood group A in the ABO system with disease severity, although the mechanisms of this effect are not yet clear. ABO system associations with infectious diseases have also been shown for infections with noroviruses, H. pylori , and falciparum malaria; however, any such associations might also be indirect markers for unrelated genes. Newer data are beginning to define human immonotype risks for more severe disease ( Mathew et al., 2020 ), as well as posssible genetic signatures of severe diseases ( Gussow et al., 2020 ). At this time, data establishing specific genetic susceptibilities remain inconclusive for most diseases, including coronavirus diseases. This is a research area likely to be important in the future, since identification of susceptibilities for human disease severity has great implications for prevention, diagnosis, and treatment. Among the most important host factors for infection and for disease emergences/re-emergences are those associated with human behaviors , e.g., population growth, crowding, human movement, and many others, including behaviors that either perturb the environment or result in new human-created ecologic niches ( Figure 3 ). Regarding human movement, both the 1347–1348 Black Death (bubonic/pneumonic plague) and the 1832 cholera pandemic (which traveled from India to Europe and then to the Western Hemisphere) were spread along major trade and travel routes. In 1831–1832, 45 years before a coherent "germ theory" would be articulated, it was clear that as cholera spread slowly westward, it moved no faster than coaches and ships traveled: it slowed down in the winter as travel slowed down, and it picked up again in the summer as travel increased. The 1889 influenza pandemic traveled westward from Asia to Europe along railroad lines and then was exported globally along shipping routes. The 1957 influenza pandemic was spread by ships, but 11 years later the 1968 influenza pandemic was spread along air routes, the first example of global pandemic spread by airplanes. In 1981, acute hemorrhagic conjunctivitis was spread between international air hubs in the tropics and some temperate zones (e.g., to Florida and North Carolina). In 2002–2003, SARS was exported by air from Hong Kong to the Western Hemisphere and Europe. In 2019–2020, SARS-CoV-2 was spread globally from China in a similar manner. These many ancient and modern examples reflect the extraordinary importance of human population growth and movement in spreading diseases: the more populous and crowded we as a species become, and the more we travel, the more we provide opportunities for emerging diseases. The Role of the Environment in the Emergence of Infectious Diseases Many other human activities related to the environment have important consequences for disease emergence ( Allen et al., 2017 ; Dobson and Carper, 1996 ; Fauci and Morens, 2012 ; Morens et al., 2020b ; Morens and Fauci, 2012 ; Morens et al., 2004 , 2008a , 2019 ; Morens and Taubenberger, 2020 ). Human water storage practices in Northern Africa, beginning about 5,000 years ago, led to the emergence of a new, exclusively human-adapted mosquito, Aedes aegypti , which created a secondary ecologic niche for the emergence of yellow fever virus and, centuries later, dengue, chikungunya, and Zika, all of which then went on to spread pandemically. Depots of used rubber tires create ideal breeding sites for a related human-adapted mosquito, Aedes albopictus , which in recent decades has spread globally across the tropical and subtropical zones, transmitting many of these same arboviral diseases widely, if less efficiently. Predictably, viruses and viral vectors have adapted to environmental influences: a single locus mutation in the Aedes aegypti -adapted Indian Ocean strain of chikungunya virus has newly adapted it, without loss of fitness for aegypti , to Aedes albopictus, widely prevalent in the region ( Tsetsarkin and Weaver, 2011 ). Land-management practices have been associated with re-emergences of Eastern equine encephalitis ( Morens et al., 2019 ); deforestation with emergences of Zika and Hendra viruses; road-building and environmental degradation with the spread of Bolivian hemorrhagic fever and HIV (infections spread by truckers and truck stop prostitution); and poverty, crowding, and poor sanitation with re-emergences of many diseases such as tuberculosis and cholera. For centuries, wars have precipitated the re-emergences of many diseases ( Dobson and Carper, 1996 ; Fauci and Morens, 2012 ; Morens and Fauci, 2012 ; Morens et al., 2004 , 2008a , 2020b ; Morens and Taubenberger, 2020 ), e.g., the Serbian typhus epidemic during World War I, which killed 150,000 people, mostly civilians. In the 1700s, when typhus had not been identified as a specific disease, textbooks listed two separate conditions under the nosologic terms "war typhus" and "jail typhus," reflecting human activities that provoked long-ago disease emergences. A classic epidemiology text published over a century ago, still studied today, is titled Epidemics Resulting from Wars ( Prinzing, 1916 ). It is suspected that SARS-CoV-2 emerged in 2019, as did SARS-CoV in 2002, and as did H5N1 and H7N9 poultry-associated influenza, in 1997 and 2013, respectively, from wet markets in China ( Morens et al., 2020a ; Morens et al., 2020b ). These four diseases may thus represent four deadly emergences, within an 18-year span, from one cultural practice in one region of the world. These and many other examples ( Dobson and Carper, 1996 ; Fauci and Morens, 2012 ; Lu et al., 2020 ; Morens and Fauci, 2012 ; Morens et al., 2004 , 2008a , 2019 , 2020b ; Morens and Taubenberger, 2020 ; Zhang and Holmes, 2020 ; Zhou et al., 2020b ) constitute a powerful argument that human activities and practices have become the key determinant of disease emergence. The Role of the Infectious Agent in the Emergence of Infectious Diseases Considerations of the emergences of infectious diseases begin with the infectious agent itself. Although many established diseases, such as tuberculosis, malaria, and cholera, are bacterial or protozoal, and yet others are caused by fungi (e.g ., cryptococci) or agents such as Rickettsia or prions, the majority of important newly emerging and re-emerging diseases in the past century have been viral. This review therefore emphasizes viruses, including SARS-CoV, and SARS-CoV-2, influenza, arboviruses, and hemorrhagic fever viruses among others. Genetic instability of microorganisms is an inherent property allowing rapid microbial evolution to adapt to ever-changing ecologic niches. This is particularly true of RNA viruses such as influenza viruses, flaviviruses, enteroviruses, and coronaviruses, which have inherently deficient or absent polymerase error-correction mechanisms and are transmitted as quasispecies or swarms of many, often hundreds or thousands of, genetic variants. Emergences of viral diseases begin with the genetic plasticity of the infectious agent, which may repeatedly encounter ecologic niches into which it can evolve and adapt under facilitative circumstances, e.g., those provided by the hosts in the context of the host environment. For viruses transmitted by person-to-person mechanisms, transmission by quasispecies may increase the likelihood that one or more viral variants within the quasispecies will be infectious for cells of a new host, leading to infection, viral amplification, and expansion of a new and different quasispecies, facilitating onward transmission (see below). Other determinants of emerging infectious agents include cell tropism, ability to circumvent innate immune responses, and antigenic immunodominance, among others. Many viruses enter cells via one or more cell receptors ( Figure 4 ) ( Dai et al., 2020 ; Jayawardena et al., 2020 ); some infect different cells via different receptors, while some cell receptors may be entry points for multiple different types of viruses. The situation is extraordinarily complex from the point of view of both virus and host, with a bewildering array of receptors, alternate receptors, and co-receptors and of countless viruses able to utilize them, reflecting that "viruses have deep evolutionary roots in the cellular world" ( Baranowski et al., 2001 ). This is exemplified by the SARS-like bat β-coronaviruses, or sarbecoviruses, whose receptor binding domains appear to be hyper-evolving by sampling a variety of mammalian receptors ( Hu et al., 2017 ). Figure 4 Variable Mechanisms of Viral Entry into Host Cells Examples of cell receptors for various DNA and RNA viruses. The cartoon image shows a spherical cell with different receptors for different categories of selected viruses. Viruses and cells are not reflective of relative sizes. The figure is suggested by the text and images of Jayawardena et al. (2020) . Many viruses enter cells via phagocytic or pinocytic endocytosis, the latter including clathrin-mediated or claveolin-mediated endocytosis, yet other viruses enter cells via fusion or direct penetration ( Dai et al., 2020 ). SARS-CoV and SARS-CoV-2 are β-coronaviruses that enter human cells via angiotensin-converting enzyme-2 (ACE-2) receptors, whose non-human counterpart receptors are ubiquitous on cells of other species ( Cui et al., 2019 ; Hasan et al., 2020 ). This means that coronaviruses of many other mammalian species may essentially be pre-adapted to human infectivity. Evidence suggests that there are many bat coronaviruses pre-adapted to emerge, and possibly to emerge pandemically ( Andersen et al., 2020 ; Hasan et al., 2020 ; Hu et al., 2017 ; Menachery et al., 2016 ; Wang et al., 2018 ; Zhou et al., 2020a ). Influenza A viruses (IAVs) infect cells via binding to terminal sialic acids found on lumenal respiratory epithelial cells of avian as well as mammalian and human hosts and contain a neuraminidase that cleaves these same receptors to allow viral release, which facilitates onward viral transmission ( Morens and Taubenberger, 2020 ; Taubenberger et al., 2019 ). Since sialic acids are found on a wide variety of mammalian and non-mammalian cells, it is not surprising that many viruses attach to these receptors, although viral affinities to receptors are complicated. For example, in vitro lectin assays suggest variable affinities of influenza for different types of terminal sialic acids, e.g., those with α-2,3 (ostensibly avian) and those with α-2,6 (ostensibly mammalian) linkages. However, autopsy studies have confirmed fatal human infections caused by IAVs that prefer either receptor, reflecting the complexity of virus-host interactions across the animal kingdom. Yet another aspect of infecting viruses is that some (e.g., HIV, influenza) express immunodominant epitopes on external proteins that dominate the elicitation of immune responses, resulting in less robust immune responses to other, often adjacent, epitopes. This may have the effect of blunting, or limiting the breadth of, an otherwise optimal host immune response. Almost all viruses have proteins that inhibit innate protective host responses, such as the host interferon response ( Blanco-Melo et al., 2020 ; Mesev et al., 2019 ). Some viruses are able to infect FcR bearing cells via antibody-dependent infection enhancement or ADE—a mechanism by which virus-IgG complexes are "tied down" by FcRs to the cell surface, facilitating cell entry via another receptor ( Morens, 1994 ; Sullivan, 2001 ). During the 2002–2003 SARS epidemic, it was found that both post-infectious and vaccine-elicited antibody caused in vitro ADE with SARS-CoV ( Jaume et al., 2012 ; Wang et al., 2016 , 2014 ; Yip et al., 2014 ). It is not known whether this phenomenon, studied mostly in vitro , has implications for natural human coronavirus infections or vaccinations ( Wan et al., 2020 ); however, it does represent a potential safety concern associated with vaccine development for SARS-CoV and SARS-CoV-2. In considering SARS-CoV-2 vaccine development and antibody treatment/prophylaxis, it is also of some concern that natural infection with the feline coronavirus (FECV) initiates a non-fatal infection that sometimes leads to development of viral variants (known as feline infectious peritonitis viruses [FIPVs]) that are macrophage-tropic and that can bind to host anti-spike protein antibody (antibody against the external viral protein that attaches to ACE-2 receptors and elicits protective immunity) to allow viruses to enter macrophages via FcRs (ADE), leading to a distinct and universally fatal disease known as feline infectious peritonitis, or FIP ( Vennema et al., 1990 ; Weiss and Scott, 1981 ). Post-infectious gene editing by the host has also been proposed as a mechanism for development of subacute sclerosing pan-encephalitis (SSPE), a fatal human disease associated with persistent natural measles virus infection complicated by generation of host-edited mutant viruses ( Cattaneo et al., 1986 ). This represents yet another variation on mechanisms of viral emergence ( Baranowski et al., 2001 ; Cattaneo et al., 1986 ; Novella et al., 2011 ). Fortunately, viruses such as those that cause SSPE, derived from in-host gene editing, are not necessarily transmissible. In experimental studies, early FIPV-associated feline deaths result from both FIPV spike protein vaccination and passive transfusion with anti-FIPV antibody ( Vennema et al., 1990 ; Weiss and Scott, 1981 ). Neither FECV nor FIPV are phylogenetically close to SARS-CoV or SARS-CoV-2, both being α-coronaviruses utilizing aminopeptidase N or other protein or glycan receptors rather than the ACE-2 receptors that bind SARS β-coronaviruses. But evidence for ADE with multiple different α- and β-coronaviruses suggests that as we proceed to develop SARS-CoV-2 vaccines and therapeutic antibodies, much remains to be learned about this complex viral family. Also of importance to the infectivity of newly emerging infectious diseases are viral genetic properties associated with pathogenicity and co-pathogenicity , exemplified most clearly with pandemic IAVs. The 1918 H1N1 pandemic virus, which killed an estimated 50 million people (equivalent to 200 million when adjusted to the 2020 population) was particularly lethal because of at least two inherent properties: (1) an avian-descended H1 hemagglutinin (HA) that is unusually cytopathic and immunopathogenic compared to the HAs of most other IAVs and (2) a marked co-pathogenic ability—the viral genetic basis of which remains poorly understood—to precipitate fatal bacterial bronchopneumonias in association with pneumopathogenic bacteria carried silently in the human upper respiratory tract ( Morens et al., 2008b ; Morens and Taubenberger, 2020 ; Taubenberger et al., 2019 ). We now know that not only avian H1s but also 4 of the other 15 avian HAs found within the wild waterfowl and shore birds ( Anseriformes and Charidriiformes ) reservoir have similar pathogenic properties, and thus they represent future threats for highly fatal pandemic emergences ( Morens and Taubenberger, 2020 ; Taubenberger et al., 2019 ). A perhaps even more shocking example of pandemic emergence associated with enhanced pathogenicity is that of Zika, a flavivirus known for decades. Zika had never caused a human epidemic; however, in 2015, it suddenly spread pandemically around the global tropical belt, causing millions of infections and severe fetal losses and birth defects ( Fauci and Morens, 2016 ). The apparent cause of the pandemic was a mutation resulting in the change of a single amino acid in the external viral glycoprotein ( Shan et al., 2020 ). Pathogenic variability undoubtedly applies to many other virus types, exemplified, for example, by comparing the highly pathogenic Ebola Zaire strain to the closely related but lowly pathogenic Ebola Reston strain ( Baseler et al., 2017 ). Although not yet adequately studied, pathogenic variability might also be a property of animal coronaviruses. That all three recently emerging human coronaviruses (the agents of SARS, MERS, and COVID-19) exact a high degree of human morbidity and mortality suggests that enzootic coronaviruses as a whole may be inherently pathogenic for humans. On the one hand, preliminary data suggest that SARS-CoV-2 may elicit an unbalanced innate immune response associated with decreased expression of interferons I and III and with increased inflammatory cytokine production ( Mesev et al., 2019 ), consistent with preliminary COVID-19 findings ( Vanderheiden et al., 2020 ). On the other hand, preliminary data suggest that viral-bacterial co-pathogenesis may be of somewhat less concern with SARS-CoV-2 than with influenza, measles, and other pathogenic respiratory viruses, although comprehensive clinical/autopsy series have not yet been published. With regard to coronavirus disease severity, it is worth considering the conventional wisdom (not always correct) that viruses that kill their hosts limit their own ability to be transmitted and that, if transmitted person to person, they would be expected to be selected for attenuation of pathogenicity over time. It is conceivable that the four endemic coronaviruses of humans—the β-coronaviruses OC43 and HKU1 and the α-coronaviruses 229E and NL63—emerged long ago as zoonotic and perhaps highly pathogenic viruses that evolved into attenuated forms over time ( Cui et al., 2019 ). Such natural attenuation of pathogenicity has not yet been observed with the agents of SARS, MERS, or COVID-19, although with a SARS-CoV-2 case-fatality in the range of 1%, and with evidence for significant asymptomatic and presymptomatic transmission in a largely susceptible population, evidence for selection pressures for attenuation may not be detectable in the short term. Moreover, viral evolution toward lower pathogenicity does not apply to all infectious diseases. For many other organisms such as cholera (expressing a bacteriophage toxin causing diarrhea) or rotavirus infection (with an NSP4 diarrhea-causing toxin), or for cough-inducing tuberculosis and many respiratory viruses, damage to the gastrointestinal tract and lungs, respectively, facilitates transmission, since diarrhea and coughing expel infectious pathogens into the environment, increasing the chance of infecting additional hosts. Disease severity thus reflects a balancing act between killing or incapacitating hosts, on the one hand, and optimizing microbial transmission, and therefore survival, on the other. Similar principles may also apply to non-viral diseases with environmental modes of transmission: for example, with enzootic anthrax, host-killing may be an important transmission mechanism, as rotting carcasses leave anthrax spores in the ground to reignite future infections ( Turner et al., 2014 ). Agent-host interaction variables are often exceedingly complex. The Role of the Host in the Emergence of Infectious Diseases Host variables that underlie the emergence of infectious diseases include those variables specific to individuals within the host population and those variables that relate to the host population as a whole ( Morens and Fauci, 2012 ; Morens et al., 2004 , 2008a ). Since a virus replicates within the cells of the host, and since viruses usually infect hosts via specific receptors on the cells of various tissues and organs, the new host must express cellular receptors or other cell-surface properties to which the virus can bind and initiate viral internalization. Major portals of host entry for infectious agents include those that are visibly external to the environment such as the skin or that can be reached directly from the environment such as the respiratory and gastrointestinal tracts, as well as organs reached systemically such as the liver, heart, and other internal organs. Human beings have many different organ systems, each with many different cell types, and with each cell having arrays of different receptors; therefore, it is not surprising that switching of a pathogen from an animal host to humans results in very different clinical and epidemiologic outcomes, including different disease manifestations and transmission mechanisms. These factors ultimately relate to the potential for establishment of infection in the new host as well as the likelihood of sustained transmission within the new host population and, as such, have a bearing on whether host-switching succeeds or fails. SARS-CoV and SARS-CoV-2 enter cells via ACE-2 receptors ( Wang et al., 2020 ), found on lung alveolar epithelial cells, gastrointestinal enterocytes, arterial and venous endothelial cells, and arterial smooth muscle cells, among other cell types ( Hamming et al., 2004 ; Wang et al., 2020 ), which explains the excretion of SARS-CoV-2 and potential transmission via the respiratory and enteric routes. With regard to the latter, although SARS-CoV-2 infects cells of the gastrointestinal tract, fecal transmission has not to date been implicated in significant person-to-person viral spread. Different viruses utilize widely different strategies for binding to, penetrating, and entering cells, e.g., polioviruses, HIV, influenza viruses, coronaviruses, and others ( Figure 4 ) ( Bowers et al., 2017 ; Cicala et al., 2011 ; Jayawardena et al., 2020 ; Laureti et al., 2018 ). Some viruses enter cells via binding to two different proximate receptors, a primary and secondary receptor, e.g., receptors for binding and for fusion (for example, numerous flaviviruses [ Laureti et al., 2018 ]). As noted, mechanisms of viral entry into cells are exceedingly variable and complex ( Jayawardena et al., 2020 ). Viruses may also infect macrophages and macrophage-like cells, as is the case with mosquito-borne and tick-borne flaviviruses. These viruses are injected into perivascular dermal tissue by their respective vectors and are taken up by dendritic cells and carried to regional lymph nodes, where they initiate systemic infection. ADE, discussed above, and other ADE-like phenomena, may also facilitate cellular infection ( Morens, 1994 ; Sullivan, 2001 ). Tissue/cell tropism also has a bearing on the types of immune responses that are elicited. For example, in a systemic infection like measles, high-level viremia is associated with infection of multiple organs, tissues, and cell types; the resulting broad systemic B and T cell responses lead to lifelong protection from reinfection. In contrast, influenza A viruses and respiratory syncytial virus (RSV), among many other respiratory viruses that infect surface epithelial cells, do not cause viremia and infect only surface respiratory epithelial cells. As a result, infectious virions do not have intimate interactions with the systemic immune system. The major site of influenza virus-immune system interaction is in the semi-organized tear duct-, nasal-, and mucosal-associated lymphoid tissues (TALT, NALT, and MALT), as well as the post-natally generated inducible bronchus-associated iBALT ( Moyron-Quiroz et al., 2007 ), leading to tissue compartmentalization of the immune response, perhaps in part explaining the weakly protective immune responses of naturally acquired or vaccine-associated influenza and RSV. Preliminary evidence from clinical and pathological studies of both SARS-CoV and SARS-CoV-2, which indicate viral infection of multiple tissues, is consistent with elicitation of robust and hopefully long-lasting protective immunity, providing a potential for control of COVID-19 with vaccines. More ominously, expression of ACE-2 receptors on endothelial and numerous other cells, and autopsy evidence of significant SARS-C0V-2 endothelial infection ( Fox et al., 2020 ), are consistent with systemic viral infection causing both pulmonary and extra-pulmonary pathology, including widespread microthrombus formation, among other outcomes. Some emerging viruses encounter pre-existing partial population immunity , e.g., pandemic influenza viruses. Most notably, in the influenza pandemics of 1968 and 2009, caused by an H3N2 and an H1N1 virus, respectively, segments of the population had pre-existing immunity that interfered with early viral spread and possibly with viral evolution. Although insufficient to prevent emergence, such population immunity did protect certain segments of the population ( Morens and Taubenberger, 2020 ; Taubenberger et al., 2019 ). In fact, in most influenza pandemics, influenza-specific case-fatality in the elderly, which increases regularly over about age 60 with seasonal influenza, is nevertheless apparently blunted by poorly characterized immune effects of prior influenza exposures, emphasizing the complexity of viral infectivity and host resistance factors ( Morens and Taubenberger, 2011 ). In contrast, there are suggestions that decreases in natural infections with, or vaccinations against, pathogens may facilitate the emergence of related organisms, e.g., increased incidence of human monkeypox after the cessation of smallpox vaccination that followed eradication of the monkeypox-related smallpox virus ( Lloyd-Smith, 2013 ), or the long-held theory that existing or newly evolving enteroviruses will emerge, or are already emerging, to fill an "ecologic niche" created by the near-eradication of the three polioviruses ( Rieder et al., 2001 ). In the case of COVID-19, some evidence suggests the absence of pre-existing population immunity, afforded by exposure to the four endemic coronaviruses, sufficient to prevent infection ( Corman et al., 2018 ). Although these endemic viruses share few epitopes that cross-react significantly with SARS-CoV-2 in serologic studies, it has been speculated that endemic coronavirus cross-protection may nevertheless prevent or at least limit the severity of disease in some, especially in young persons ( Nickbakhsh et al., 2020 ). There is preliminary but growing evidence that infectious disease severity or even susceptibility may in some cases be related to host genetic variables associated with the innate immune response, as is the case with epidemiologic information concerning severe disease caused by H5N1 poultry-associated influenza ( Morens and Taubenberger, 2015 ). Several host genes have been provisionally linked to susceptibility to such severe viral disease outcomes ( Nguyen et al., 2020 ; Tang et al., 2008 ), including the interferon-related transmembrane protein 3 (IFITM3) SNP rs22522-C allele ( Everitt et al., 2012 ; Zheng et al., 2017 ), which has been proposed (based on very preliminary data) to be involved in severity of SARS-CoV-2 disease ( Zhang et al., 2020 ). In addition, differential interactions with IFITMs such as IFITM3 include inhibiting the human endemic α-coronavirus while enhancing entry of SARS-CoV and MERS-CoV β-coronaviruses ( Huang et al., 2011 ; Zhao et al., 2018 ). With respect to COVID-19, several studies have associated blood group A in the ABO system with disease severity, although the mechanisms of this effect are not yet clear. ABO system associations with infectious diseases have also been shown for infections with noroviruses, H. pylori , and falciparum malaria; however, any such associations might also be indirect markers for unrelated genes. Newer data are beginning to define human immonotype risks for more severe disease ( Mathew et al., 2020 ), as well as posssible genetic signatures of severe diseases ( Gussow et al., 2020 ). At this time, data establishing specific genetic susceptibilities remain inconclusive for most diseases, including coronavirus diseases. This is a research area likely to be important in the future, since identification of susceptibilities for human disease severity has great implications for prevention, diagnosis, and treatment. Among the most important host factors for infection and for disease emergences/re-emergences are those associated with human behaviors , e.g., population growth, crowding, human movement, and many others, including behaviors that either perturb the environment or result in new human-created ecologic niches ( Figure 3 ). Regarding human movement, both the 1347–1348 Black Death (bubonic/pneumonic plague) and the 1832 cholera pandemic (which traveled from India to Europe and then to the Western Hemisphere) were spread along major trade and travel routes. In 1831–1832, 45 years before a coherent "germ theory" would be articulated, it was clear that as cholera spread slowly westward, it moved no faster than coaches and ships traveled: it slowed down in the winter as travel slowed down, and it picked up again in the summer as travel increased. The 1889 influenza pandemic traveled westward from Asia to Europe along railroad lines and then was exported globally along shipping routes. The 1957 influenza pandemic was spread by ships, but 11 years later the 1968 influenza pandemic was spread along air routes, the first example of global pandemic spread by airplanes. In 1981, acute hemorrhagic conjunctivitis was spread between international air hubs in the tropics and some temperate zones (e.g., to Florida and North Carolina). In 2002–2003, SARS was exported by air from Hong Kong to the Western Hemisphere and Europe. In 2019–2020, SARS-CoV-2 was spread globally from China in a similar manner. These many ancient and modern examples reflect the extraordinary importance of human population growth and movement in spreading diseases: the more populous and crowded we as a species become, and the more we travel, the more we provide opportunities for emerging diseases. The Role of the Environment in the Emergence of Infectious Diseases Many other human activities related to the environment have important consequences for disease emergence ( Allen et al., 2017 ; Dobson and Carper, 1996 ; Fauci and Morens, 2012 ; Morens et al., 2020b ; Morens and Fauci, 2012 ; Morens et al., 2004 , 2008a , 2019 ; Morens and Taubenberger, 2020 ). Human water storage practices in Northern Africa, beginning about 5,000 years ago, led to the emergence of a new, exclusively human-adapted mosquito, Aedes aegypti , which created a secondary ecologic niche for the emergence of yellow fever virus and, centuries later, dengue, chikungunya, and Zika, all of which then went on to spread pandemically. Depots of used rubber tires create ideal breeding sites for a related human-adapted mosquito, Aedes albopictus , which in recent decades has spread globally across the tropical and subtropical zones, transmitting many of these same arboviral diseases widely, if less efficiently. Predictably, viruses and viral vectors have adapted to environmental influences: a single locus mutation in the Aedes aegypti -adapted Indian Ocean strain of chikungunya virus has newly adapted it, without loss of fitness for aegypti , to Aedes albopictus, widely prevalent in the region ( Tsetsarkin and Weaver, 2011 ). Land-management practices have been associated with re-emergences of Eastern equine encephalitis ( Morens et al., 2019 ); deforestation with emergences of Zika and Hendra viruses; road-building and environmental degradation with the spread of Bolivian hemorrhagic fever and HIV (infections spread by truckers and truck stop prostitution); and poverty, crowding, and poor sanitation with re-emergences of many diseases such as tuberculosis and cholera. For centuries, wars have precipitated the re-emergences of many diseases ( Dobson and Carper, 1996 ; Fauci and Morens, 2012 ; Morens and Fauci, 2012 ; Morens et al., 2004 , 2008a , 2020b ; Morens and Taubenberger, 2020 ), e.g., the Serbian typhus epidemic during World War I, which killed 150,000 people, mostly civilians. In the 1700s, when typhus had not been identified as a specific disease, textbooks listed two separate conditions under the nosologic terms "war typhus" and "jail typhus," reflecting human activities that provoked long-ago disease emergences. A classic epidemiology text published over a century ago, still studied today, is titled Epidemics Resulting from Wars ( Prinzing, 1916 ). It is suspected that SARS-CoV-2 emerged in 2019, as did SARS-CoV in 2002, and as did H5N1 and H7N9 poultry-associated influenza, in 1997 and 2013, respectively, from wet markets in China ( Morens et al., 2020a ; Morens et al., 2020b ). These four diseases may thus represent four deadly emergences, within an 18-year span, from one cultural practice in one region of the world. These and many other examples ( Dobson and Carper, 1996 ; Fauci and Morens, 2012 ; Lu et al., 2020 ; Morens and Fauci, 2012 ; Morens et al., 2004 , 2008a , 2019 , 2020b ; Morens and Taubenberger, 2020 ; Zhang and Holmes, 2020 ; Zhou et al., 2020b ) constitute a powerful argument that human activities and practices have become the key determinant of disease emergence. Emergence of Diseases Leading to Epidemicity and Endemicity Whatever human behaviors lead to the emergence of infectious diseases, the newly emerged organism cannot survive continually in humans without adapting to one of several direct or indirect mechanisms of person-to-person spread ( Table 4 ). Many emerging organisms, such as the hantaviral agents of hantavirus pulmonary syndrome and Korean hemorrhagic fever, or the arenaviruses causing Argentine, Bolivian, and Lassa hemorrhagic fevers, result from dead-end exposures to reservoir rodents and rarely spread from person to person. Preventing and controlling these types of emergences thus focus on the enzootic reservoir. Table 4 Mechanisms of Infection Transmission from Person to Person and from Animal to Person Respiratory including Environmental Gastrointestinal including Environmental Inoculation Direct Vectorborne Influenza ∗ Cholera Anthrax 2 Chikungunya ∗ Human coronaviruses Noroviruses Dracunculiasis 3 Dengue ∗ Measles Rotaviruses Gonorrhea Lyme disease ∗ Rhinoviruses Salmonellosis Hepatitis B and C Malaria ∗ SARS, ∗ COVID-19, ∗ MERS 1 , ∗ Typhoid fever HIV ∗ Yellow fever Some human enteroviruses ∗ Some human enteroviruses Syphilis Zika ∗ Selected endemic and emerging infectious diseases transmitted to humans via multiple mechanisms. Many diseases are transmitted by more than one mechanism. For example, most of the respiratory-transmitted diseases are also transmitted by hands and on fomites; most of the gastrointestinal diseases are uncommonly transmitted directly and more commonly transmitted by water, food, and on fomites. Some enteroviruses are predominantly transmitted by the fecal-oral (gastrointestinal) route, whereas others are predominantly transmitted by the respiratory route. Anthrax is transmitted by respiratory, gastrointestinal, and inoculation routes. Understanding mechanisms of transmission is important not only because it helps control those diseases that emerge but also because it provides opportunities to control multiple diseases transmitted by the same mechanisms. ∗ New/important emergences and re-emergences in the past 4 decades. 1 MERS is largely a zoonotic disease, although person-to-person spread can occur with relative inefficiency. 2 Anthrax is acquired by inoculation, inhalation, or ingestion of spores, but the cutaneous form of anthrax can be spread person to person. 3 Dracunculiasis is environmentally acquired as part of a complex life cycle including human expulsion of pathogens into the water and water-borne transmission to other humans. Person-to-person transmissibility, a necessary step in establishing epidemicity and pandemicity after host-switching, can occur via one or more of four basic mechanisms ( Table 4 ): respiratory;gastrointestinal; environmental spread via an intermediate environmental state such as waterborne, foodborne, and fomite-mediated; and environmental spread via inoculation, including vectorborne. It is noteworthy that these mechanisms reflect not only human societal practices such as sanitary practices, embracing physical closeness, and sexual activities but also ways in which people interact with the environment such as storage and drinking of water from different sources, food procurement, and food preparation practice. Even so, many variables interact to produce different clinical and epidemiologic aspects of disease transmitted from person to person. For example, while both Ebola and SARS-CoV-2 are potentially fatal diseases transmitted person to person, SARS-CoV-2 is, like most respiratory diseases, associated with asymptomatic infection in many and is often transmitted to others by an asymptomatic or pre-symptomatic infected person. Ebola, on the other hand, has a lower frequency of asymptomatic infection and is of low transmissibility right up to the time of illness onset. Moreover, the mechanisms of person-to-person transmission are distinct for these two diseases. SARS-CoV-2 is transmitted via hands and fomites, respiratory droplets and aerosol, including transmission by superspreading events where large numbers of individuals are infected by a single person, almost invariably in closed crowded settings. Ebola, on the other hand, is transmitted via virus-contaminated bodily secretions, and infection is usually acquired by persons touching contaminated fluids or fomites, for example in providing nursing care, burial services, handling towels, bed linens, utensils, etc. ( Baseler et al., 2017 ). Although general principles of infection transmission are understood, specific diseases may vary in clinical and epidemiologic features that bear upon the type and intensity of transmission. In this context, ongoing research into the origin of SARS-CoV-2 seeks to learn where, how, and why the virus emerged as a human pandemic disease ( Boni et al., 2020 ; Latinne et al., 2020 ). SARS-CoV-2 clusters phylogenetically within an extensive but still not fully characterized universe of wild bat β-coronaviruses found in many species over much of the globe ( Anthony et al., 2017 ; Cui et al., 2019 ; Hu et al., 2017 ; Letko et al., 2020 ; Lu et al., 2020 ; Morens et al., 2020a ; Zhang and Holmes, 2020 ; Zhou et al., 2020b ). The discovery that its closest identified viral relatives are enzootic in horseshoe ( Rhinolophus ) bats ( Zhou et al., 2020a , 2020b ) indicates that SARS-CoV-2 probably emerged from an as-yet-unidentified bat reservoir either directly or after infection of an intermediate host such as a pangolin ( Boni et al., 2020 ; Letko et al., 2020 ; Li et al., 2020 ; Zhou et al., 2020a , 2020b ). As was true of SARS-CoV 18 years ago, the specific determinants of SARS-CoV-2 emergence remain obscure. Gaining a better understanding of the enormous reservoir of bat coronaviruses has been an urgent priority since the 2002 SARS epidemic, and remains so today. Considerable surveillance and phylogenetic and experimental work remains to be done. In 2020, it is among our most urgent research priorities ( Latinne et al., 2020 ). One of the most important unanswered questions we face in the ongoing COVID-19 pandemic relates to the evolutionary potential of human-adapted SARS-CoV-2. Will it, similar to human IAVs in recent centuries, evolve to persist as a permanent human pathogen by mutating to escape the population herd immunity it creates? And if it does persist, will it attenuate over time, as the four endemic coronaviruses may have done centuries ago? Or, on the other hand, will it increase in pathogenicity as the pandemic H3N2 IAV has done over the past 52 years? Because SARS-CoV-2 lacks a segmented genome, we are spared at least one genetic trick (gene reassortment) underlying IAV pandemic emergences; however, like human IAVs, SARS-CoV-2 can be expected to evolve by mutation as it spreads through human populations, and it has the additional capacity of evolving by genetic recombination. Enzootic evidence supports a very high degree of recombination of SARS-like CoVs in nature. Will these capacities allow SARS-CoV-2 to escape from population immunity elicited by natural infection or future vaccination? Only additional time, and much important research, will begin to answer these questions. The Enigma of Host-Switching Perhaps the single biggest mystery of emerging diseases is how microorganisms, including animal-adapted microorganisms, switch species to infect humans. Host-switching from animal to human has long been conceptualized as the crossing of a fitness valley , in which a virus adapted to transmitting between members of host species A must somehow simultaneously develop the capacity to productively infect cells of new host species B ( Figure 5 ) ( Dolan et al., 2018 ; Geoghegan and Holmes, 2018 ; Kuiken et al., 2006 ; Parrish et al., 2008 ). In this paradigm, the depth of the fitness valley reflects host-to-host barrier challenges that have to be overcome. Our understanding of mechanisms of viral evolution and host-switching is nevertheless incomplete, in part because of inability to reconcile experimental within-host viral evolutionary findings and natural-world findings that examine viral evolution during outbreaks, including new disease emergences ( Geoghegan and Holmes, 2018 ). Among many complicating factors, rapid intense person-to-person transmission of new viruses in large human populations often produces complex genetic diversity, confounding attempts to link viral variation to viral phenotypic changes and selection pressures eliciting them. Figure 5 Proposed Molecular Mechanisms of Host-Switching Proposed mechanisms of cross-species host-switching of infectious agents (after Kuiken et al., 2006 ). Steep (A) and shallow (B) fitness valleys between donor and recipient host species reflect adaptational barriers that need to be crossed. (A) and (C) show that a greater number of sequential adaptational mutations are needed in (A) to cross the fitness valley and then adapt to the new host, as compared to the situation shown in (B) and (D), where greater donor-host similarities facilitate switching. (C) and (D) represent the associated phylogenetic trees: in (C), the donor host and recipient host viruses most go through significant adaptational steps, including those associated with initial transmission in the new host. In (D), the new host receives an infectious agent that is partially pre-adapted; successful emergence requires fewer adaptational mutations. The situation is particularly complicated for the most important category of newly emerging disease agents: the RNA viruses, which include SARS-CoV, SARS-CoV-2, MERS, Ebola, and influenza, as well as dengue, Zika, and other arboviruses. These viruses evolve as, and are transmitted as, complex quasispecies, or viral swarms, which contain many viral variants of differing degrees of relatedness. It is unclear whether and/or to what extent transmission/host-switching reflects Darwinian evolution of novel virus variants, as opposed to evolution based on whole-quasispecies fitness ( Geoghegan and Holmes, 2018 ). In the latter conceptualization, viral quasispecies evolve together as a diverse array of optimally fit, less optimally fit, and least optimally fit variants, trading off perfect host fitness for adaptational flexibility. While it has long been assumed that the major determinants of host-switching are the evolutionary closeness of hosts A and B and the diversity of their transmitting quasispecies, recent research suggests that pathogen opportunity may be the major determinant of host-switching ( Anishchenko et al., 2006 ; Araujo et al., 2015 ). In essence, even a virus poorly fit to a potential host can adapt to infect that host if given enough chances. The implications are profound. If host-switching is opportunity driven, e.g., for SARS- CoV and CoV-2, Nipah, and Hendra, then prevention and control will have to focus not only on the infectious agents themselves but also on human behaviors, such as the animal-human interface, represented by shopping in live animal markets, preparation and consumption of bush meat, intensive farming/animal husbandry, environmental degradation, and other human behaviors ( Allen et al., 2017 ; Carroll et al., 2018 ). Looked at another way, infectious diseases may be emerging into humans with greater frequency than appreciated; however, historically they have not usually achieved sustained-enough transmission that would lead to detection of the emergence. For example, before the recognized emergence of Zika, low human seroprevalence rates had been detected in enzootic areas for decades, but without detection of human outbreaks. Before the emergence of SARS-CoV-2, antibodies to it or to closely related sarbecoviruses were detected in humans exposed to bat coronaviruses ( Wang et al., 2018 ). MERS has emerged from dromedary camels to humans on multiple occasions; however, despite limited chains of subsequent human-to-human spread, it has not become an established human disease. H5N1 avian influenza has infected millions of humans; however, only a small percentage manifest a recognized disease (often severe or fatal), and human-to-human transmission has been rarely reported ( Morens and Taubenberger, 2015 ). These and many other examples suggest that disease emergence into humans may be common; however, sustained transmission between humans has been a rare occurrence resulting from incompletely understood factors. But in looking at the recent spate of deadly emergences noted above, we must now ask whether human behaviors that perturb the human-microbial status quo have reached a tipping point that forecasts the inevitability of an acceleration of disease emergences. Such enigmas need to be better understood if we are to control emergences of infectious diseases ( Dobson et al., 2020 ). Better understanding of the obscure mechanisms of emergences might allow us to anticipate emergence risks by (1) surveilling and characterizing taxonomic groups of potentially pre-emergent viruses, including viruses that may be the nearest to emerging, e.g., coronaviruses, henipaviruses, flaviviruses, arenaviruses, and filoviruses; (2) conducting intensive research on suspected high-risk viral categories to identify, in experimental animals, conserved epitopes for vaccine development and targets for antiviral therapies; (3) characterizing mechanisms of potential human transmission in experimental animal studies; (4) developing control mechanisms in areas such as animal husbandry, wildlife interaction, and rodent, vector, and mosquito/tick control; (5) devising environmental, land management, wildlife conservation/control programs; and (6) utilizing new virologic, informatic, and technological approaches to understand viral evolution and even predict emergence potential ( Allen et al., 2017 ). There are many examples where disease emergences reflect our increasing inability to live in harmony with nature. Nipah virus emergence followed agricultural burning of forests, which led to displacement of infected bats; bats then went to roost in trees that shaded intensively farmed pigs that were crowded into small areas, which led to infection of pigs via bat droppings, which in turn led to human outbreaks in pig farmers ( Morens et al., 2004 ). Fish meal farming by fishermen in waters around the globe damages ecosystems by over-fishing and deprives local residents of food sources, leading to poverty and human movement, which exacerbate the potential for disease emergence. In Asia, the fish meal is used for breeding and over-feeding of farmed animals, often in intensive farming practices, which increase the likelihood of zoonotic diseases. Yellow fever, dengue, chikungunya, and Zika are all associated with urban crowding, poor sanitation, and water storage. Over many centuries, urbanization and crowding has led to rodent infestation and to rodent-borne diseases such as plague, murine typhus, and rat-bite fever. The ongoing COVID-19 pandemic reminds us that overcrowding in dwellings and places of human congregation (sports venues, bars, restaurants, beaches, airports), as well as human geographic movement, catalyzes disease spread. Living in greater harmony with nature will require changes in human behavior as well as other radical changes that may take decades to achieve: rebuilding the infrastructures of human existence, from cities to homes to workplaces, to water and sewer systems, to recreational and gatherings venues. In such a transformation we will need to prioritize changes in those human behaviors that constitute risks for the emergence of infectious diseases. Chief among them are reducing crowding at home, work, and in public places as well as minimizing environmental perturbations such as deforestation, intense urbanization, and intensive animal farming. Equally important are ending global poverty, improving sanitation and hygiene, and reducing unsafe exposure to animals, so that humans and potential human pathogens have limited opportunities for contact. It is a useful "thought experiment" to note that until recent decades and centuries, many deadly pandemic diseases either did not exist or were not significant problems. Cholera, for example, was not known in the West until the late 1700s and became pandemic only because of human crowding and international travel, which allowed new access of the bacteria in regional Asian ecosystems to the unsanitary water and sewer systems that characterized cities throughout the Western world. This realization leads us to suspect that some, and probably very many, of the living improvements achieved over recent centuries come at a high cost that we pay in deadly disease emergences. Since we cannot return to ancient times, can we at least use lessons from those times to bend modernity in a safer direction? These are questions to be answered by all societies and their leaders, philosophers, builders, and thinkers and those involved in appreciating and influencing the environmental determinants of human health. Summary and Conclusions SARS-CoV-2 is a deadly addition to the long list of microbial threats to the human species. It forces us to adapt, react, and reconsider the nature of our relationship to the natural world. Emerging and re-emerging infectious diseases are epiphenomena of human existence and our interactions with each other, and with nature. As human societies grow in size and complexity, we create an endless variety of opportunities for genetically unstable infectious agents to emerge into the unfilled ecologic niches we continue to create. There is nothing new about this situation, except that we now live in a human-dominated world in which our increasingly extreme alterations of the environment induce increasingly extreme backlashes from nature. Science will surely bring us many life-saving drugs, vaccines, and diagnostics; however, there is no reason to think that these alone can overcome the threat of ever more frequent and deadly emergences of infectious diseases. Evidence suggests that SARS, MERS, and COVID-19 are only the latest examples of a deadly barrage of coming coronavirus and other emergences. The COVID-19 pandemic is yet another reminder, added to the rapidly growing archive of historical reminders, that in a human-dominated world, in which our human activities represent aggressive, damaging, and unbalanced interactions with nature, we will increasingly provoke new disease emergences. We remain at risk for the foreseeable future. COVID-19 is among the most vivid wake-up calls in over a century. It should force us to begin to think in earnest and collectively about living in more thoughtful and creative harmony with nature, even as we plan for nature's inevitable, and always unexpected, surprises.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4894344/

Fundamental Properties of One-Dimensional Zinc Oxide Nanomaterials and Implementations in Various Detection Modes of Enhanced Biosensing

Recent bioapplications of one-dimensional (1D) zinc oxide (ZnO) nanomaterials, despite the short development period, have shown promising signs as new sensors and assay platforms offering exquisite biomolecular sensitivity and selectivity. The incorporation of 1D ZnO nanomaterials has proven beneficial to various modes of biodetection owing to their inherent properties. The more widely explored electrochemical and electrical approaches tend to capitalize on the reduced physical dimensionality, yielding a high surface-to-volume ratio, as well as on the electrical properties of ZnO. The newer development of the use of 1D ZnO nanomaterials in fluorescence-based biodetection exploits the innate optical property of their high anisotropy. This review considers stimulating research advances made to identify and understand fundamental properties of 1D ZnO nanomaterials, and examines various biosensing modes utilizing them, while focusing on the unique optical properties of individual and ensembles of 1D ZnO nanomaterials specifically pertaining to their bio-optical applications in simple and complex fluorescence assays.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4515727/

Climate Change, Drought and Human Health in Canada

Droughts have been recorded all across Canada and have had significant impacts on individuals and communities. With climate change, projections suggest an increasing risk of drought in Canada, particularly in the south and interior. However, there has been little research on the impacts of drought on human health and the implications of a changing climate. A review of the Canadian, U.S. and international literature relevant to the Canadian context was conducted to better define these impacts and adaptations available to protect health. Drought can impact respiratory health, mental health, illnesses related to exposure to toxins, food/water security, rates of injury and infectious diseases (including food-, water- and vector-borne diseases). A range of direct and indirect adaptation (e.g., agricultural adaptation) options exist to cope with drought. Many have already been employed by public health officials, such as communicable disease monitoring and surveillance and public education and outreach. However, gaps exist in our understanding of the impacts of short-term vs. prolonged drought on the health of Canadians, projections of drought and its characteristics at the regional level and the effectiveness of current adaptations. Further research will be critical to inform adaptation planning to reduce future drought-related risks to health. 1. Introduction The Intergovernmental Panel on Climate Change (IPCC) recognizes that climate change and the altered frequencies of associated extreme events such as drought and floods are expected to have negative impacts on human health [ 1 ]. For Canada, projections call for more frequent and severe extreme weather events including drought [ 2 ]. Research into the health impacts of drought and how climate change may enhance these effects is still in its infancy. A number of key health concerns related to climate change including food-, vector- and water-borne diseases may be exacerbated by drought. In addition, there is the potential for broad and lengthy exposure of Canadians to these episodes. There is particular concern about changes in the frequency, duration and severity of drought in the future due to climate change [ 3 ]. For these reasons, understanding of drought-related health risks is a climate change research priority. This paper reviews current understanding of the health impacts associated with drought that can be anticipated with climate change in the Canadian context drawing on Canadian, U.S. and international literature sources. It assesses the state of knowledge about adaptation options to protect health in the face of drought conditions and suggests future research directions to support adaptation efforts by health sector officials. 1.1. Canadian Drought Features There is currently no single, clear, consistent definition of drought. For the purposes of this discussion, drought will be considered a prolonged period of abnormally dry weather that depletes water resources for human and environmental needs [ 4 , 5 ]. Droughts stress water availability by lowering lake and reservoir levels, reducing stream flows, diminishing groundwater supplies and depleting soil moisture. Droughts can also create major environmental hazards such as reduced water quality, wetland loss, soil erosion and degradation, ecological habitat destruction and increased risk of wildfires. Prolonged, large-area droughts are one of Canada's major natural disasters and can have large scale impacts on various sectors including human health and society, agriculture, forestry, industry, municipalities and recreation [ 4 ]. Droughts differ from other disasters (e.g., floods) since they have longer durations, lack easily identified onsets and terminations and their recurrence is practically certain in most environments including many regions of Canada [ 6 ]. Climate anomalies, lasting from a month to one or more years, are the root of most droughts, although human impacts on resources and climate and the changing demand for water are major contributing factors [ 7 ]. There is no one "straightforward" manifestation of drought. Each drought manifests itself differently depending on factors, such as area affected, duration, intensity and society's capability to adapt to water shortages. Thus, although the general features of drought tend to be similar, the actual features vary between droughts and evolve within a single drought. This complexity may lead to varying impacts [ 8 , 9 ]. Some precipitation normally occurs even within a period of drought. It is not necessarily the case that the number of days with precipitation actually decreases during a drought; the relative lack of large precipitation events can for example be a contributor to drought formation. When it occurs during drought, precipitation tends to come in small amounts and to be highly scattered spatially due to convective processes, although more widespread rain can fall as well [ 9 ]. Sometimes, droughts can even be linked with heavy precipitation events. This can occur when the atmospheric forcing that is causing reduced precipitation in one region results in nearby areas receiving more rain than usual. The boundary between these regions can shift, thereby creating a situation in which conditions in some locations can alternate between both extremes. In some cases, the "end" or significant reduction of a drought can come in the form of a single heavy precipitation event. For example dry conditions in 2002 across the Prairies were alleviated by a torrential multi-day event [ 10 ]. Surface air temperatures are often high/elevated during drought, although this is not always the case. For some long duration droughts, there are periods when temperatures are below normal [ 11 ]. Droughts can also be windy or calm. Moderate to high winds can lead to erosion and transport of desiccated soil, although the number of these instances has generally declined, in part due to altered farming practices. 1.2. Historical Occurrence of Canadian Droughts Canada is considered a country with abundant water resources; however, these resources and the demands on them are distributed unevenly across the country [ 12 ] resulting in varied vulnerability to, and impacts from, drought. For example, 98% of the Canadian population resides in the South, which only provides 38% of the water yield (Water yield is the volume of freshwater resources for a defined geographic area and time period. It is an estimate of the amount of renewable water and non-renewable glacial melt) [ 12 ]. In the North, the per capita water yield is 98 times greater than in the South [ 12 ]. The water yield in the Prairies is a fraction (3%–12%) of that in the Great Lakes, Maritime Coastal or Pacific Coastal drainage regions and decreased between 1971 and 2004 [ 12 ]. During the 20th century, meteorological measurements have shown significant warming during all seasons but the greatest rates have been during winter and spring. They have also revealed general increases in precipitation, although considerable decadal-scale variability is present over many regions of the country. For example, Zhang et al. [ 13 ] determined that mean annual air temperature has increased by an average of 0.9 °C over southern Canada between 1900 and 1998 with the greatest rates seen in the West during the winter and spring (southern Canada broadly refers to south of 60N; an exception is [ 314 ] and mention in [ 315 ] where southern Canada refers to south of 55N. See Figure 5 in [ 4 ] for additional details. For Statistics Canada discussions, see [ 12 ] for a description of standard drainage area classifications). Subsequent analysis extending the target period to 2010 confirmed the warming has continued and even accelerated with an increase in the mean annual temperature of 1.5°C for southern Canada over this longer period [ 16 ]. Zhang et al. [ 13 ] also found that annual total precipitation increased significantly from 1990–1998 over most of southern Canada with the exception of southern Alberta and Saskatchewan. Analysis expanded to the period 1900 to 2009 also found an 8.7% increase in annual rainfall for southern Canada [ 17 ]. Snowfall mainly increased in northern Canada while at the same time decreasing significantly in the Southwest for the period 1950–2009 [ 17 ]. An analysis of changes in extreme temperature indices over southern Canada from 1900–2003 and the entire country from 1950–2003 found that over most of the country, there was an increase in the number of warm days where the maximum temperature (>90th percentile) and the number of warm summer days (maximum temperature >25°C) during the last half century [ 18 ]. The number of hot days (>30°C) and summer warm days (>90th percentile) also increased over all of Canada, but were only significant in southwestern regions [ 19 ]. The analysis revealed little change in the number of summer warm spells. An examination of changes in extreme temperature and precipitation indices over southern Canada from 1900–2003 and the entire country from 1950–2003 found a decrease in the consecutive number of days with no measurable precipitation in British Columbia and Atlantic Canada. Trends in drought occurrence and severity in Canada are often monitored using meteorological drought indices, such as the Standardized Precipitation Index (SPI), the Palmer Drought Severity Index (PDSI), and the Standardized Precipitation Evapotranspiration Index (SPEI) [ 4 ]. These indices are designed to assess regional precipitation anomalies (the SPI), or a region's water balance using both precipitation and temperature data (PDSI and SPEI), to account for evapotranspiration effects. Results from trend investigations incorporating these various drought indices have shown variable results over Canada, often being dependent on the time period used in the trend calculation (mainly due to the aforementioned decadal-scale variability in precipitation). However, an increase in drought (i.e., a decreasing Palmer Drought Severity Index (PDSI) trend) over most of the country was observed between 1950 and 2002 [ 4 ]. Western Canada, notably the southern Canadian Prairies and interior valleys of British Columbia, are more susceptible to drought because they lie in the lee of major mountain ranges and, as a result, generally receive little precipitation but with high variability [ 20 ]. During the past 100+ years, several long-duration droughts have occurred in western Canada. The Canadian Prairies have experienced large-area, multi-year dry episodes during the 1890s, 1930s, late 1950s/early 1960s, 1980s and more recently from 1999–2005 [ 21 , 22 ]. During the notable drought period from 2001–2002, below average stream flows were reported across the country extending from western Canada to Ontario and the Atlantic Provinces [ 23 ]. Water levels of most closed-basin lakes have declined throughout the 20th century across southcentral Alberta and Saskatchewan in western Canada, although levels have increased in the past few years [ 24 ]. Farther east, droughts do occur in southern Ontario and Quebec, although these events are usually shorter, smaller in area, less frequent and less intense [ 20 ]. Nonetheless, there have been major drought occurrences in the East during the last century [ 25 ] resulting in severe impacts, including low water levels in the Great Lakes [ 26 ]. Between 1920 and 1999, southern Ontario experienced major droughts in 1930, 1933, 1934, 1936, 1963, 1998, and 1999 [ 25 ]. Ontario was also impacted by drought in 2001–2002 [ 23 ]. Great Lakes' water levels have shown considerable decadal-scale variability during the 20th century with no evidence of any long-term trend. However, lower Great Lakes' water levels have coincided with the droughts of the 1930s, early 1960s and from early 2000 to the present [ 20 ].In 2011, water levels in the Great Lakes–St. Lawrence Rivers system reached record low levels and some activities/sectors in Ontario watersheds were asked to reduce water consumption under the Ontario Low Water Response Plan (OLWR)[ 23 ]. Déry et al. [ 27 ] found that freshwater discharge rates decreased over a period of 37 years for the majority of key rivers feeding the Hudson, James and Ungava Bays. This watershed is fed by water from five provinces, the territory of Nunavut andfour U.S. states, covering an area that is equivalent to one-third of Canada [ 27 ]. Along the East Coast, droughts in the Atlantic provinces are infrequent. However, a side effect of their rarity is that these regions have lower adaptive capacity to deal with droughts, making the region more susceptible to drought impacts [ 28 ]. The Atlantic provinces were affected by the drought of 2001–2002, with river flows in the Atlantic Provinces dropping to a 20-year low [ 23 ]. An analysis of meteorological indices (southern Canada 1900–2003; entire country 1950–2003) indicated a decrease in the number of days with no measurable precipitation in Atlantic Canada [ 18 ]. The context surrounding this shift will be discussed further below. 1.3. Future Droughts All Global Climate Models (GCMs) project future increases of summer continental interior drying and associated risk of droughts. The greater overall risk is ascribed to further increases in temperature and resultant potential evapotranspiration not being offset by precipitation increases [ 3 ]. In fact, the recent Fifth Assessment Report from the IPCC [ 29 ] stated that an increase in the intensity and duration of future droughts on a regional to global scale is likely ( i.e. , "medium" confidence) by the end of the 21st century. This statement especially applies to southern Europe and the Mediterranean region, central Europe, central North America, Central America and Mexico, northeast Brazil, and southern Africa. Elsewhere there is overall "low" confidence because of inconsistent projections of drought changes (dependent both on the GCM model and dryness index). Definitional issues, lack of observational data, and the inability of models to include all the factors that influence droughts preclude stronger confidence than "medium" in drought projections (see [ 30 ]). Some global studies have indicated that droughts will be slightly more frequent and longer in duration as compared to present-day conditions by the second half of the 21st century [ 31 , 32 ]. Specifically for North America, Trenberth [ 33 ], using several GCM simulations, showed that future warming will increase both evaporation and surface drying, thereby leading to longer and more intense drought particularly in the summer over continental interior regions of the continent. A recent study found that the southwest and central plains regions in the continental U.S. will be significantly drier by the end of the century when compared with historical periods. This was the case even across various GCMs, and different metrics of soil moisture in both moderate and high greenhouse gas emission scenarios [ 34 ]. Overall, future precipitation/water resource projections suggest that in Canada, the North will become wetter and the South drier especially in summer [ 35 ]. However, limited research has been carried out regarding potential future drought occurrence across Canada and most investigations have focused on the particularly drought-prone Canadian Prairies. A first-order assessment of future drought occurrence over southern Canada (2041–2070) by Bonsal and Regier [ 14 ] found future annual precipitation will increase over most of the southern region of Canada but this will be more than offset by increased temperatures that lead to higher evaporation so increases in drought are likely. The future mean PDSI estimates indicate the prevalence of more frequent drought for the southern half of the Canadian Prairies in the latter half of the 21stcentury with persistent negative PDSI values expected after 2040 [ 15 ]. This suggests the possibility of a regime shift to a more arid climate in the southern Prairies due to the effects of projected summer temperature increases. Future projections of PDSI also revealed that multi-year droughts lasting 10 or more years are more likely than what has been historically observed. Projections suggest approximately three such events per 100 years as compared to the approximately once every 100 years that has been observed. Similar results for the southern Canadian Prairies were also documented by PaiMazumder et al. [ 36 ] using simulations from the Canadian Regional Climate Model for the period 2041–2070. Although published studies for the rest of Canada are limited, the authors noted that in more northern areas of the Prairies, drought events are projected to be less severe and less frequent. Considerable uncertainty exists with respect to future precipitation events, particularly on a regional and intra-seasonal basis. Bonsal et al. [ 15 ] found considerable range in projections in their assessment of changes to the temporal and spatial characteristics of future (2011–2100) PDSI and SPI using statistically downscaled daily temperature and precipitation data over the southern Canadian Prairies, from three GCMs with different greenhouse gas emission scenarios. In summary, although there are no comprehensive analyses of future drought events across all of Canada to provide detailed regional and seasonal drought information, both global investigations and a handful of studies conducted for the Canadian Prairies suggest that future droughts will likely increase, particularly in those regions that already experience these events ( i.e. , more southern and interior areas of the country). 1.1. Canadian Drought Features There is currently no single, clear, consistent definition of drought. For the purposes of this discussion, drought will be considered a prolonged period of abnormally dry weather that depletes water resources for human and environmental needs [ 4 , 5 ]. Droughts stress water availability by lowering lake and reservoir levels, reducing stream flows, diminishing groundwater supplies and depleting soil moisture. Droughts can also create major environmental hazards such as reduced water quality, wetland loss, soil erosion and degradation, ecological habitat destruction and increased risk of wildfires. Prolonged, large-area droughts are one of Canada's major natural disasters and can have large scale impacts on various sectors including human health and society, agriculture, forestry, industry, municipalities and recreation [ 4 ]. Droughts differ from other disasters (e.g., floods) since they have longer durations, lack easily identified onsets and terminations and their recurrence is practically certain in most environments including many regions of Canada [ 6 ]. Climate anomalies, lasting from a month to one or more years, are the root of most droughts, although human impacts on resources and climate and the changing demand for water are major contributing factors [ 7 ]. There is no one "straightforward" manifestation of drought. Each drought manifests itself differently depending on factors, such as area affected, duration, intensity and society's capability to adapt to water shortages. Thus, although the general features of drought tend to be similar, the actual features vary between droughts and evolve within a single drought. This complexity may lead to varying impacts [ 8 , 9 ]. Some precipitation normally occurs even within a period of drought. It is not necessarily the case that the number of days with precipitation actually decreases during a drought; the relative lack of large precipitation events can for example be a contributor to drought formation. When it occurs during drought, precipitation tends to come in small amounts and to be highly scattered spatially due to convective processes, although more widespread rain can fall as well [ 9 ]. Sometimes, droughts can even be linked with heavy precipitation events. This can occur when the atmospheric forcing that is causing reduced precipitation in one region results in nearby areas receiving more rain than usual. The boundary between these regions can shift, thereby creating a situation in which conditions in some locations can alternate between both extremes. In some cases, the "end" or significant reduction of a drought can come in the form of a single heavy precipitation event. For example dry conditions in 2002 across the Prairies were alleviated by a torrential multi-day event [ 10 ]. Surface air temperatures are often high/elevated during drought, although this is not always the case. For some long duration droughts, there are periods when temperatures are below normal [ 11 ]. Droughts can also be windy or calm. Moderate to high winds can lead to erosion and transport of desiccated soil, although the number of these instances has generally declined, in part due to altered farming practices. 1.2. Historical Occurrence of Canadian Droughts Canada is considered a country with abundant water resources; however, these resources and the demands on them are distributed unevenly across the country [ 12 ] resulting in varied vulnerability to, and impacts from, drought. For example, 98% of the Canadian population resides in the South, which only provides 38% of the water yield (Water yield is the volume of freshwater resources for a defined geographic area and time period. It is an estimate of the amount of renewable water and non-renewable glacial melt) [ 12 ]. In the North, the per capita water yield is 98 times greater than in the South [ 12 ]. The water yield in the Prairies is a fraction (3%–12%) of that in the Great Lakes, Maritime Coastal or Pacific Coastal drainage regions and decreased between 1971 and 2004 [ 12 ]. During the 20th century, meteorological measurements have shown significant warming during all seasons but the greatest rates have been during winter and spring. They have also revealed general increases in precipitation, although considerable decadal-scale variability is present over many regions of the country. For example, Zhang et al. [ 13 ] determined that mean annual air temperature has increased by an average of 0.9 °C over southern Canada between 1900 and 1998 with the greatest rates seen in the West during the winter and spring (southern Canada broadly refers to south of 60N; an exception is [ 314 ] and mention in [ 315 ] where southern Canada refers to south of 55N. See Figure 5 in [ 4 ] for additional details. For Statistics Canada discussions, see [ 12 ] for a description of standard drainage area classifications). Subsequent analysis extending the target period to 2010 confirmed the warming has continued and even accelerated with an increase in the mean annual temperature of 1.5°C for southern Canada over this longer period [ 16 ]. Zhang et al. [ 13 ] also found that annual total precipitation increased significantly from 1990–1998 over most of southern Canada with the exception of southern Alberta and Saskatchewan. Analysis expanded to the period 1900 to 2009 also found an 8.7% increase in annual rainfall for southern Canada [ 17 ]. Snowfall mainly increased in northern Canada while at the same time decreasing significantly in the Southwest for the period 1950–2009 [ 17 ]. An analysis of changes in extreme temperature indices over southern Canada from 1900–2003 and the entire country from 1950–2003 found that over most of the country, there was an increase in the number of warm days where the maximum temperature (>90th percentile) and the number of warm summer days (maximum temperature >25°C) during the last half century [ 18 ]. The number of hot days (>30°C) and summer warm days (>90th percentile) also increased over all of Canada, but were only significant in southwestern regions [ 19 ]. The analysis revealed little change in the number of summer warm spells. An examination of changes in extreme temperature and precipitation indices over southern Canada from 1900–2003 and the entire country from 1950–2003 found a decrease in the consecutive number of days with no measurable precipitation in British Columbia and Atlantic Canada. Trends in drought occurrence and severity in Canada are often monitored using meteorological drought indices, such as the Standardized Precipitation Index (SPI), the Palmer Drought Severity Index (PDSI), and the Standardized Precipitation Evapotranspiration Index (SPEI) [ 4 ]. These indices are designed to assess regional precipitation anomalies (the SPI), or a region's water balance using both precipitation and temperature data (PDSI and SPEI), to account for evapotranspiration effects. Results from trend investigations incorporating these various drought indices have shown variable results over Canada, often being dependent on the time period used in the trend calculation (mainly due to the aforementioned decadal-scale variability in precipitation). However, an increase in drought (i.e., a decreasing Palmer Drought Severity Index (PDSI) trend) over most of the country was observed between 1950 and 2002 [ 4 ]. Western Canada, notably the southern Canadian Prairies and interior valleys of British Columbia, are more susceptible to drought because they lie in the lee of major mountain ranges and, as a result, generally receive little precipitation but with high variability [ 20 ]. During the past 100+ years, several long-duration droughts have occurred in western Canada. The Canadian Prairies have experienced large-area, multi-year dry episodes during the 1890s, 1930s, late 1950s/early 1960s, 1980s and more recently from 1999–2005 [ 21 , 22 ]. During the notable drought period from 2001–2002, below average stream flows were reported across the country extending from western Canada to Ontario and the Atlantic Provinces [ 23 ]. Water levels of most closed-basin lakes have declined throughout the 20th century across southcentral Alberta and Saskatchewan in western Canada, although levels have increased in the past few years [ 24 ]. Farther east, droughts do occur in southern Ontario and Quebec, although these events are usually shorter, smaller in area, less frequent and less intense [ 20 ]. Nonetheless, there have been major drought occurrences in the East during the last century [ 25 ] resulting in severe impacts, including low water levels in the Great Lakes [ 26 ]. Between 1920 and 1999, southern Ontario experienced major droughts in 1930, 1933, 1934, 1936, 1963, 1998, and 1999 [ 25 ]. Ontario was also impacted by drought in 2001–2002 [ 23 ]. Great Lakes' water levels have shown considerable decadal-scale variability during the 20th century with no evidence of any long-term trend. However, lower Great Lakes' water levels have coincided with the droughts of the 1930s, early 1960s and from early 2000 to the present [ 20 ].In 2011, water levels in the Great Lakes–St. Lawrence Rivers system reached record low levels and some activities/sectors in Ontario watersheds were asked to reduce water consumption under the Ontario Low Water Response Plan (OLWR)[ 23 ]. Déry et al. [ 27 ] found that freshwater discharge rates decreased over a period of 37 years for the majority of key rivers feeding the Hudson, James and Ungava Bays. This watershed is fed by water from five provinces, the territory of Nunavut andfour U.S. states, covering an area that is equivalent to one-third of Canada [ 27 ]. Along the East Coast, droughts in the Atlantic provinces are infrequent. However, a side effect of their rarity is that these regions have lower adaptive capacity to deal with droughts, making the region more susceptible to drought impacts [ 28 ]. The Atlantic provinces were affected by the drought of 2001–2002, with river flows in the Atlantic Provinces dropping to a 20-year low [ 23 ]. An analysis of meteorological indices (southern Canada 1900–2003; entire country 1950–2003) indicated a decrease in the number of days with no measurable precipitation in Atlantic Canada [ 18 ]. The context surrounding this shift will be discussed further below. 1.3. Future Droughts All Global Climate Models (GCMs) project future increases of summer continental interior drying and associated risk of droughts. The greater overall risk is ascribed to further increases in temperature and resultant potential evapotranspiration not being offset by precipitation increases [ 3 ]. In fact, the recent Fifth Assessment Report from the IPCC [ 29 ] stated that an increase in the intensity and duration of future droughts on a regional to global scale is likely ( i.e. , "medium" confidence) by the end of the 21st century. This statement especially applies to southern Europe and the Mediterranean region, central Europe, central North America, Central America and Mexico, northeast Brazil, and southern Africa. Elsewhere there is overall "low" confidence because of inconsistent projections of drought changes (dependent both on the GCM model and dryness index). Definitional issues, lack of observational data, and the inability of models to include all the factors that influence droughts preclude stronger confidence than "medium" in drought projections (see [ 30 ]). Some global studies have indicated that droughts will be slightly more frequent and longer in duration as compared to present-day conditions by the second half of the 21st century [ 31 , 32 ]. Specifically for North America, Trenberth [ 33 ], using several GCM simulations, showed that future warming will increase both evaporation and surface drying, thereby leading to longer and more intense drought particularly in the summer over continental interior regions of the continent. A recent study found that the southwest and central plains regions in the continental U.S. will be significantly drier by the end of the century when compared with historical periods. This was the case even across various GCMs, and different metrics of soil moisture in both moderate and high greenhouse gas emission scenarios [ 34 ]. Overall, future precipitation/water resource projections suggest that in Canada, the North will become wetter and the South drier especially in summer [ 35 ]. However, limited research has been carried out regarding potential future drought occurrence across Canada and most investigations have focused on the particularly drought-prone Canadian Prairies. A first-order assessment of future drought occurrence over southern Canada (2041–2070) by Bonsal and Regier [ 14 ] found future annual precipitation will increase over most of the southern region of Canada but this will be more than offset by increased temperatures that lead to higher evaporation so increases in drought are likely. The future mean PDSI estimates indicate the prevalence of more frequent drought for the southern half of the Canadian Prairies in the latter half of the 21stcentury with persistent negative PDSI values expected after 2040 [ 15 ]. This suggests the possibility of a regime shift to a more arid climate in the southern Prairies due to the effects of projected summer temperature increases. Future projections of PDSI also revealed that multi-year droughts lasting 10 or more years are more likely than what has been historically observed. Projections suggest approximately three such events per 100 years as compared to the approximately once every 100 years that has been observed. Similar results for the southern Canadian Prairies were also documented by PaiMazumder et al. [ 36 ] using simulations from the Canadian Regional Climate Model for the period 2041–2070. Although published studies for the rest of Canada are limited, the authors noted that in more northern areas of the Prairies, drought events are projected to be less severe and less frequent. Considerable uncertainty exists with respect to future precipitation events, particularly on a regional and intra-seasonal basis. Bonsal et al. [ 15 ] found considerable range in projections in their assessment of changes to the temporal and spatial characteristics of future (2011–2100) PDSI and SPI using statistically downscaled daily temperature and precipitation data over the southern Canadian Prairies, from three GCMs with different greenhouse gas emission scenarios. In summary, although there are no comprehensive analyses of future drought events across all of Canada to provide detailed regional and seasonal drought information, both global investigations and a handful of studies conducted for the Canadian Prairies suggest that future droughts will likely increase, particularly in those regions that already experience these events ( i.e. , more southern and interior areas of the country). 2. Experimental Section Canadian and U.S. literature was searched using PubMed, SCOPUS and Cochrane Collection databases. The following search terms were included: drought, arid * ( * = the wildcard search character), desiccation, dry spell, water AND (shortage * OR scarcity), (low OR drop OR reduc * ) AND (water AND (level OR table)), Canada, Canadian, health. Articles published over the 20 years from 1993 to 2013 were considered for the initial search. Articles were screened by scanning titles and/or abstracts. Full articles were scanned where abstracts were missing or unclear. Articles were included where the linkages to human health and drought were made explicit. Articles from reference lists and/or in-text citations meeting these criteria, as well as from climate change and vector-borne disease expert-provided sources (peer-reviewed and grey literature) were also included. The international literature was searched using the OVID Medline and Web of Knowledge databases. The following search terms were included: climate, climate change, health, public health, drought, adaptation, response. Articles published at any time up to December 2013 were considered. Relevant websites, such as the World Health Organization, IPCC, United Nations Framework Convention on Climate Change were also searched using the same key words. Where appropriate, the entire article was reviewed for more detailed information. Articles were included if they discussed drought and human health effects and made explicit the link between the two, or if they discussed potential adaptations or responses to drought conditions. They were also included if the study was not limited to the examination of drought in a developing country/region and/or the results were applicable to the Canadian context. A number of articles were excluded given their limited relevance to the Canadian context including those focused on malnutrition, or infectious diseases that are rare in Canada, such as malaria, chikungunya, schistosoma and tick-borne relapsing fever [ 37 , 38 , 39 , 40 ]. Vector-borne disease expert articles were also included. The review of the Canadian and U.S. literature identified 64 records and the international literature yielded 72 records (duplicates were removed). This total (136) included 22 publications identified by climate change or vector-borne disease experts and 59 records identified within the in-text citations and references from the sources mentioned above. Broadly, evidence from developed countries and particularly from the U.S., Australia and the UK was considered to have greater potential relevancy than research focused on developing countries for this review, given their greater similarity to the Canadian context and of their populations in aspects that could modify health impacts related to drought (e.g., lifestyle, infrastructure and public services). 3. Results 3.1. Health Risks Associated with Drought Drought can affect human health through a broad range of pathways including those shown in Figure 1 . They are discussed in greater detail in the following sections. Generally speaking, these broadly fall into environmental services and socioeconomic areas. There is agreement that globally, drought leads to an increase in morbidity and mortality [ 41 , 42 , 43 , 44 ]. Although discussions have pointed to direct drought-related deaths in low-income countries [ 45 ], a broad range of public health impacts associated with drought has been identified in the U.S. [ 46 , 47 , 48 ]. Research also indicates that climate change may impact drought-related health outcomes in Canada [ 2 , 49 , 50 ]. In fact, some evidence of drought-related health impacts has been identified for Canadians living on the Prairies [ 51 ]. For example, there is some evidence of respiratory impacts related to dust and increased risk of water-borne disease [ 51 ]. Furthermore, reviews of impacts from Ontario [ 52 ] and the U.S. [ 53 ] have identified climate change and future drought as having the potential to affect food-, vector- and water-borne diseases [ 54 , 55 ]. 3.2. Water Quality and Human Health 3.2.1. Source Water for Drinking Water Drought can negatively impact the quality of drinking water [ 2 , 23 , 46 , 50 ]. Decreases in drinking water quality can, in turn, lead to water-related disease [ 42 , 56 , 57 , 58 , 59 , 60 ] and the exacerbation of heat stress [ 61 ]. The pattern of reduced water levels, stream flow and resultant stagnation during drought can increase contaminant concentrations in ground and surface waters [ 46 , 51 , 62 , 63 ]. For example, measurements taken during a drought in 2003 over the Rhine and Meuse rivers in Europe showed not only a substantial reduction in river discharge relative to the year before, but also that chloride concentrations were inversely related to the flow (m 3 /s) of these rivers [ 64 ]. This effect is thought to increase when contaminants continue to be added to the system [ 64 ]. Lower water levels can also facilitate an increase in reactions between the water, contaminants and sediments [ 62 ]. This can be further reinforced by greater soil erosion during drought [ 23 ]. In the U.S., elevated nitrate, orthophosphates, chlorides and sulfates in groundwater have also been associated with drought, in some cases exceeding the U.S. EPA's Maximum Contaminant Level [ 63 ]. Frequent detection in wells near agriculture and septic systems suggested these systems were at risk of contamination [ 63 ]. Baures et al. [ 65 ] found that variation in total organic carbon and nitrate was associated with changes to river flow in France, with a high ratio of total organic carbon to nitrate concentrations under conditions of very low river flow. A UK study found that drought can lead to greater contamination of drinking water source water from dissolved organic carbons and can increase difficulty in contaminant removal [ 66 ]. Similar patterns have been reported in Canada. Elevated nutrient concentrations ( i.e. , phosphorous and nitrogen) have been observed in association with low water periods in Alberta [ 67 ]. Figure 1 Pathways through which drought impacts human health in the context of climate change. Nutrient-loading is known to lead to eutrophication and is associated with cyanobacterial (algal) bloom development in all provinces [ 68 ]. Some cyanobacteria produce toxins that are harmful to humans and blooms can impact water bodies used as sources of drinking water and for recreational water activities [ 2 ]. In Nebraska (U.S.), for example, drought was considered a contributing factor to a cyanobacteria bloom in 2004 that resulted in reports to public health of impacts such as gastrointestinal illness and the issuance of health alerts and advisories for nearly 100 lakes [ 69 ]. However, Barbeau et al. [ 70 ] carried out a hypothetical analysis of the vulnerability of Quebec's provincial drinking water systems to toxic cyanobacteria and their results suggest that existing systems in that province would be able to cope with this risk, even within the context of climate change. In Ontario, there was a significant increase in algal blooms between 1994 and 2009 [ 71 ]. Climate change may lead to changes in the frequency and duration of algal blooms, as well as increases in the amount of toxins they produce or an increase in the prevalence of toxin-producing species overall in temperate regions [ 70 , 71 ]. Warmer temperatures anticipated with climate change may promote algal blooms through mechanisms such as reducing the mixing of water, allowing algae to rise to the surface more quickly and by increasing the salinity of freshwater bodies [ 72 ]. The warmer temperatures associated with drought and decreases in water flow and/or volume may promote the survival and growth of pathogens in water, including those responsible for infectious disease [ 46 , 55 , 73 ]. Reduced water flow can lead to an increase in bacterial concentration and reduced water quality as there is less water available to dilute these contaminants [ 64 , 74 ]. Climate change-related impacts on water, including drought, are recognized as leading to greater risks from water-borne diseases such as hepatitis A [ 52 ]. However, there appears to be limited discussion of the relationship between drought and water-borne pathogens and the interactions between climate change and water-borne diseases are complex. In the U.S., the survival and multiplication of bacteria leading to leptospirosis outbreaks may have been altered by temperature and pH changes resulting from the stagnation associated with drought [ 75 ]. Similarly, although Naegleria fowleri is rare, increased water temperature during drought can allow for the amoeba to grow in greater concentrations [ 46 ]. Drought conditions may concentrate pathogens while at the same time inactivate them (e.g., Cryptosporidium oocysts) [ 54 ]. In Canada, it is important to note that the greatest burden from enteric, food and water-borne diseases, drawing on 2012 data, is associated with campylobacteriosis (10,174 cases), followed by salmonellosis (6828 cases) and giardiasis (3862 cases). Other reportable diseases in this category have a far smaller number of cases, such as shigellosis with 1,068 reported cases and approximately 500 cases each of E. coli and cryptosporidiosis in 2012 [ 76 ].In addition, while at its peak in 1970, there were 12,283 cases of hepatitis A reported in Canada; this has since fallen to approximately 200–300 cases per year [ 76 ]. Future changes to the frequency and/or severity of drought could modify the burden associated with these diseases in Canada. In coastal areas, drought can result in the contamination of groundwater drinking water sources by the intrusion of salt water [ 46 , 50 ]. The mingling of saline and fresh groundwater that occurs in such regions can progress inland affecting freshwater wells, under a number of circumstances, including drier periods [ 77 ]. Saltwater intrusion can be exacerbated by climate change and sea-level rise [ 50 , 77 ]. The combination of drought with extreme precipitation may also impact water-, food- and vector- borne diseases. The most commonly cited impact of climate change on water-borne disease is the anticipated effects of increasing frequencies and amounts of heavy rainfall impacting the likelihood that untreated and pathogen-carrying water is consumed by Canadians [ 52 ]. Drought can amplify the impacts of a subsequent extreme rainfall event. During a drought event the soil can become compacted, thereby increasing runoff and the likelihood of water contamination and water-borne illness from precipitation events that follow [ 46 , 54 , 55 , 78 ]. Conditions associated with drought, such as low rainfall, can also concentrate Giardia and Cryptosporidium cysts in groundwater sources and where water is stored. Water sources may then be contaminated by the cysts after rainfall events [ 79 ]. Evidence from the Prairies suggests that the prevalence of livestock farming may affect water-borne diseases that follow such events if manure is able to contaminate water sources (e.g., Eschericia (E). coli , Cryptosporidium , Giardia ) [ 80 , 81 , 82 , 83 ]. Campylobacteriosis outbreaks are also frequently associated with heavy rainfall, often when they follow a period of drought [ 54 ]. Although a much smaller risk for Canadians, with typically only 0–2 cases in Canada each year [ 76 ], the pattern of drought conditions followed by a wet spring could also activate anthrax [ 52 ]. 3.2.2. Treated Drinking Water During drought, there can be a drop in pressure at water treatment plants [ 52 ], which can increase turbidity leading to a greater potential risk of water contamination and consequently a greater need for water treatment [ 84 , 85 ]. The impact of drought on contaminant levels in source water has been discussed above. In the event that well water becomes unavailable during drought, turning to alternative water sources can also increase risks to health if appropriate safeguards for those treatment measures are not in place [ 86 ]. Inadequate water treatment is important as pathogens, including bacteria and viruses, are known to be taken up by particles in water [ 87 ]. Costs also increase with the additional treatment required to address higher levels of turbidity [ 88 , 89 ]. Changes to water flow and levels associated with climate change could further contribute to stress on water treatment systems and increase the cost of treatment [ 50 ]. In northern Ontario, some First Nations communities may already be near the limit of their capacity in this regard [ 50 ]. 3.2.3. Recreational Water Drought can also affect the quantity and quality of recreational water [ 59 , 60 , 61 , 65 , 90 , 91 , 92 , 93 , 94 ]. The hot and dry weather that can be associated with drought is thought to increase recreational water use [ 46 , 75 ]. At the same time, the increased concentration of pathogens encouraged under these conditions, as described above, are thought to increase the likelihood of infection among recreational water users [ 46 ]. Although leptospirosis cases and outbreaks are frequently associated with heavy rainfall events [ 57 ], in the U.S., they have also been associated with drought and with swimming in lakes [ 75 ]. In Canada, even under typical (non-drought) conditions, acute gastrointestinal illness (AGI) is a risk for swimmers, with the risk of AGI at approximately 3%–8% [ 95 ]. Although the ingestion of water during recreational use would be lower than that from the drinking water sources discussed above, recreational water users could be exposed to water-borne pathogens that lead to infection through direct contact or inhalation [ 96 , 97 ]. 3.3. Sanitation/Hygiene Water plays a key role in maintaining hygiene, which is directly associated with preventing disease [ 46 , 60 , 98 ]. Health care facilities may depend heavily on water to protect the health of patients and workers [ 46 ]. Studies outside of Canada suggest strong linkages between changes in precipitation, drought and diarrheal disease [ 99 , 100 ]. The World Health Organization has estimated that globally, the risk factor "water, sanitation and hygiene" accounted for 4% of all deaths [ 101 ]. A global cross-sectional study found the incidence of diarrhoea in children under the age of five years increased 4% with each 10mm per month decrease in precipitation [ 102 ]. Modeling done on a global scale by Motoshita et al. [ 103 ] based on recent historical data suggested that health damage from infectious diseases resulting from domestic water scarcity would be relatively low in the U.S. compared with other regions, although the health damage could not be modeled for Canada. However, in England and Wales, both low and excessive rainfall over the short-term was linked to a rise in the incidence of diarrhoea related to Giardia , Cryptosporidium , E.coli , S. typhi , S. paratyphi , Campylobacter and Streptobacillus moniliformis (This is not a reportable disease in Canada and infections are thought to be lower in Canada and other Western countries than in the US; close to 200 cases of "rat bite fever" from Streptobacillus moniliformis had been documented in the U.S. at the time of a 2007 publication [ 333 ]) [ 104 ]. Furthermore, people displaced by drought are considered likely to experience loss of sanitation and routine hygiene [ 45 ]. Human behavior may play a role, since the relaxing of hygiene practices could lead to an increase in diarrheal illnesses during drought [ 74 ]. As described above, Giardia , Cryptosporidium , E. coli and Salmonella cause a considerable number of disease cases in Canada each year. Alteration of the rates of such diseases by drought could have implications for public health. 3.4. Food-borne Diseases The conditions associated with drought could impact food-borne illness. Hot and dry conditions often associated with drought are thought to favour the proliferation of some food-borne pathogens. The impact in Canada could be significant, as currently there are an estimated 11 million cases of food-borne illness annually across the country [ 105 ]. Among reportable diseases, 10,174 cases of campylobacteriosis and 6,828 cases of salmonellosis were reported in Canada in 2012 [ 76 ]. Illness caused by Clostridium perfringens is not nationally notifiable in Canada, but ranks among the most common food-borne diseases in the industrialized world [ 106 ]; in the U.S., C. perfringens is thought to cause close to one million cases of food-borne illness each year [ 107 ]. In some cases the environmental conditions created by drought are favorable to pathogens, such as C. perfringens , which does well in hot and dry conditions [ 52 ]. Campylobacterios is also expected to benefit from some seasonal conditions under climate change in the future [ 54 ]. However, the expected impacts of climate change-related drought conditions remain uncertain for other pathogens, including Salmonella , Listeria and Norovirus [ 54 ]. For example, under experimental desiccation Salmonella (typhimurium) lost infectivity and cultivability [ 108 ]. Meanwhile, in contrast to earlier studies, Zhang et al. [ 109 ], in comparing models to forecast salmonellosis cases in Adelaide (Australia) found decreased rainfall was associated with salmonellosis cases. It is possible that this finding was due to interaction with other variables such as temperature or the unique local climate [ 109 ]. For example, Ge et al. [ 110 ] suggest that under extreme conditions, including drought, when the soil contains high concentrations of Salmonella there may be greater internalization of the pathogen in lettuce relative to optimal irrigation conditions. During drought, cultivated crops, fish and shellfish consumed for food [ 52 ] may pose an increased risk of disease, which could have implications for human health [ 46 ]. If treated municipal sewage is used by the agricultural sector for irrigation or to process products due to water shortages there can be an increase in the risk of food-borne illness (e.g., Salmonella , E. coli ) [ 46 ]. Local food production practices and food imported from some countries may need to be monitored more closely under such conditions [ 46 ]. 3.5. Vector-Borne Diseases 3.5.1. West Nile and St. Louis Encephalitis Vector-borne diseases—particularly mosquito-borne ones—are frequently associated with increased precipitation [ 111 ] because standing water is needed for the development of immature stages of mosquitoes. Mosquitoes are important disease vectors in Canada as they can transmit West Nile virus (WNV). There were 110 total clinical cases of West Nile virus reported in Canada between January and the end of November, 2013, including three cases potentially related to travel outside the province where they were reported [ 112 ]. High precipitation in the prairie provinces in 2007 was thought responsible for an unprecedented abundance of Culex tarsalis mosquitoes that in turn led, later that year, to the largest ever outbreak of WNV in Canada (2,401 cases) [ 76 , 113 , 114 ]. Studies on the mosquitoes Cx. pipiens and Cx. restuans , which are the main WNV vectors in eastern Canada, also suggest that precipitation is positively associated with their abundance [ 115 ]. However, in some circumstances periods of drought have been directly associated with increased risk from mosquito-borne infections, such as WNV in southeastern, northeastern, central and western U.S., the Canadian prairies, as well as in Europe [ 115 , 116 , 117 , 118 , 119 , 120 , 121 , 122 ]. A number of explanations have been proposed for these observations, including impacts of drought on competitors and predators of WNV mosquito vectors [ 116 ] which can lead to mosquito population increases [ 115 ] and changes in communities and densities of the bird species that are reservoir hosts of WNV, resulting in increased WNV prevalence in mosquitoes [ 123 , 124 ]. Semi-permanent rather than permanent wetlands are thought to be particularly prone to the effects of drought given the fluctuations they experience in response to precipitation. Irrigation during times of drought may create new breeding sites for WNV mosquito vectors and thus shift the local geographic focus of risk from WNV [ 125 ]. In the matrix of mosquito breeding habitat found in urban and suburban environments, small numbers of locations can contribute to mosquito breeding and abundance and these are frequently man-made water courses, sewage overflow systems and holding pools, etc . [ 126 ], which are likely to continue to hold water during drought events. Associations of WNV incidence with droughts or low precipitation (e.g., in Virginia, U.S. [ 127 ]) may be due to reduced flushing of mosquito larvae and eggs from breeding sites in urban areas during rainfall; this flushing reduces mosquito abundance [ 128 ]. Frequently, the effect of drought on vector-borne diseases is seen when drought is followed by periods of wetter weather [ 129 , 130 ], but the relationships amongst mosquito abundance, infection prevalence, WNV risk and seasonal variations in drought and rainfall can be complex, geographically variable, habitat/environment-specific and difficult to predict [ 131 ]. In addition to altering the mosquito vectors' environment, drought can impact risk from mosquito-borne diseases by affecting contact rates between the vectors and reservoir hosts, thereby influencing the prevalence of infection in mosquitoes [ 46 , 118 ]. Evidence for this mechanism comes particularly from studies on mosquito-bird transmitted viruses [ 118 ] including WNV [ 129 ] and St. Louis encephalitis (SLE) virus [ 132 ].The congregation of reservoir host birds around dwindling water sources may be one way that drought affects WNV infection prevalence in mosquitoes by increasing rates of contact between avian reservoirs and mosquitoes thereby amplifying transmission [ 129 , 132 ]. In California (U.S.) drought may have contributed to an increase in human WNV cases by reducing populations of non-competent host birds, allowing for greater contact of mosquitoes with reservoir-competent host birds and possibly humans [ 131 ]. In the southeastern U.S., SLE outbreaks are associated with rainfall following drought, which is largely due to the behavior of the main vector Culex nigripalpis [ 133 ]. In drought conditions, gravid Culex nigripalpis females will increase in numbers around standing water as they wait for rainfall to feed and lay their eggs together, resulting in synchronous production of mosquitoes [ 134 ]. In addition, enhanced transmission due to the concentrating of mosquitoes and bird reservoirs in habitats that hold water is thought to help amplify the SLE virus [ 132 , 135 ]. 3.5.2. Lyme Disease Lyme disease, the most frequent vector-borne disease diagnosed in North America, is emerging in Canada [ 136 , 137 ]. Drought has a consistently negative effect on risk from tick-borne diseases such as Lyme disease because of (i) direct killing of ticks by desiccation in severe drought; (ii) reduction in host-finding success by reducing the time that ticks can spend seeking hosts outside of moisture-holding refugia in the habitats in which they occur; and (iii) increased energy expenditure due to ticks returning more frequently to moisture-holding refugia to rehydrate [ 138 , 139 ]. 3.5.3. Directly Transmitted Wildlife-borne Zoonoses The effect of drought on zoonoses transmitted directly from wildlife to humans is likely a reduction in the abundance of wild animal hosts and consequently, reduced pathogen transmission. When drought is followed by rainfall, wild animal host densities may increase again, but as most individuals will be naïve of infection, the potential for epidemics exists. Rodent-borne hantavirus is a paradigm for this in which drought followed by heavy rainfall (particularly associated with El Nino) is thought to drive epidemics in the southwestern U.S. [ 53 , 140 , 141 ]. 3.6. Fungal Diseases Drought may encourage the growth of various fungi associated with health impacts. Warmer, drier summers may have contributed to the establishment of Cryptococcus gattii in Canada [ 142 ]. Low moisture has been associated with C. gattii soil colonization [ 143 ]. The fungus was first detected on Vancouver Island in 1999, with more than 100 human cases identified [ 142 ]. Between 1999 and 2007, there were 218 reported cases of C. gattii in British Columbia, the region with the largest C. gattii infected population in Canada and the world [ 144 ]. Drought may also positively affect the growth and dissemination of coccidioidomycosis (Valley fever). Rainfall patterns and timing (including length of the drought) have been identified as modifiers of the incidence of this disease [ 46 , 142 , 145 , 146 , 147 , 148 ]. Although the fungus is concentrated in the Southwestern U.S., recent cases that appeared to originate in Washington State (U.S.) suggest that it may be capable of moving to other areas [ 149 ]. Valley fever is presently rare in Canada but climate change may impact the incidence of the disease among Canadian travelers [ 150 ]. Since 1952, there have been a total of 128 cases of coccidiodomycosis in Canada. These Canadian cases were acquired during travel outside Canada [ 150 ]. Although associated with warming rather than drought, projections of increased precipitation in the winter months followed by reduced precipitation in the summer months for North America are thought to favor the dispersal of Blastomyces dermatitidis spores, which causes blastomycosis [ 142 ]. Currently, this fungus occurs mainly in northwestern Ontario [ 142 ]. Between 1994 and 2003, there were 309 cases of blastomycosis in Ontario with more than half of the cases occurring between 2001 and 2003 [ 151 ]. Geographically, northern Ontario accounted for 61% of the cases, followed by 21% from the Toronto region [ 151 ]. 3.7. Respiratory Health 3.7.1. Dust, Particulates and Allergens Drought is often accompanied by dry, dusty conditions and dust storms, which can impact health [ 83 , 152 , 153 ]. Climate change can lead to an increase in fine particulate matter, allergen and dust concentrations in the air in drought-prone areas [ 154 ] which can have significant health impacts. A study of a lake desiccated through persistent drought in Saskatchewan found a greater prevalence of coughing, wheezing and eye and nasal irritation among residents living near the lake than among those in a control group [ 155 ]. PM 10 (air particle) increases during and following dust storms have also been found to cause small increases in emergency room visits for bronchitis and sinusitis in the U.S. [ 156 ] although dust storms were not associated with an increased risk of mortality [ 157 ]. Air-borne toxins from algal blooms, which can be increased due to drought conditions, can also irritate the respiratory system, nose and eyes [ 46 , 153 , 155 ]. It has been anecdotally suggested that there were numerous deaths from "dust pneumonia" related to dust inhalation during the drought of the 1930s in the U.S. [ 152 ]. In Ontario, Charron et al. [ 52 ] noted the possibility that drought with strong winds could lead to the transmission of Rickettsial diseases to humans from infected livestock. 3.7.2. Silo-Filler's Disease Dry conditions can increase nitrate levels in corn plants and encourage the accumulation of nitrogen dioxide (NO 2 ), thereby elevating the risk of "silo-filler's disease" [ 158 ]. A greater number of "silo-filler's disease" cases were reported in New York State after a dry growing season in 1995 [ 158 ]. There are approximately five cases of silo-filler's disease per 100,000 silo-associated farm workers in New York State annually [ 159 ]. 3.7.3. Wildfires and Extreme Heat Drought has been associated with wildfires in both Canada and the U.S. and is expected to increase the number of wildfires in the Prairies specifically [ 46 , 51 , 83 ]. The international literature links wildfires to respiratory, cardiovascular, ophthalmic and psychiatric illnesses [ 43 , 160 , 161 ]. In Brazil, a study of young children linked an increase in respiratory disease incidence with drought-related wildfires [ 162 ]. During a prolonged forest fire in British Columbia, Canada in 2003 one community experienced a 46%–78% increase in physician visits for respiratory diseases [ 163 ]. Wildfires and other extreme weather events that might lead to displacement can be a source of significant stress for those directly affected [ 51 , 80 , 83 , 160 , 164 ]. The significant loss of property and infrastructure during a wildfire can also have negative impacts on mental health [ 51 ] and may lead to post-traumatic stress disorder [ 160 , 164 ]. More recently, in 2011 the northern Prairies experienced severe drought and catastrophic fires during which Slave Lake, Alberta lost one-third of homes and businesses to wildfire [ 165 , 166 ].In the Northwest Territories, a dry winter combined with only half the typical amount of rainfall over the summer of 2014 and above average temperatures, was accompanied by the worst fire season in 30 years [ 167 ]. Record warm and dry conditions in British Columbia over the same summer and preceding winter also set the stage for a wildfire season that saw more than seven and a half times the 20 year average in land area burned in 2014 [ 167 ]. Wildfires are expected to become more frequent and severe in many parts of Canada with climate change [ 168 ]. Further, drought has been associated with periods of extreme heat which can present a range of health impacts that need to be addressed. In addition to its direct impact on human health, extreme heat can add to drought-related stress on agriculture with knock-on effects discussed below. For example, an extreme heat event that took place in Russia over July and August, 2010 reduced the country's grain yield by 25%, resulting in increases in food prices and losses estimated at $15 billion [ 169 ]. 3.8. Injuries Although severe injuries directly resulting from drought may be unlikely in mid and higher income countries [ 45 ], during periods of drought an increased incidence of spinal cord injuries has been recorded in the U.S. from diving into shallow water bodies. Accidents from debris can also be a concern as lower water levels can expose previously "hidden" hazards in bodies of water [ 46 , 170 , 171 ]. Affected individuals may have been seeking relief from warm temperatures accompanying dry conditions [ 171 ]. Although carelessness and alcohol may play a role, injured individuals were often familiar with local bodies of water and believed the water to be deeper than it actually was [ 170 , 171 ]. Dust storms, discussed above in the context of drought, have also been associated with traffic accidents on the Prairies [ 23 , 153 , 172 ]. Dust storms can reduce visibility for drivers [ 172 ]. 3.9. Food/Water Insecurity Drought may increase the difficulty that individuals face in obtaining adequate water. In the U.S. there have been reports of risk of drinking water shortages [ 173 ]. In Canada, there have been reports of costly water-related infrastructure such as water cisterns, septic tanks and wells damaged or at risk during drought [ 80 , 174 , 175 , 176 ]. The IPCC recognizes a broad range of impacts due to climate change and the altered frequencies of associated extreme events, such as drought and floods that can have negative effects on crop production [ 177 , 178 , 179 ]. Climate change can also alter the geographic distribution and productivity of crops and livestock and the distribution of pests and disease [ 180 , 181 , 182 ]. Food security can be affected through drought-related decreases in agricultural output, food shortages and higher food prices. Crop failures and reduced crop yields are of particular concern because they can have immediate and direct effects on both local and global food security [ 183 , 184 , 185 , 186 , 187 ].Climate change could lead to a decrease in the zinc and iron content of staple crops such as wheat [ 188 ]. Such a change in nutritional content could exacerbate iron deficiency, which occurs in both developed and developing countries [ 189 ]. Meanwhile, increasing homogeneity among crop species globally threatens food security with increased interdependence of food systems around the world and reduced genetic diversity [ 190 ].As a net exporter of food and the 9th largest exporter in the world [ 191 ], Canada is unlikely to experience the types of food shortages that could occur in other countries. However, food insecurity may be reflected in food price volatility caused by international market shortages [ 192 ], which could impact low income Canadians who find it difficult to pay higher prices for nutritious food. In addition, some individuals may be forced to use their resources to purchase water instead of food [ 46 , 49 ]. The media reported a surge in food bank users and pointed to high household spending on water in parts of California (U.S.) during their recent drought [ 193 ]. The widespread U.S. drought in 2012 was thought to be one of the factors contributing to an increase in food prices in countries, including Canada [ 194 , 195 ].In addition, threats from climate change to the food security of First Nations and Inuit populations in northern Canada are increasing [ 2 ]. While it appears to be complex, a recent Australian study suggests that drought may play a moderating role in the relationship between food insecurity and psychological distress, possibly increasing distress [ 196 ]. For those who rely on subsistence fishing, such as Aboriginal populations, health can be affected when drought affects the health of aquatic organisms [ 46 ]. Drought can increase the likelihood of disease spread and/or exposure to an accumulation of contaminants in the water for these aquatic organisms [ 46 ]. They can also be impacted by decreases in the quality and quantity of water during drought or through contamination introduced by the rains that follow drought [ 52 ]. 3.10. Socioeconomic Pathways Leading to Human Health Impacts Drought can affect social and economic wellbeing within a community through changes to water availability that can impact agriculture [ 176 ]. The Canadian agricultural sector produced crop and livestock products totaling over $41 billion in 2010 [ 197 ]. Costs associated with a drought can be significant. For example, total costs of drought across the Prairies in 1990 exceeded $800 million (equivalent 2011 Canadian dollars) [ 198 ]. The more recent drought of 2001–2002 was responsible for $3.6 billion in direct losses related to agricultural production [ 199 ]. Impacts on agriculture are passed on to the broader community and may lead to temporary unemployment [ 200 ] and increase stress levels [ 176 , 201 ]. The U.S. drought in the 1930s and its impact on agriculture contributed to the Great Depression; there was high unemployment, increased demands on government aid programs and community impacts as millions of people affected by the drought moved to other communities [ 202 ]. The impact of drought on different segments of a society can vary [ 203 ]. For example, the 2001–2002 drought brought significant impacts on agricultural production in Alberta's "Special Areas", but it was mainly an inconvenience for residents and businesses outside the agricultural sector [ 203 ]. Similarly, the 1987–1992 drought in California (U.S.) led to primarily behavioral rather than economic impacts for residents [ 200 ]. Loss of agricultural output in one area may drive up the demand and prices for farmers unaffected by the drought suggesting that some economic impacts may be highly localized [ 204 ]. Globally, the impact of climate change on agriculture in different regions and countries, such as other developed grain-exporting countries can affect the competitiveness of Canada's agricultural sector, particularly for wheat [ 174 ]. A widespread U.S. drought in 2012 increased crop prices, resulting in new record highs for Canadian farm incomes overall [ 205 ]. The international literature suggests that drought can impact communities by shifting resources from services such as public health to address more immediate needs (e.g., sanitation), as well as leading to conflict over water between groups and regions [ 2 , 42 , 59 , 90 , 206 , 207 , 208 ]. Recent media attention around conflicts over water in the western U.S. states among farmers, activists, developers and governments at state and local levels suggests that conflict over water in developed countries such as Canada could be a concern [ 209 ]. During periods of drought, water-related recreational activities may be cancelled or closed, such as during the 2001–2002 drought in the Prairies and the 1987–1992 drought in the U.S., which can have social impacts [ 174 , 176 , 200 ]. In the Canadian context, data from 2005 show that over 90% of the water withdrawn in the country was to support economic activity [ 12 ]. The impact of drought on economic activities overall, in particular through reductions in power generation capacity, could be considerable. For example, between 1999 (winter) and 2004 (fall), seasonal decreases in precipitation of up to 60% compared to normal conditions were observed in the Canadian cities of Calgary, Edmonton and Saskatoon [ 9 ]. Hydroelectricity generation accounts for 62% of electricity generation in Canada [ 12 ] (A large amount of hydroelectricity is generated in Canada and when in-stream use is taken into account, this industry accounts for the largest amount of water use in Canada [ 12 ]. The volume of water used for hydroelectricity generation was over 70 times the total of all other combined water use in Canada in 2005 [ 12 ]. Thermal (nuclear and fossil fuel) power generation and manufacturing are responsible for the largest and second largest total withdrawals of freshwater in Canada [ 12 , 334 ]). Shortfalls in this area could further amplify costs and stresses associated with drought, for example through the lack of revenue or the need to purchase alternative energy (see section 3.12.2 for additional discussion of the health impacts of these shortfalls). During the 2001 drought, hydroelectric power generation in Saskatchewan fell to 66% of the past four year average in the province resulting in power being purchased from other jurisdictions to meet electricity needs [ 174 ]. During the 1987–1992 drought in California (U.S.), hydroelectricity generation fell to under 60% of normal levels [ 200 ]. Additional energy was purchased from more expensive sources [ 200 ] and led to more than a $3 billion increase in electricity costs for consumers [ 204 ]. Drought may exacerbate the substantial impacts anticipated with climate change. The climate change-related decrease in water levels anticipated across the Great Lakes is expected to result in direct economic impacts between 2014 and 2030 of $9.61 billion (2012 USD) due to impacts on recreation, shipping and harbours, hydroelectricity generation, waterfront property values and groundwater use [ 210 ]. 3.11. Mental Health Drought and its associated financial burden can lead to stress [ 51 , 176 , 201 , 211 , 212 , 213 , 214 ] and is associated with physical and mental health problems, such as anxiety, emotional and psychological distress, loss and grief [ 46 , 215 , 216 , 217 , 218 ]. The pathways leading to these impacts are illustrated in Figure 2 . For farm families, financial stress [ 51 ] and drought [ 219 ] can amplify the other stressors they experience. For consumers with low socio-economic status, stress related to an increase in food costs or the decreased availability of food can also negatively impact mental health [ 46 ]. Figure 2 Pathways through which climate change-related disasters can affect mental health. Source: Reproduced from Berry et al. [ 220 ]. Australia has produced numerous studies regarding drought's relationship to mental health. Negative drought-related impacts on mental health have been identified for older women and Aboriginal groups in Australia [ 216 , 221 ], as well as for rural Australian youth [ 222 ], with emotional stress increasing with the length of the drought [ 217 ]. Dryland salinity, a phenomenon similar to drought, has also been associated with an increased risk of hospitalization for depression in Australia [ 223 ]. Although Stehlik et al. [ 214 ] found that drought-related stress negatively impacted relationships within farm families (Central Queensland, New South Wales, Australia), a later large-scale study across Australian agricultural communities did not find family relationships were negatively impacted by drought [ 224 ]. For drought that was both constant and long (although not for either alone), there was a difference in the distress levels in rural and urban populations in Australia, with relatively higher levels among rural populations [ 225 ]. The degradation of the physical environment associated with drying could also contribute to psychological distress for those who feel connected to the landscape (solastalgia) [ 226 ]. Research from New South Wales (Australia) found a statistically significant relationship between drought and suicide from 1964 to 2001 [ 227 ]. Limited studies suggest that drought is a contributing factor to suicide among men in rural areas in Australia [ 228 , 229 ]. However, Page and Fragar [ 230 ] found a greater rate of suicide between 1988 and 1997 among male farm managers and agricultural labourers than among the male Australian population or the rural population overall, independent of drought. 3.12. Populations Vulnerable to the Health Impacts of Drought Populations at higher risk to the health impacts of climate change have been identified in the literature. They include infants and children, elderly people, the socially and economically disadvantaged, pregnant women, people who spend time outdoors and those with chronic diseases and compromised immune systems ( Table 1 ) [ 1 , 59 , 231 , 232 , 233 ]. Overlap exists between populations vulnerable to the health impacts of drought and populations at risk of adverse impacts on health resulting from other extreme weather events, including those expected to increase in frequency and intensity with climate change [ 46 , 59 ]. For example, people of low-socioeconomic status have more difficulty taking protective actions during extreme heat events [ 234 ] and may also suffer greater impacts from drought [ 49 ]. As an important pathway for increased exposure to injuries, air pollution, food, water and vector-borne diseases as discussed above, drought is associated with many higher risk populations. The existence of common vulnerability factors to multiple climate-related vulnerabilities increases overall risk to public health, but also offers significant opportunities to take adaptive actions that convey benefits among a broad spectrum of risk areas, including drought. For example, Cheng and Berry [ 235 ] identify efforts to enhance social capital as a climate change adaptation that has multiple health co-benefits. ijerph-12-08359-t001_Table 1 Table 1 Vulnerability to climate-sensitive health outcomes by subpopulation [ 233 ]. Groups with Increased Vulnerability Climate-related Vulnerabilities Infants and children Heat stress, air pollution, water-borne/food-borne diseases, vector-borne diseases, malnutrition Pregnant women Heat stress, extreme weather events, water-borne/food-borne diseases, vector-borne diseases Elderly people and people with chronic medical conditions Heat stress, air pollution, extreme weather events, water-borne/food-borne diseases, vector-borne diseases Impoverished/low socioeconomic status Heat stress, air pollution, extreme weather events, water-borne/food-borne diseases, vector-borne diseases Outdoor workers Heat stress, air pollution, vector-borne diseases, ultraviolet light (UV) exposure The following sections describe how specific populations may be at increased vulnerability to the health impacts of drought. 3.12.1. Agricultural Workers and Rural Populations Agricultural workers and others who rely on water and/or rainfall for their livelihoods are at greater risk of experiencing the mental health impacts of drought [ 46 , 51 , 83 , 174 , 176 , 212 ]. Over time new agricultural practices and technologies have increased agricultural output, but have also led to increased costs and vulnerability for farmers during drought [ 83 ]. Given their occupational environment, agricultural workers are at increased risk of "silo-filler's disease" (NO 2 gas poisoning within silos) [ 158 ]. Horticulturalists and people working for nurseries, garden supply stores and recreational facilities have also been identified as being at higher risk of negative health outcomes [ 46 ]. Research in Australia suggests that drought-related stress on agricultural workers can be widespread. A survey of rural farmers in New South Wales, Australia found that 71.8% of rural farmers reported stress due to drought [ 213 ]. Limited evidence suggests that drought is a contributing factor to suicide among men in rural areas [ 228 , 229 ]. In some studies, younger farmers were found to experience higher levels of stress than their older counterparts between 55–64 years but not their oldest counterparts aged 65–74 years [ 212 , 236 ]. The mental health impacts of drought can extend beyond farmers themselves to their families [ 237 ]. Rural Australian youth have experienced negative mental health effects during prolonged drought [ 222 ], with their emotional stress increasing with the length of the drought [ 217 ]. Older women and Aboriginals have also been identified as being more vulnerable to these impacts [ 216 , 221 ]. In general, rural and urban populations can experience different levels of drought-related distress. For drought that was both constant and long (although not for either alone) in Australia, distress levels in rural populations were observed to be higher relative to urban populations [ 225 ]. 3.12.2. People with Chronic Health Conditions, Pregnant Women, Young Children and Older Adults People with chronic health conditions are at increased risk of health impacts associated with air quality compromised by drought including the effects of wildfires [ 46 , 47 , 49 , 83 , 154 , 155 ]. This group is also thought to be at greater risk of drought-related infectious diseases [ 49 ]. Pregnant and nursing women, dialysis patients and immune-compromised individuals are considered to be at increased risk from health impacts of drought and heat [ 47 ]. Pregnant women in their third trimester, diabetics and immunosuppressed individuals were thought to have increased risk of developing a disseminated case of Valley fever [ 238 ]. Young children and seniors are at increased risk from drought-related water-borne diseases [ 49 ]. In addition to the very young (particularly those who have received rotavirus vaccines) and the elderly, those who are immunocompromised are at greater risk of developing acute gastrointestinal illness from water-borne diseases [ 95 ]. Very young children and seniors overlap with those susceptible to heat as identified above and may also be at greater risk from decreases in water quality and availability [ 60 ]. Children under five years of age and adults over 50 years old are also at increased risk of developing a disseminated case of Valley fever [ 146 ]. Individuals under institutional care (e.g., in hospitals, nursing homes) who depend on electrical equipment may be impacted by hydroelectricity power shortages during a drought [ 46 ]. Drought can also lead to disruptions of heating and cooling systems that rely on water to function, and these disruptions may result in health impacts on individuals when it is not possible to maintain health-protective temperatures [ 46 ]. 3.12.3. People with Low Socioeconomic Status People with low socioeconomic status are at greater risk of health impacts from climate change [ 59 ]. For example, families who are already facing challenges meeting their needs or who engage in subsistence fishing or farming may be at risk of food insecurity with drought [ 46 , 49 ]. In Canada, 8.3% of households were food insecure in 2011–2012, with 36.7% of households in Nunavut food insecure during this period [ 30 ]. 3.12.4. People Engaged in Activities Where Risk May Increase with Drought Individuals who spend more time outdoors have increased exposure to certain drought-related risks including those for respiratory diseases resulting from exposures related to wildfires and increased dust (e.g., dust-borne Valley fever) [ 59 , 146 ]. Exposure to vector-borne diseases, such as WNV or Eastern Equine Encephalitis, that may be linked to drought is increased in people spending greater amounts of time outdoors for recreation and/or due to their occupation (e.g., agricultural workers, parks services), placing them at greater risk of contracting these diseases [ 47 , 76 ]. Although the overall risk is considered to be low, individuals with greater exposure to the droppings of infected rodents are at increased risk of contracting rodent-borne diseases (e.g., Hantavirus) [ 239 ]. Recreational water users are at greater risk of contracting water-borne diseases, especially from pathogens favored by warm, shallow waters that often accompany drought [ 46 , 49 , 75 ]. In addition to biological pathogens, they may also have greater exposure to chemical contaminants [ 46 ]. Spinal cord injuries related to diving into shallow waters largely occur in males between 15–25 years of age [ 170 ]. 3.13. Adaptation to the Human Health Impacts of Drought (Direct) A broad range of adaptation strategies are available that help reduce the health impacts of drought. They generally attempt to reduce the impacts directly or they modify the environmental services or socioeconomic pathways through which drought affects health ( Table 2 ). However, there has been little evaluation of the effectiveness of many of these adaptations [ 183 , 240 ]. Limited knowledge of the effectiveness of current measures to reduce health risks from drought increases the possibility of "maladaptation" as communities and individuals prepare for climate change. Ensuring that adaptation strategies are developed with evidence-based information and through consultation with a wide range of partners mitigates the risk of maladaptation. ijerph-12-08359-t002_Table 2 Table 2 Examples of adaptations to reduce the health impacts of drought. Adaptations to the human health impacts Water quality and water treatment monitoring Disease surveillance and monitoring Air quality monitoring and warnings Public education and outreach Mental health prevention and awareness programs Adaptations to address changes to environmental services or socioeconomic pathways Mainstreaming climate change into source water protection planning Enhancing water service delivery and watershed management Water capture and storage Increasing water use efficiency and water conservation Modifying or adopting new agricultural practices, such as increasing reliance on crop irrigation Introducing drought-resistant crops and diversifying crops and income sources Using seasonal climate outlooks for crop planning Developing watershed management and climate change adaptation plans Insurance and other financial assistance programs Irrigation and altering the timing and use of water 3.13.1. Monitoring Water Quality and Water Treatment In addition to monitoring indicators for drought itself [ 47 ] or drought-related drivers of health impacts [ 241 ], increased water quality testing and monitoring of drinking water and/or recreational water during droughts can help safeguard health [ 43 , 59 , 65 , 74 , 92 ]. For example, the province of Quebec has begun to adapt by increasing the monitoring of surface water and adopting more stringent standards for water quality surveillance. Water treatment is also an important adaptation to protect against water-borne disease [ 59 ] and boil-water advisories may be a supportive tool, although they are designed to address fecal and not algal or chemical contamination. During the 2001 drought within the Kanai Blood Indian Reserve in Alberta (Canada), there was an increase in the number of boil water advisories called to protect residents against potential contamination [ 174 ]. 3.13.2. Disease Surveillance and Monitoring Surveillance and monitoring of diseases sensitive to drought such as those caused by E. coli , C. perfringens and WNV are important adaptations [ 43 , 52 , 53 , 56 , 74 , 94 , 242 , 243 , 244 ]. Surveillance and monitoring can help to direct the provision of medical services and the development of needed vaccines to reduce risks to health [ 59 , 74 , 94 , 242 , 243 ]. Modeling could play a role alongside surveillance [ 125 ]. Control of disease vectors may also be implemented [ 59 ]. Primary care physicians can be part of monitoring efforts and help identify and address cases of malnutrition during drought episodes [ 207 ]. 3.13.3. Air Quality Monitoring and Warnings When drought threatens air quality, for example by increasing particulate matter, air quality monitoring and warnings for the public can help reduce health risks [ 51 ]. In over 60 communities and 10 provinces across Canada, the Air Quality Health Index (AQHI) provides individuals with information on the immediate health risks associated with local air quality [ 245 ]. The AQHI includes information on how to reduce exposure to air pollution that is tailored to the general public and to vulnerable populations (e.g., parents with children and infants, seniors and those with cardiovascular and respiratory diseases). It is regularly updated and accessible through the Internet [ 246 ]. However, it may be challenging to capture a quickly-developing event such as a forest fire smoke episode through the Index. Environment Canada informs the public of severe weather by providing special weather statements, advisories and watches [ 247 ]. In responding to wildfires in western Canada in 2014, Environment Canada issued alerts through wildfire smoke warnings and special air quality statements [ 248 ]. Provincial and community-level agencies further monitor events such as forest fires and provide information to those affected through a number of platforms, including websites and social media [ 249 ]. 3.13.4. Public Education and Outreach Public education is a key adaptation to protect health from drought. Personal protection measures that can be promoted by public health officials can include reducing water- and vector-borne disease exposures [ 59 ], wearing respiratory masks on days when ambient air quality could impact health [ 51 ], and the promotion of hand washing [ 74 ]. Education for travelers to and from Canada can raise awareness about diseases such as malaria [ 80 ], or tuberculosis, which may be increasingly transmitted under social conditions created by drought in developing countries [ 52 ]. To prevent spinal cord injuries resulting from diving in shallow waters, some jurisdictions have implemented public education, including the posting of notices where diving is not safe [ 170 ]. For example, brain and spinal cord injury prevention programs have been carried out in Saskatchewan and Ontario (Canada) and in Washington State (U.S.) [ 250 , 251 , 252 ]. However, evaluations of these programs suggest that program delivery needs to be well designed and robust to be effective. Although the initiative in Saskatchewan was found to change knowledge among students, in the Toronto (Ontario) case only 16% of students in schools where the program video was distributed viewed the information [ 250 ]. The one time intervention in Washington State, where students attended a session which included a video, lecture and personal testimony around traumatic brain and spinal cord injury, had little impact on knowledge or behavior [ 252 ]. Similarly, evidence suggests that boil water advisories may not be adhered to by everyone. During the gastroenteritis outbreak in Walkerton, Ontario (Canada), some residents under the boil water advisory continued to use and occasionally drink the water without boiling it [ 85 ]. The link between water and hygienic practices should be considered when developing drought adaptation measures. Conservation should take into consideration on-farm agricultural practices, regional water management and treatment and allow for and encourage individual hygienic practices such as hand washing [ 46 ]. In the later example, homes and businesses could also promote both water conservation and hand washing by installing low-flow faucet aerators [ 46 ]. Education of health care providers can also benefit efforts to address drought impacts on health [ 53 ]. 3.13.5. Addressing Mental Health Impacts Adaptive actions may be taken by individuals and professionals within primary care and community services settings to reduce the mental health impacts of drought. Personal hopefulness and positive attitudes towards the rural lifestyle have been highlighted as useful adaptations to drought [ 218 , 253 ]. According to Caldwell and Boyd [ 254 ] rural Australian families used a combination of both positive (e.g., problem-focused coping, optimism and positive appraisal) and less beneficial strategies (e.g., cognitive dissonance, denial and avoidance of negative social influences) in their response to drought. Primary care practitioners can provide support to reduce symptoms of mental illness exacerbated by drought and ineffective coping mechanisms such as substance abuse. They may also provide information about financial support resources or community drought support groups [ 207 , 255 ]. Mental health programs in schools and in occupational settings and suicide prevention and awareness programs have also been identified as possible adaptation measures [ 59 , 216 , 222 , 256 , 257 ]. Newspapers may be another pathway to communicate broadly about mental health during drought [ 258 ].In Australia rural support workers help traditional mental health professionals identify farmers experiencing mental distress and refer them to support services [ 259 ]. Training in mental health first aid can also increase the capacity of rural support workers to address mental health needs in drought-affected rural areas [ 260 ]. One study suggested that the Rural Adversity Mental Health Program (RAMHP), which is based on a community development model, is effective in building capacity and resiliency to cope with the impacts of drought in rural New South Wales (Australia) [ 261 ]. Linking the services and programs focused on mental health, community or social needs with those on practical assistance for farming communities is considered key for addressing mental health concerns associated with or exacerbated by drought [ 262 ]. Access to support services for older individuals is another consideration [ 256 ]. Community-building [ 263 ] and events which increase social capital can further help decrease psychological stress [ 254 ]. High levels of social capital in many Canadian (Prairie) rural communities [ 176 , 264 ], based upon high levels of trust, networking and participation in organizations help reduce stress associated with drought [ 176 , 264 ]. However, Australian farm families faced heavier workloads during the drought of 2002–2003, which left less time for community participation, volunteer activities and hence could negatively impact social capital and the community [ 201 ]. 3.14. Adaptation to Address Changes to Environmental Services or Socioeconomic Pathways (Indirect) 3.14.1. Adaptation in Urban Environments Effective adaptation to climate change-related impacts, including drought, requires "mainstreaming" in which climate change information is routinely considered and incorporated into existing programs, policies and planning [ 2 , 62 ]. For example, in Ontario, de Loe and Berg [ 62 ] point to opportunities to "mainstream" climate change into source water protection planning within the Clean Water Act to help address anticipated changes to water supply and demand, including an increased frequency of drought [ 62 ]. Mainstreaming can be assisted by integrating learnings about effective risk management options and adaptations from experiences with current climate variability, particularly extreme weather events. However, mainstreaming also requires consideration and use of information about risks expected from future climate change in adaptation planning. A range of measures to adapt to changes in water quality and quantity in the context of climate change, such as leak detection, outreach, adjustments to the building code and management policies have been reported in the province of Quebec [ 49 ]. Many sectors will be affected by drought and play a role in adaptation to reduce risks to individuals and communities. In urban centers communities may adapt to drought by enhancing water delivery systems and through watershed management [ 59 , 74 , 204 ]. In Atlantic Canada, communities facing shortages or reduced water quality have made efforts to improve water treatment systems, conserve and protect surface water and develop new water sources through, for example, water capture [ 50 ]. Dams, desalination and water recycling are also being used to cope with the increased variability in water supply, although the cost and impacts of such actions are not well established [ 204 ]. The health sector must actively be engaged in drought management at all levels of government. For example, in Brazil actions to protect public health are embedded into all stages of drought-related disaster risk reduction [ 241 ]. Increasing water use efficiency is advocated as an important adaptation when water infrastructure already exists [ 204 ]. Water conservation has also been highlighted as an important measure for easing drought-related costs [ 50 , 51 , 204 ]. This may be done through voluntary or mandatory restrictions on water usage during vulnerable periods [ 43 ] or education campaigns to promote conservation [ 50 , 51 ]. During the 1987–1988 drought in California (U.S.) education was used in conjunction with mandatory and voluntary water conservation programs to target a 10%–25% reduction in water use [ 265 ]. As this drought stretched on, impacts spread from dryland agricultural areas to impacts on state level electrical power supply. Lessons learned in a study of this drought identified inadequacies in water resource planning, the existence of responsible agencies along political rather than affected area boundaries and a lack of inter-jurisdictional communication as key factors. Water shortages continue to occur in California despite national freshwater draws (per capita) having dropped below levels seen in 1975 [ 265 ], with the exception of municipal and industrial use which are also expected to decline as regulations and technologies for more efficient use are adopted. In the most recent 2014 drought in California the entire state was under moderate to exceptional drought and water customers in Santa Cruz were required to decrease their water consumption by 25% with financial penalties for non-compliance [ 266 ]. 3.14.2. Adaptation in the Agriculture Sector Farmers have a great deal of experience in coping with changes in conditions, including drought and are seen to be one population group with a high capacity to adapt [ 174 , 203 ]. In agricultural systems adaptation to drought occurs through the use of modified or new agricultural practices, such as increasing reliance on crop irrigation and the introduction of drought-resistant crops, as well as diversifying crops and income sources [ 61 , 244 , 254 , 267 , 268 , 269 , 270 , 271 , 272 ]. With the added challenge of climate change, seasonal climate outlooks are another tool that farmers in the U.S. are turning to in order to better determine which crops will be successful in a particular season [ 273 ]. The international literature also highlights the importance of drought forecasts and preparing for drought in advance [ 243 , 267 ]. The Southern African Development Communities Drought Monitoring Centre is an example of such an alerting body [ 274 ]. In Canada, agricultural adaptation over the past century has included the development of new agricultural practices and infrastructure to conserve and enhance water supplies and distribution and targeted institutional activities to improve water management [ 174 , 203 , 275 ]. For example, less frequent tilling and reductions of other field management practices that lowered soil moisture and increased erosion aided in lowering the vulnerability of farmers to the 2001–2002 drought relative to the drought of the 1930s [ 203 , 275 ]. Governments have also increased regional resilience by producing drought plans and enacting legislation to support coordinated water management. In England, water resource management plans have been incorporated into climate change adaptation plans [ 276 ]. In the U.S., drought plans have been produced by a number of states [ 277 ]. Water management and climate change adaptation plans are being developed by various watershed, municipal and regional authorities across Canada [ 278 , 279 ]. However, climate, human and other factors affecting water sources and reserves are extensive and rarely conform to the jurisdictional boundaries at which watershed management is required [ 280 , 281 ]. Programs to sustain farmers through times of very low production or crop failure due to natural hazards such as drought are expected to be important in the future [ 282 ], but developing proactive adaptation actions is key to increasing the resilience of existing agricultural systems to drought [ 283 , 284 , 285 , 286 ]. In fact, in the Prairies there are some areas in which conditions such as soil erosion and the accumulation of soluble salts in the root zone may warrant reclassification in use [ 287 ]. A media analysis of adaptations undertaken in Canada during the 2001–2002 drought showed that the most frequently mentioned actions included measures related to irrigation, conservation and the management of water resources, insurance and other financial assistance programs (e.g., Canadian Farm Income Program, Net Income Stabilization Account), as well as government based low water responses [ 288 ]. Irrigation in particular is one of the most broadly referenced agricultural adaptations to drought in the literature [ 289 ]. The use of irrigation systems is increasing across Canada and irrigation is becoming more and more common in regions that had previously relied on rain and subsurface water (Saskatchewan has an Irrigation Strategy to grow and extend the use of irrigation across the province as part of their Water Security Plan http://www.agriculture.gov.sk.ca/Irrigation-Strategy-2014 ; see also [ 335 ]). In Prince Edward Island the growing reliance on irrigation in potato farming and requests for new deep well licenses created serious concerns regarding the sustainability of groundwater supplies and whether or not a moratorium instituted over a decade earlier was still required [ 290 , 291 ]. In November, 2014, a decision was made to uphold the moratorium. In more arid regions where water supplies are finite, like the Okanagan region of southern British Columbia, water demands are managed through a combination of monitoring and management among users (e.g., managing water catchments, use of high elevation dams, reservoirs and instituting seasonal use restrictions). In Alberta's Oldman Dam water shortages have even been managed using a system of "sharing the shortage" in which users agree to share the reduced allocation of water due to drought [ 292 ]. In 2001, partial allotments of 60% were successfully used to ensure the delivery of water supplies to all licensees. Water consumption can also be reduced by managing the timing of irrigation so that it targets the sensitive stages of crop development. In the case of corn, for example, yield losses may be reduced significantly by preventing water deficits at critical times in development; notably during the reproductive stages [ 293 , 294 ]. Even when plants are subjected to drought conditions, the timing and duration of the event must also be considered since plant processes can often recover following a period of drought if critical management actions are taken at the correct time (e.g., irrigation during flowering). Water conservation strategies and the timing and use of water (water management and irrigation timing) represent simple solutions that can be managed in response to need, when and where required ( i.e. , the farm scale). In contrast, technological and transgenic approaches will take time to develop [ 295 ]. Moreover, since drought resistance is linked to overall plant performance, cultivars that are tolerant to extreme drought conditions may result in reduced production during less extreme or episodic drought conditions [ 296 , 297 , 298 ]. As such, farmers may need to balance trade-offs between production and the risk of drought in their choice of what to plant next season. Planting seeds representing a diversity of cultivar tolerance may prove beneficial in this case [ 299 ]. The combined use of existing drought tolerant cultivars with physiologically sensitive timing of irrigation can serve to reduce water demands during key stages of development. Furthermore, since drought stressed plants and livestock are also more vulnerable to diseases, competition from invasive species and pests, strategies that increase genetic and crop diversity across fields and the landscape may represent adaptations that will provide overall benefits to both health and production [ 288 , 300 ]. Looking to the future, against a global backdrop of diminishing lands suited to agriculture and multiple cropping due to climate change, Canada is expected to experience an increase in the area of land that is marginally or moderately suitable for agriculture in farther northern regions [ 301 ]. 3.14.3. Adaptation to Build Individual and Community Resilience At the individual level, personal savings may not be enough to support families during droughts, hence loans, savings strategies and insurance may be important adaptation options [ 267 , 270 , 302 ]. In the U.S., many families went into bankruptcy under the combined drought and Great Depression in the 1930s and had to draw on government aid for support [ 202 ]. Others moved to different communities to earn a living [ 202 ]. In 2012, following a particularly dry summer in Ontario, over 5000 crop damage reports were submitted by farmers [ 303 ]. Although drought-related crop losses in the Prairies from 2002 were offset by farm insurance [ 174 ], not all farmers are registered for insurance [ 303 , 304 ]. Moreover, while smaller farms tend to have more diversified crops which serve in part to mitigate risk ( i.e. , act as adaptation strategies; [ 305 , 306 ]), they may also be less likely to have insurance coverage. Another key adaptation to variability has been to increase income from off-farm sources [ 174 ] including from oil and gas leases [ 203 ]. The long-term viability of such tactics and their impact on social cohesion in communities is unclear. Socioeconomic challenges can impact the adaptive capacity of communities [ 307 ], but so can more external influences such as global scale changes in agricultural production, demand and markets [ 308 ]. As such, the presence of well-developed social institutions and access to highly qualified personnel and information are particularly important for remote rural communities [ 309 ]. In Canada, drought resiliency in some rural and agricultural communities is related to proximity to urban centers. A study by the International Institute for Sustainable Development and Agriculture and Agri-Food Canada [ 309 ] in a 550,000 square kilometer prairie region (comprising 53 Census Divisions across Alberta, Saskatchewan and Manitoba) found that the areas exhibiting the highest adaptive capacity index values were clustered within corridors between major urban centers. Proximity to urban centers is thought to bring benefits related to off-farm incomes, access to technology (including computers/computing technologies), information networks (email, Internet) and access to agricultural institutions. Conversely, the lowest adaptive capacity was observed in areas more remote from urban centers, which were situated in more northern zones of this prairie belt. The importance of social networks and connectivity may be more obvious during periods of crisis, such as during the 2002 drought on the Prairies when farmers in eastern Canada who were less affected by drought shipped 64,000 bales of hay to western Canada through "Hay West" to support the worst struck areas [ 175 ]. In 2012, some western Canadian farmers attempted to reciprocate with hay to eastern farmers [ 175 ]. Adaptations to address the impact of drought on mental health and social capital within the community are also discussed above. 3.15. Discussion 3.15.1. Knowledge of Current and Future Drought Impacts on Health In recent years information about the potential health impacts of drought has increased. Evidence pointing to health impacts that may be modified by climate change has continued to build since earlier reviews (e.g., by Smoyer-Tomic [ 51 ] in 2004). The number of Canadian-U.S. studies is increasing, with more than half of the records for this region published after 2004. However, Canadian studies continue to be focused on the Prairies. In addition to studies situated within the Canadian context broadly, more than half of the records identified focused specifically on western Canada and all but one targeted the Prairies specifically. Research linking mental health impacts with wildfire and drought has also expanded significantly. However, research around drought and health remains narrow. Approximately three-quarters of the primary international studies identified were focused on mental health and nearly all focus on Australia. In the Canadian-U.S. literature, more than one-third of the primary studies are focused on mosquito-borne disease, largely WNV and the related SLE. This also appears to be the area of current interest, with these records accounting for more than half of the primary Canadian-U.S. studies published after 2004. Limited data on actual health outcomes results in uncertainty about the severity or magnitude of these impacts. Consequently projections of climate change effects on drought-related health impacts remain difficult to make especially in the Canadian context. At the same time that more evidence is needed to understand the impact of drought on health outcomes, public health is challenged by the uncertainty surrounding the probable occurrence of future drought in Canada. There has been limited research directed toward the probability and assessment of future droughts specifically over Canada in the context of climate change and variability. Greater certainty around the characteristics of drought anticipated in the future and their linkages with specific health impacts could help adaptation planners better identify and prepare for key risks in their regions, including those related to human health. For example, the literature reviewed in this paper suggests that certain features of drought (e.g., subsequent heavy rainfall events) may significantly increase risks of some infectious diseases among Canadians. Knowledge of geographic areas likely to be affected by drought, as well as potential changes to frequency, duration and severity of events is critical for the development of adaptation strategies which can directly or indirectly reduce health impacts. Coordinated drought research is limited and fragmented in Canada, often carried out only in response to severe drought experience. Although most regions of Canada have experienced drought, the majority of studies have focused on the Prairies due to the greater frequency of droughts in this region. The variety in the forms of drought experienced in Canada needs to be taken into consideration when planning for possible health impacts. Given the complex interactions with water, a discussion of the impact of drought on soil through the health lens may add to the literature on water and water content alone. 3.15.2. Knowledge of Adaptation Options A recent 2013 systematic review of the effectiveness of public health interventions to reduce the health impacts of climate change highlighted a gap in the literature around adaptation to drought [ 310 ]. While there has been some effort to assess agricultural adaptation to drought [ 311 ] and public health promotion activities that touch on related impacts, such as spinal cord injury [ 250 , 251 ], none of the public health adaptations specific to drought identified in the international literature had been evaluated. Here, again, a lack of specificity around drought among other climate change-related extreme weather events limits the relevance of these discussions for public health planning. Potential adaptations to directly or indirectly reduce health risks from drought are plentiful, particularly those that address agricultural impacts of climate change and weather variability. Few adaptations are discussed specifically in the context of drought but are often put forward to address secondary concerns (e.g., vector-borne diseases). Factors influencing agriculture and agricultural practices are complex and sometimes contradictory, ranging from public level issues driven by policy, legislation and international agreements, to environmental and community issues relating to how land is used and the local impacts [ 287 ]. To a large extent adaptations to drought are emerging from regional risk and adaptation case studies [ 282 ]. There is an opportunity to address multiple climate change-related health impacts under broad public health adaptations. In addition, adaptation activities and programs in other sectors can contribute co-benefits to the agricultural sector. Programs that have been developed for other objectives, such as protecting wildlife or water, have been found to provide incidental benefits to the agricultural sector during drought periods. Such co-benefits are not well recognized, but constitute just short of two-thirds of activities captured in a national inventory of programs for managing drought and agricultural water in Canada [ 311 ]. 3.15.3. Challenges Adapting to Drought Measures taken to adapt to drought between 2000 and 2006 over the Canadian Prairies were expensive, their effectiveness varied and challenges were associated with their implementation [ 199 ]. Gaps in knowledge and misplaced assumptions have been identified as obstacles to drought adaptation. A number of barriers to agricultural adaptation to drought were highlighted during the drought of 2001–2002 in Canada, including the lack of research and a lack of funds, as well as problems relating to institutions and processes [ 288 ]. A study of water management systems in the North American Pacific Northwest found that water management decisions were centred around an assumption of stable conditions and this was compounded by a lack of access to regionally specific climate information [ 312 ]. Challenges have also been noted in simply obtaining high quality local hydrologic information from climate change datasets [ 184 , 185 , 313 ]. The paucity of such information makes it difficult for communities to develop more integrated climate adaptation plans which reduce drought risks. Severe drought can exceed the adaptive capacity of agricultural producers and communities. In their analysis of the 2001–2002 Canadian drought, Wheaton et al. [ 288 ] found that some adaptations were ineffective and that drought brought about negative impacts even when adaptation measures were put in place. Drought can be a prolonged event, which also suggests the importance of considering both long-term and short-term adaptations and approaches to mitigate any long-term health impacts. Despite rural areas having greater social capital than other parts of Canada [ 176 , 264 ], there is much reliance on technologies which may not be able to counter the full impacts of climate change on the production of key crops [ 178 ]. Adaptations derived from new technologies will take time to understand, develop and institute (e.g., transgenics or traditional breeding for drought tolerance) [ 314 ]. Furthermore, the resources required to generate necessary information about future climate and weather conditions and the required analyses may not be available to provide timely planning direction to farmers, municipalities and other local planners [ 286 , 315 ]. The literature reviewed in this paper highlighted that remote areas and communities may have significantly less adaptive capacity to reduce drought impacts than communities closer to urban centres. 3.15.4. Key Knowledge Gaps Key knowledge gaps that need to be addressed to inform drought adaptation include: Projections of drought for Canada (beyond western Canada) including the features of those droughts and regional susceptibility to increased frequency and/or intensity of major droughts; The impacts of drought on infectious diseases which represent a substantial burden of illness in Canada already (e.g., C. perfringens , Campylobacteriosis ), as well as the influence of climate change on these impacts; The impacts of drought on disease rates associated with fungi that are now established in Canada (e.g., Cryptococcus gattii and Blastomyces dermatidis ), as well as the influence of climate change on these impacts; The impact of drought on particle levels in already stressed airsheds and implications for health, as well as the influence of climate change on these impacts; The applicability of mental health impact studies of drought to the Canadian context; Impacts of drought on dust and wildfire-related health impacts; The effectiveness of adaptations to drought, including the impacts of non-health adaptations (e.g., for the agricultural sector) on reducing negative health outcomes. 3.1. Health Risks Associated with Drought Drought can affect human health through a broad range of pathways including those shown in Figure 1 . They are discussed in greater detail in the following sections. Generally speaking, these broadly fall into environmental services and socioeconomic areas. There is agreement that globally, drought leads to an increase in morbidity and mortality [ 41 , 42 , 43 , 44 ]. Although discussions have pointed to direct drought-related deaths in low-income countries [ 45 ], a broad range of public health impacts associated with drought has been identified in the U.S. [ 46 , 47 , 48 ]. Research also indicates that climate change may impact drought-related health outcomes in Canada [ 2 , 49 , 50 ]. In fact, some evidence of drought-related health impacts has been identified for Canadians living on the Prairies [ 51 ]. For example, there is some evidence of respiratory impacts related to dust and increased risk of water-borne disease [ 51 ]. Furthermore, reviews of impacts from Ontario [ 52 ] and the U.S. [ 53 ] have identified climate change and future drought as having the potential to affect food-, vector- and water-borne diseases [ 54 , 55 ]. 3.2. Water Quality and Human Health 3.2.1. Source Water for Drinking Water Drought can negatively impact the quality of drinking water [ 2 , 23 , 46 , 50 ]. Decreases in drinking water quality can, in turn, lead to water-related disease [ 42 , 56 , 57 , 58 , 59 , 60 ] and the exacerbation of heat stress [ 61 ]. The pattern of reduced water levels, stream flow and resultant stagnation during drought can increase contaminant concentrations in ground and surface waters [ 46 , 51 , 62 , 63 ]. For example, measurements taken during a drought in 2003 over the Rhine and Meuse rivers in Europe showed not only a substantial reduction in river discharge relative to the year before, but also that chloride concentrations were inversely related to the flow (m 3 /s) of these rivers [ 64 ]. This effect is thought to increase when contaminants continue to be added to the system [ 64 ]. Lower water levels can also facilitate an increase in reactions between the water, contaminants and sediments [ 62 ]. This can be further reinforced by greater soil erosion during drought [ 23 ]. In the U.S., elevated nitrate, orthophosphates, chlorides and sulfates in groundwater have also been associated with drought, in some cases exceeding the U.S. EPA's Maximum Contaminant Level [ 63 ]. Frequent detection in wells near agriculture and septic systems suggested these systems were at risk of contamination [ 63 ]. Baures et al. [ 65 ] found that variation in total organic carbon and nitrate was associated with changes to river flow in France, with a high ratio of total organic carbon to nitrate concentrations under conditions of very low river flow. A UK study found that drought can lead to greater contamination of drinking water source water from dissolved organic carbons and can increase difficulty in contaminant removal [ 66 ]. Similar patterns have been reported in Canada. Elevated nutrient concentrations ( i.e. , phosphorous and nitrogen) have been observed in association with low water periods in Alberta [ 67 ]. Figure 1 Pathways through which drought impacts human health in the context of climate change. Nutrient-loading is known to lead to eutrophication and is associated with cyanobacterial (algal) bloom development in all provinces [ 68 ]. Some cyanobacteria produce toxins that are harmful to humans and blooms can impact water bodies used as sources of drinking water and for recreational water activities [ 2 ]. In Nebraska (U.S.), for example, drought was considered a contributing factor to a cyanobacteria bloom in 2004 that resulted in reports to public health of impacts such as gastrointestinal illness and the issuance of health alerts and advisories for nearly 100 lakes [ 69 ]. However, Barbeau et al. [ 70 ] carried out a hypothetical analysis of the vulnerability of Quebec's provincial drinking water systems to toxic cyanobacteria and their results suggest that existing systems in that province would be able to cope with this risk, even within the context of climate change. In Ontario, there was a significant increase in algal blooms between 1994 and 2009 [ 71 ]. Climate change may lead to changes in the frequency and duration of algal blooms, as well as increases in the amount of toxins they produce or an increase in the prevalence of toxin-producing species overall in temperate regions [ 70 , 71 ]. Warmer temperatures anticipated with climate change may promote algal blooms through mechanisms such as reducing the mixing of water, allowing algae to rise to the surface more quickly and by increasing the salinity of freshwater bodies [ 72 ]. The warmer temperatures associated with drought and decreases in water flow and/or volume may promote the survival and growth of pathogens in water, including those responsible for infectious disease [ 46 , 55 , 73 ]. Reduced water flow can lead to an increase in bacterial concentration and reduced water quality as there is less water available to dilute these contaminants [ 64 , 74 ]. Climate change-related impacts on water, including drought, are recognized as leading to greater risks from water-borne diseases such as hepatitis A [ 52 ]. However, there appears to be limited discussion of the relationship between drought and water-borne pathogens and the interactions between climate change and water-borne diseases are complex. In the U.S., the survival and multiplication of bacteria leading to leptospirosis outbreaks may have been altered by temperature and pH changes resulting from the stagnation associated with drought [ 75 ]. Similarly, although Naegleria fowleri is rare, increased water temperature during drought can allow for the amoeba to grow in greater concentrations [ 46 ]. Drought conditions may concentrate pathogens while at the same time inactivate them (e.g., Cryptosporidium oocysts) [ 54 ]. In Canada, it is important to note that the greatest burden from enteric, food and water-borne diseases, drawing on 2012 data, is associated with campylobacteriosis (10,174 cases), followed by salmonellosis (6828 cases) and giardiasis (3862 cases). Other reportable diseases in this category have a far smaller number of cases, such as shigellosis with 1,068 reported cases and approximately 500 cases each of E. coli and cryptosporidiosis in 2012 [ 76 ].In addition, while at its peak in 1970, there were 12,283 cases of hepatitis A reported in Canada; this has since fallen to approximately 200–300 cases per year [ 76 ]. Future changes to the frequency and/or severity of drought could modify the burden associated with these diseases in Canada. In coastal areas, drought can result in the contamination of groundwater drinking water sources by the intrusion of salt water [ 46 , 50 ]. The mingling of saline and fresh groundwater that occurs in such regions can progress inland affecting freshwater wells, under a number of circumstances, including drier periods [ 77 ]. Saltwater intrusion can be exacerbated by climate change and sea-level rise [ 50 , 77 ]. The combination of drought with extreme precipitation may also impact water-, food- and vector- borne diseases. The most commonly cited impact of climate change on water-borne disease is the anticipated effects of increasing frequencies and amounts of heavy rainfall impacting the likelihood that untreated and pathogen-carrying water is consumed by Canadians [ 52 ]. Drought can amplify the impacts of a subsequent extreme rainfall event. During a drought event the soil can become compacted, thereby increasing runoff and the likelihood of water contamination and water-borne illness from precipitation events that follow [ 46 , 54 , 55 , 78 ]. Conditions associated with drought, such as low rainfall, can also concentrate Giardia and Cryptosporidium cysts in groundwater sources and where water is stored. Water sources may then be contaminated by the cysts after rainfall events [ 79 ]. Evidence from the Prairies suggests that the prevalence of livestock farming may affect water-borne diseases that follow such events if manure is able to contaminate water sources (e.g., Eschericia (E). coli , Cryptosporidium , Giardia ) [ 80 , 81 , 82 , 83 ]. Campylobacteriosis outbreaks are also frequently associated with heavy rainfall, often when they follow a period of drought [ 54 ]. Although a much smaller risk for Canadians, with typically only 0–2 cases in Canada each year [ 76 ], the pattern of drought conditions followed by a wet spring could also activate anthrax [ 52 ]. 3.2.2. Treated Drinking Water During drought, there can be a drop in pressure at water treatment plants [ 52 ], which can increase turbidity leading to a greater potential risk of water contamination and consequently a greater need for water treatment [ 84 , 85 ]. The impact of drought on contaminant levels in source water has been discussed above. In the event that well water becomes unavailable during drought, turning to alternative water sources can also increase risks to health if appropriate safeguards for those treatment measures are not in place [ 86 ]. Inadequate water treatment is important as pathogens, including bacteria and viruses, are known to be taken up by particles in water [ 87 ]. Costs also increase with the additional treatment required to address higher levels of turbidity [ 88 , 89 ]. Changes to water flow and levels associated with climate change could further contribute to stress on water treatment systems and increase the cost of treatment [ 50 ]. In northern Ontario, some First Nations communities may already be near the limit of their capacity in this regard [ 50 ]. 3.2.3. Recreational Water Drought can also affect the quantity and quality of recreational water [ 59 , 60 , 61 , 65 , 90 , 91 , 92 , 93 , 94 ]. The hot and dry weather that can be associated with drought is thought to increase recreational water use [ 46 , 75 ]. At the same time, the increased concentration of pathogens encouraged under these conditions, as described above, are thought to increase the likelihood of infection among recreational water users [ 46 ]. Although leptospirosis cases and outbreaks are frequently associated with heavy rainfall events [ 57 ], in the U.S., they have also been associated with drought and with swimming in lakes [ 75 ]. In Canada, even under typical (non-drought) conditions, acute gastrointestinal illness (AGI) is a risk for swimmers, with the risk of AGI at approximately 3%–8% [ 95 ]. Although the ingestion of water during recreational use would be lower than that from the drinking water sources discussed above, recreational water users could be exposed to water-borne pathogens that lead to infection through direct contact or inhalation [ 96 , 97 ]. 3.2.1. Source Water for Drinking Water Drought can negatively impact the quality of drinking water [ 2 , 23 , 46 , 50 ]. Decreases in drinking water quality can, in turn, lead to water-related disease [ 42 , 56 , 57 , 58 , 59 , 60 ] and the exacerbation of heat stress [ 61 ]. The pattern of reduced water levels, stream flow and resultant stagnation during drought can increase contaminant concentrations in ground and surface waters [ 46 , 51 , 62 , 63 ]. For example, measurements taken during a drought in 2003 over the Rhine and Meuse rivers in Europe showed not only a substantial reduction in river discharge relative to the year before, but also that chloride concentrations were inversely related to the flow (m 3 /s) of these rivers [ 64 ]. This effect is thought to increase when contaminants continue to be added to the system [ 64 ]. Lower water levels can also facilitate an increase in reactions between the water, contaminants and sediments [ 62 ]. This can be further reinforced by greater soil erosion during drought [ 23 ]. In the U.S., elevated nitrate, orthophosphates, chlorides and sulfates in groundwater have also been associated with drought, in some cases exceeding the U.S. EPA's Maximum Contaminant Level [ 63 ]. Frequent detection in wells near agriculture and septic systems suggested these systems were at risk of contamination [ 63 ]. Baures et al. [ 65 ] found that variation in total organic carbon and nitrate was associated with changes to river flow in France, with a high ratio of total organic carbon to nitrate concentrations under conditions of very low river flow. A UK study found that drought can lead to greater contamination of drinking water source water from dissolved organic carbons and can increase difficulty in contaminant removal [ 66 ]. Similar patterns have been reported in Canada. Elevated nutrient concentrations ( i.e. , phosphorous and nitrogen) have been observed in association with low water periods in Alberta [ 67 ]. Figure 1 Pathways through which drought impacts human health in the context of climate change. Nutrient-loading is known to lead to eutrophication and is associated with cyanobacterial (algal) bloom development in all provinces [ 68 ]. Some cyanobacteria produce toxins that are harmful to humans and blooms can impact water bodies used as sources of drinking water and for recreational water activities [ 2 ]. In Nebraska (U.S.), for example, drought was considered a contributing factor to a cyanobacteria bloom in 2004 that resulted in reports to public health of impacts such as gastrointestinal illness and the issuance of health alerts and advisories for nearly 100 lakes [ 69 ]. However, Barbeau et al. [ 70 ] carried out a hypothetical analysis of the vulnerability of Quebec's provincial drinking water systems to toxic cyanobacteria and their results suggest that existing systems in that province would be able to cope with this risk, even within the context of climate change. In Ontario, there was a significant increase in algal blooms between 1994 and 2009 [ 71 ]. Climate change may lead to changes in the frequency and duration of algal blooms, as well as increases in the amount of toxins they produce or an increase in the prevalence of toxin-producing species overall in temperate regions [ 70 , 71 ]. Warmer temperatures anticipated with climate change may promote algal blooms through mechanisms such as reducing the mixing of water, allowing algae to rise to the surface more quickly and by increasing the salinity of freshwater bodies [ 72 ]. The warmer temperatures associated with drought and decreases in water flow and/or volume may promote the survival and growth of pathogens in water, including those responsible for infectious disease [ 46 , 55 , 73 ]. Reduced water flow can lead to an increase in bacterial concentration and reduced water quality as there is less water available to dilute these contaminants [ 64 , 74 ]. Climate change-related impacts on water, including drought, are recognized as leading to greater risks from water-borne diseases such as hepatitis A [ 52 ]. However, there appears to be limited discussion of the relationship between drought and water-borne pathogens and the interactions between climate change and water-borne diseases are complex. In the U.S., the survival and multiplication of bacteria leading to leptospirosis outbreaks may have been altered by temperature and pH changes resulting from the stagnation associated with drought [ 75 ]. Similarly, although Naegleria fowleri is rare, increased water temperature during drought can allow for the amoeba to grow in greater concentrations [ 46 ]. Drought conditions may concentrate pathogens while at the same time inactivate them (e.g., Cryptosporidium oocysts) [ 54 ]. In Canada, it is important to note that the greatest burden from enteric, food and water-borne diseases, drawing on 2012 data, is associated with campylobacteriosis (10,174 cases), followed by salmonellosis (6828 cases) and giardiasis (3862 cases). Other reportable diseases in this category have a far smaller number of cases, such as shigellosis with 1,068 reported cases and approximately 500 cases each of E. coli and cryptosporidiosis in 2012 [ 76 ].In addition, while at its peak in 1970, there were 12,283 cases of hepatitis A reported in Canada; this has since fallen to approximately 200–300 cases per year [ 76 ]. Future changes to the frequency and/or severity of drought could modify the burden associated with these diseases in Canada. In coastal areas, drought can result in the contamination of groundwater drinking water sources by the intrusion of salt water [ 46 , 50 ]. The mingling of saline and fresh groundwater that occurs in such regions can progress inland affecting freshwater wells, under a number of circumstances, including drier periods [ 77 ]. Saltwater intrusion can be exacerbated by climate change and sea-level rise [ 50 , 77 ]. The combination of drought with extreme precipitation may also impact water-, food- and vector- borne diseases. The most commonly cited impact of climate change on water-borne disease is the anticipated effects of increasing frequencies and amounts of heavy rainfall impacting the likelihood that untreated and pathogen-carrying water is consumed by Canadians [ 52 ]. Drought can amplify the impacts of a subsequent extreme rainfall event. During a drought event the soil can become compacted, thereby increasing runoff and the likelihood of water contamination and water-borne illness from precipitation events that follow [ 46 , 54 , 55 , 78 ]. Conditions associated with drought, such as low rainfall, can also concentrate Giardia and Cryptosporidium cysts in groundwater sources and where water is stored. Water sources may then be contaminated by the cysts after rainfall events [ 79 ]. Evidence from the Prairies suggests that the prevalence of livestock farming may affect water-borne diseases that follow such events if manure is able to contaminate water sources (e.g., Eschericia (E). coli , Cryptosporidium , Giardia ) [ 80 , 81 , 82 , 83 ]. Campylobacteriosis outbreaks are also frequently associated with heavy rainfall, often when they follow a period of drought [ 54 ]. Although a much smaller risk for Canadians, with typically only 0–2 cases in Canada each year [ 76 ], the pattern of drought conditions followed by a wet spring could also activate anthrax [ 52 ]. 3.2.2. Treated Drinking Water During drought, there can be a drop in pressure at water treatment plants [ 52 ], which can increase turbidity leading to a greater potential risk of water contamination and consequently a greater need for water treatment [ 84 , 85 ]. The impact of drought on contaminant levels in source water has been discussed above. In the event that well water becomes unavailable during drought, turning to alternative water sources can also increase risks to health if appropriate safeguards for those treatment measures are not in place [ 86 ]. Inadequate water treatment is important as pathogens, including bacteria and viruses, are known to be taken up by particles in water [ 87 ]. Costs also increase with the additional treatment required to address higher levels of turbidity [ 88 , 89 ]. Changes to water flow and levels associated with climate change could further contribute to stress on water treatment systems and increase the cost of treatment [ 50 ]. In northern Ontario, some First Nations communities may already be near the limit of their capacity in this regard [ 50 ]. 3.2.3. Recreational Water Drought can also affect the quantity and quality of recreational water [ 59 , 60 , 61 , 65 , 90 , 91 , 92 , 93 , 94 ]. The hot and dry weather that can be associated with drought is thought to increase recreational water use [ 46 , 75 ]. At the same time, the increased concentration of pathogens encouraged under these conditions, as described above, are thought to increase the likelihood of infection among recreational water users [ 46 ]. Although leptospirosis cases and outbreaks are frequently associated with heavy rainfall events [ 57 ], in the U.S., they have also been associated with drought and with swimming in lakes [ 75 ]. In Canada, even under typical (non-drought) conditions, acute gastrointestinal illness (AGI) is a risk for swimmers, with the risk of AGI at approximately 3%–8% [ 95 ]. Although the ingestion of water during recreational use would be lower than that from the drinking water sources discussed above, recreational water users could be exposed to water-borne pathogens that lead to infection through direct contact or inhalation [ 96 , 97 ]. 3.3. Sanitation/Hygiene Water plays a key role in maintaining hygiene, which is directly associated with preventing disease [ 46 , 60 , 98 ]. Health care facilities may depend heavily on water to protect the health of patients and workers [ 46 ]. Studies outside of Canada suggest strong linkages between changes in precipitation, drought and diarrheal disease [ 99 , 100 ]. The World Health Organization has estimated that globally, the risk factor "water, sanitation and hygiene" accounted for 4% of all deaths [ 101 ]. A global cross-sectional study found the incidence of diarrhoea in children under the age of five years increased 4% with each 10mm per month decrease in precipitation [ 102 ]. Modeling done on a global scale by Motoshita et al. [ 103 ] based on recent historical data suggested that health damage from infectious diseases resulting from domestic water scarcity would be relatively low in the U.S. compared with other regions, although the health damage could not be modeled for Canada. However, in England and Wales, both low and excessive rainfall over the short-term was linked to a rise in the incidence of diarrhoea related to Giardia , Cryptosporidium , E.coli , S. typhi , S. paratyphi , Campylobacter and Streptobacillus moniliformis (This is not a reportable disease in Canada and infections are thought to be lower in Canada and other Western countries than in the US; close to 200 cases of "rat bite fever" from Streptobacillus moniliformis had been documented in the U.S. at the time of a 2007 publication [ 333 ]) [ 104 ]. Furthermore, people displaced by drought are considered likely to experience loss of sanitation and routine hygiene [ 45 ]. Human behavior may play a role, since the relaxing of hygiene practices could lead to an increase in diarrheal illnesses during drought [ 74 ]. As described above, Giardia , Cryptosporidium , E. coli and Salmonella cause a considerable number of disease cases in Canada each year. Alteration of the rates of such diseases by drought could have implications for public health. 3.4. Food-borne Diseases The conditions associated with drought could impact food-borne illness. Hot and dry conditions often associated with drought are thought to favour the proliferation of some food-borne pathogens. The impact in Canada could be significant, as currently there are an estimated 11 million cases of food-borne illness annually across the country [ 105 ]. Among reportable diseases, 10,174 cases of campylobacteriosis and 6,828 cases of salmonellosis were reported in Canada in 2012 [ 76 ]. Illness caused by Clostridium perfringens is not nationally notifiable in Canada, but ranks among the most common food-borne diseases in the industrialized world [ 106 ]; in the U.S., C. perfringens is thought to cause close to one million cases of food-borne illness each year [ 107 ]. In some cases the environmental conditions created by drought are favorable to pathogens, such as C. perfringens , which does well in hot and dry conditions [ 52 ]. Campylobacterios is also expected to benefit from some seasonal conditions under climate change in the future [ 54 ]. However, the expected impacts of climate change-related drought conditions remain uncertain for other pathogens, including Salmonella , Listeria and Norovirus [ 54 ]. For example, under experimental desiccation Salmonella (typhimurium) lost infectivity and cultivability [ 108 ]. Meanwhile, in contrast to earlier studies, Zhang et al. [ 109 ], in comparing models to forecast salmonellosis cases in Adelaide (Australia) found decreased rainfall was associated with salmonellosis cases. It is possible that this finding was due to interaction with other variables such as temperature or the unique local climate [ 109 ]. For example, Ge et al. [ 110 ] suggest that under extreme conditions, including drought, when the soil contains high concentrations of Salmonella there may be greater internalization of the pathogen in lettuce relative to optimal irrigation conditions. During drought, cultivated crops, fish and shellfish consumed for food [ 52 ] may pose an increased risk of disease, which could have implications for human health [ 46 ]. If treated municipal sewage is used by the agricultural sector for irrigation or to process products due to water shortages there can be an increase in the risk of food-borne illness (e.g., Salmonella , E. coli ) [ 46 ]. Local food production practices and food imported from some countries may need to be monitored more closely under such conditions [ 46 ]. 3.5. Vector-Borne Diseases 3.5.1. West Nile and St. Louis Encephalitis Vector-borne diseases—particularly mosquito-borne ones—are frequently associated with increased precipitation [ 111 ] because standing water is needed for the development of immature stages of mosquitoes. Mosquitoes are important disease vectors in Canada as they can transmit West Nile virus (WNV). There were 110 total clinical cases of West Nile virus reported in Canada between January and the end of November, 2013, including three cases potentially related to travel outside the province where they were reported [ 112 ]. High precipitation in the prairie provinces in 2007 was thought responsible for an unprecedented abundance of Culex tarsalis mosquitoes that in turn led, later that year, to the largest ever outbreak of WNV in Canada (2,401 cases) [ 76 , 113 , 114 ]. Studies on the mosquitoes Cx. pipiens and Cx. restuans , which are the main WNV vectors in eastern Canada, also suggest that precipitation is positively associated with their abundance [ 115 ]. However, in some circumstances periods of drought have been directly associated with increased risk from mosquito-borne infections, such as WNV in southeastern, northeastern, central and western U.S., the Canadian prairies, as well as in Europe [ 115 , 116 , 117 , 118 , 119 , 120 , 121 , 122 ]. A number of explanations have been proposed for these observations, including impacts of drought on competitors and predators of WNV mosquito vectors [ 116 ] which can lead to mosquito population increases [ 115 ] and changes in communities and densities of the bird species that are reservoir hosts of WNV, resulting in increased WNV prevalence in mosquitoes [ 123 , 124 ]. Semi-permanent rather than permanent wetlands are thought to be particularly prone to the effects of drought given the fluctuations they experience in response to precipitation. Irrigation during times of drought may create new breeding sites for WNV mosquito vectors and thus shift the local geographic focus of risk from WNV [ 125 ]. In the matrix of mosquito breeding habitat found in urban and suburban environments, small numbers of locations can contribute to mosquito breeding and abundance and these are frequently man-made water courses, sewage overflow systems and holding pools, etc . [ 126 ], which are likely to continue to hold water during drought events. Associations of WNV incidence with droughts or low precipitation (e.g., in Virginia, U.S. [ 127 ]) may be due to reduced flushing of mosquito larvae and eggs from breeding sites in urban areas during rainfall; this flushing reduces mosquito abundance [ 128 ]. Frequently, the effect of drought on vector-borne diseases is seen when drought is followed by periods of wetter weather [ 129 , 130 ], but the relationships amongst mosquito abundance, infection prevalence, WNV risk and seasonal variations in drought and rainfall can be complex, geographically variable, habitat/environment-specific and difficult to predict [ 131 ]. In addition to altering the mosquito vectors' environment, drought can impact risk from mosquito-borne diseases by affecting contact rates between the vectors and reservoir hosts, thereby influencing the prevalence of infection in mosquitoes [ 46 , 118 ]. Evidence for this mechanism comes particularly from studies on mosquito-bird transmitted viruses [ 118 ] including WNV [ 129 ] and St. Louis encephalitis (SLE) virus [ 132 ].The congregation of reservoir host birds around dwindling water sources may be one way that drought affects WNV infection prevalence in mosquitoes by increasing rates of contact between avian reservoirs and mosquitoes thereby amplifying transmission [ 129 , 132 ]. In California (U.S.) drought may have contributed to an increase in human WNV cases by reducing populations of non-competent host birds, allowing for greater contact of mosquitoes with reservoir-competent host birds and possibly humans [ 131 ]. In the southeastern U.S., SLE outbreaks are associated with rainfall following drought, which is largely due to the behavior of the main vector Culex nigripalpis [ 133 ]. In drought conditions, gravid Culex nigripalpis females will increase in numbers around standing water as they wait for rainfall to feed and lay their eggs together, resulting in synchronous production of mosquitoes [ 134 ]. In addition, enhanced transmission due to the concentrating of mosquitoes and bird reservoirs in habitats that hold water is thought to help amplify the SLE virus [ 132 , 135 ]. 3.5.2. Lyme Disease Lyme disease, the most frequent vector-borne disease diagnosed in North America, is emerging in Canada [ 136 , 137 ]. Drought has a consistently negative effect on risk from tick-borne diseases such as Lyme disease because of (i) direct killing of ticks by desiccation in severe drought; (ii) reduction in host-finding success by reducing the time that ticks can spend seeking hosts outside of moisture-holding refugia in the habitats in which they occur; and (iii) increased energy expenditure due to ticks returning more frequently to moisture-holding refugia to rehydrate [ 138 , 139 ]. 3.5.3. Directly Transmitted Wildlife-borne Zoonoses The effect of drought on zoonoses transmitted directly from wildlife to humans is likely a reduction in the abundance of wild animal hosts and consequently, reduced pathogen transmission. When drought is followed by rainfall, wild animal host densities may increase again, but as most individuals will be naïve of infection, the potential for epidemics exists. Rodent-borne hantavirus is a paradigm for this in which drought followed by heavy rainfall (particularly associated with El Nino) is thought to drive epidemics in the southwestern U.S. [ 53 , 140 , 141 ]. 3.5.1. West Nile and St. Louis Encephalitis Vector-borne diseases—particularly mosquito-borne ones—are frequently associated with increased precipitation [ 111 ] because standing water is needed for the development of immature stages of mosquitoes. Mosquitoes are important disease vectors in Canada as they can transmit West Nile virus (WNV). There were 110 total clinical cases of West Nile virus reported in Canada between January and the end of November, 2013, including three cases potentially related to travel outside the province where they were reported [ 112 ]. High precipitation in the prairie provinces in 2007 was thought responsible for an unprecedented abundance of Culex tarsalis mosquitoes that in turn led, later that year, to the largest ever outbreak of WNV in Canada (2,401 cases) [ 76 , 113 , 114 ]. Studies on the mosquitoes Cx. pipiens and Cx. restuans , which are the main WNV vectors in eastern Canada, also suggest that precipitation is positively associated with their abundance [ 115 ]. However, in some circumstances periods of drought have been directly associated with increased risk from mosquito-borne infections, such as WNV in southeastern, northeastern, central and western U.S., the Canadian prairies, as well as in Europe [ 115 , 116 , 117 , 118 , 119 , 120 , 121 , 122 ]. A number of explanations have been proposed for these observations, including impacts of drought on competitors and predators of WNV mosquito vectors [ 116 ] which can lead to mosquito population increases [ 115 ] and changes in communities and densities of the bird species that are reservoir hosts of WNV, resulting in increased WNV prevalence in mosquitoes [ 123 , 124 ]. Semi-permanent rather than permanent wetlands are thought to be particularly prone to the effects of drought given the fluctuations they experience in response to precipitation. Irrigation during times of drought may create new breeding sites for WNV mosquito vectors and thus shift the local geographic focus of risk from WNV [ 125 ]. In the matrix of mosquito breeding habitat found in urban and suburban environments, small numbers of locations can contribute to mosquito breeding and abundance and these are frequently man-made water courses, sewage overflow systems and holding pools, etc . [ 126 ], which are likely to continue to hold water during drought events. Associations of WNV incidence with droughts or low precipitation (e.g., in Virginia, U.S. [ 127 ]) may be due to reduced flushing of mosquito larvae and eggs from breeding sites in urban areas during rainfall; this flushing reduces mosquito abundance [ 128 ]. Frequently, the effect of drought on vector-borne diseases is seen when drought is followed by periods of wetter weather [ 129 , 130 ], but the relationships amongst mosquito abundance, infection prevalence, WNV risk and seasonal variations in drought and rainfall can be complex, geographically variable, habitat/environment-specific and difficult to predict [ 131 ]. In addition to altering the mosquito vectors' environment, drought can impact risk from mosquito-borne diseases by affecting contact rates between the vectors and reservoir hosts, thereby influencing the prevalence of infection in mosquitoes [ 46 , 118 ]. Evidence for this mechanism comes particularly from studies on mosquito-bird transmitted viruses [ 118 ] including WNV [ 129 ] and St. Louis encephalitis (SLE) virus [ 132 ].The congregation of reservoir host birds around dwindling water sources may be one way that drought affects WNV infection prevalence in mosquitoes by increasing rates of contact between avian reservoirs and mosquitoes thereby amplifying transmission [ 129 , 132 ]. In California (U.S.) drought may have contributed to an increase in human WNV cases by reducing populations of non-competent host birds, allowing for greater contact of mosquitoes with reservoir-competent host birds and possibly humans [ 131 ]. In the southeastern U.S., SLE outbreaks are associated with rainfall following drought, which is largely due to the behavior of the main vector Culex nigripalpis [ 133 ]. In drought conditions, gravid Culex nigripalpis females will increase in numbers around standing water as they wait for rainfall to feed and lay their eggs together, resulting in synchronous production of mosquitoes [ 134 ]. In addition, enhanced transmission due to the concentrating of mosquitoes and bird reservoirs in habitats that hold water is thought to help amplify the SLE virus [ 132 , 135 ]. 3.5.2. Lyme Disease Lyme disease, the most frequent vector-borne disease diagnosed in North America, is emerging in Canada [ 136 , 137 ]. Drought has a consistently negative effect on risk from tick-borne diseases such as Lyme disease because of (i) direct killing of ticks by desiccation in severe drought; (ii) reduction in host-finding success by reducing the time that ticks can spend seeking hosts outside of moisture-holding refugia in the habitats in which they occur; and (iii) increased energy expenditure due to ticks returning more frequently to moisture-holding refugia to rehydrate [ 138 , 139 ]. 3.5.3. Directly Transmitted Wildlife-borne Zoonoses The effect of drought on zoonoses transmitted directly from wildlife to humans is likely a reduction in the abundance of wild animal hosts and consequently, reduced pathogen transmission. When drought is followed by rainfall, wild animal host densities may increase again, but as most individuals will be naïve of infection, the potential for epidemics exists. Rodent-borne hantavirus is a paradigm for this in which drought followed by heavy rainfall (particularly associated with El Nino) is thought to drive epidemics in the southwestern U.S. [ 53 , 140 , 141 ]. 3.6. Fungal Diseases Drought may encourage the growth of various fungi associated with health impacts. Warmer, drier summers may have contributed to the establishment of Cryptococcus gattii in Canada [ 142 ]. Low moisture has been associated with C. gattii soil colonization [ 143 ]. The fungus was first detected on Vancouver Island in 1999, with more than 100 human cases identified [ 142 ]. Between 1999 and 2007, there were 218 reported cases of C. gattii in British Columbia, the region with the largest C. gattii infected population in Canada and the world [ 144 ]. Drought may also positively affect the growth and dissemination of coccidioidomycosis (Valley fever). Rainfall patterns and timing (including length of the drought) have been identified as modifiers of the incidence of this disease [ 46 , 142 , 145 , 146 , 147 , 148 ]. Although the fungus is concentrated in the Southwestern U.S., recent cases that appeared to originate in Washington State (U.S.) suggest that it may be capable of moving to other areas [ 149 ]. Valley fever is presently rare in Canada but climate change may impact the incidence of the disease among Canadian travelers [ 150 ]. Since 1952, there have been a total of 128 cases of coccidiodomycosis in Canada. These Canadian cases were acquired during travel outside Canada [ 150 ]. Although associated with warming rather than drought, projections of increased precipitation in the winter months followed by reduced precipitation in the summer months for North America are thought to favor the dispersal of Blastomyces dermatitidis spores, which causes blastomycosis [ 142 ]. Currently, this fungus occurs mainly in northwestern Ontario [ 142 ]. Between 1994 and 2003, there were 309 cases of blastomycosis in Ontario with more than half of the cases occurring between 2001 and 2003 [ 151 ]. Geographically, northern Ontario accounted for 61% of the cases, followed by 21% from the Toronto region [ 151 ]. 3.7. Respiratory Health 3.7.1. Dust, Particulates and Allergens Drought is often accompanied by dry, dusty conditions and dust storms, which can impact health [ 83 , 152 , 153 ]. Climate change can lead to an increase in fine particulate matter, allergen and dust concentrations in the air in drought-prone areas [ 154 ] which can have significant health impacts. A study of a lake desiccated through persistent drought in Saskatchewan found a greater prevalence of coughing, wheezing and eye and nasal irritation among residents living near the lake than among those in a control group [ 155 ]. PM 10 (air particle) increases during and following dust storms have also been found to cause small increases in emergency room visits for bronchitis and sinusitis in the U.S. [ 156 ] although dust storms were not associated with an increased risk of mortality [ 157 ]. Air-borne toxins from algal blooms, which can be increased due to drought conditions, can also irritate the respiratory system, nose and eyes [ 46 , 153 , 155 ]. It has been anecdotally suggested that there were numerous deaths from "dust pneumonia" related to dust inhalation during the drought of the 1930s in the U.S. [ 152 ]. In Ontario, Charron et al. [ 52 ] noted the possibility that drought with strong winds could lead to the transmission of Rickettsial diseases to humans from infected livestock. 3.7.2. Silo-Filler's Disease Dry conditions can increase nitrate levels in corn plants and encourage the accumulation of nitrogen dioxide (NO 2 ), thereby elevating the risk of "silo-filler's disease" [ 158 ]. A greater number of "silo-filler's disease" cases were reported in New York State after a dry growing season in 1995 [ 158 ]. There are approximately five cases of silo-filler's disease per 100,000 silo-associated farm workers in New York State annually [ 159 ]. 3.7.3. Wildfires and Extreme Heat Drought has been associated with wildfires in both Canada and the U.S. and is expected to increase the number of wildfires in the Prairies specifically [ 46 , 51 , 83 ]. The international literature links wildfires to respiratory, cardiovascular, ophthalmic and psychiatric illnesses [ 43 , 160 , 161 ]. In Brazil, a study of young children linked an increase in respiratory disease incidence with drought-related wildfires [ 162 ]. During a prolonged forest fire in British Columbia, Canada in 2003 one community experienced a 46%–78% increase in physician visits for respiratory diseases [ 163 ]. Wildfires and other extreme weather events that might lead to displacement can be a source of significant stress for those directly affected [ 51 , 80 , 83 , 160 , 164 ]. The significant loss of property and infrastructure during a wildfire can also have negative impacts on mental health [ 51 ] and may lead to post-traumatic stress disorder [ 160 , 164 ]. More recently, in 2011 the northern Prairies experienced severe drought and catastrophic fires during which Slave Lake, Alberta lost one-third of homes and businesses to wildfire [ 165 , 166 ].In the Northwest Territories, a dry winter combined with only half the typical amount of rainfall over the summer of 2014 and above average temperatures, was accompanied by the worst fire season in 30 years [ 167 ]. Record warm and dry conditions in British Columbia over the same summer and preceding winter also set the stage for a wildfire season that saw more than seven and a half times the 20 year average in land area burned in 2014 [ 167 ]. Wildfires are expected to become more frequent and severe in many parts of Canada with climate change [ 168 ]. Further, drought has been associated with periods of extreme heat which can present a range of health impacts that need to be addressed. In addition to its direct impact on human health, extreme heat can add to drought-related stress on agriculture with knock-on effects discussed below. For example, an extreme heat event that took place in Russia over July and August, 2010 reduced the country's grain yield by 25%, resulting in increases in food prices and losses estimated at $15 billion [ 169 ]. 3.7.1. Dust, Particulates and Allergens Drought is often accompanied by dry, dusty conditions and dust storms, which can impact health [ 83 , 152 , 153 ]. Climate change can lead to an increase in fine particulate matter, allergen and dust concentrations in the air in drought-prone areas [ 154 ] which can have significant health impacts. A study of a lake desiccated through persistent drought in Saskatchewan found a greater prevalence of coughing, wheezing and eye and nasal irritation among residents living near the lake than among those in a control group [ 155 ]. PM 10 (air particle) increases during and following dust storms have also been found to cause small increases in emergency room visits for bronchitis and sinusitis in the U.S. [ 156 ] although dust storms were not associated with an increased risk of mortality [ 157 ]. Air-borne toxins from algal blooms, which can be increased due to drought conditions, can also irritate the respiratory system, nose and eyes [ 46 , 153 , 155 ]. It has been anecdotally suggested that there were numerous deaths from "dust pneumonia" related to dust inhalation during the drought of the 1930s in the U.S. [ 152 ]. In Ontario, Charron et al. [ 52 ] noted the possibility that drought with strong winds could lead to the transmission of Rickettsial diseases to humans from infected livestock. 3.7.2. Silo-Filler's Disease Dry conditions can increase nitrate levels in corn plants and encourage the accumulation of nitrogen dioxide (NO 2 ), thereby elevating the risk of "silo-filler's disease" [ 158 ]. A greater number of "silo-filler's disease" cases were reported in New York State after a dry growing season in 1995 [ 158 ]. There are approximately five cases of silo-filler's disease per 100,000 silo-associated farm workers in New York State annually [ 159 ]. 3.7.3. Wildfires and Extreme Heat Drought has been associated with wildfires in both Canada and the U.S. and is expected to increase the number of wildfires in the Prairies specifically [ 46 , 51 , 83 ]. The international literature links wildfires to respiratory, cardiovascular, ophthalmic and psychiatric illnesses [ 43 , 160 , 161 ]. In Brazil, a study of young children linked an increase in respiratory disease incidence with drought-related wildfires [ 162 ]. During a prolonged forest fire in British Columbia, Canada in 2003 one community experienced a 46%–78% increase in physician visits for respiratory diseases [ 163 ]. Wildfires and other extreme weather events that might lead to displacement can be a source of significant stress for those directly affected [ 51 , 80 , 83 , 160 , 164 ]. The significant loss of property and infrastructure during a wildfire can also have negative impacts on mental health [ 51 ] and may lead to post-traumatic stress disorder [ 160 , 164 ]. More recently, in 2011 the northern Prairies experienced severe drought and catastrophic fires during which Slave Lake, Alberta lost one-third of homes and businesses to wildfire [ 165 , 166 ].In the Northwest Territories, a dry winter combined with only half the typical amount of rainfall over the summer of 2014 and above average temperatures, was accompanied by the worst fire season in 30 years [ 167 ]. Record warm and dry conditions in British Columbia over the same summer and preceding winter also set the stage for a wildfire season that saw more than seven and a half times the 20 year average in land area burned in 2014 [ 167 ]. Wildfires are expected to become more frequent and severe in many parts of Canada with climate change [ 168 ]. Further, drought has been associated with periods of extreme heat which can present a range of health impacts that need to be addressed. In addition to its direct impact on human health, extreme heat can add to drought-related stress on agriculture with knock-on effects discussed below. For example, an extreme heat event that took place in Russia over July and August, 2010 reduced the country's grain yield by 25%, resulting in increases in food prices and losses estimated at $15 billion [ 169 ]. 3.8. Injuries Although severe injuries directly resulting from drought may be unlikely in mid and higher income countries [ 45 ], during periods of drought an increased incidence of spinal cord injuries has been recorded in the U.S. from diving into shallow water bodies. Accidents from debris can also be a concern as lower water levels can expose previously "hidden" hazards in bodies of water [ 46 , 170 , 171 ]. Affected individuals may have been seeking relief from warm temperatures accompanying dry conditions [ 171 ]. Although carelessness and alcohol may play a role, injured individuals were often familiar with local bodies of water and believed the water to be deeper than it actually was [ 170 , 171 ]. Dust storms, discussed above in the context of drought, have also been associated with traffic accidents on the Prairies [ 23 , 153 , 172 ]. Dust storms can reduce visibility for drivers [ 172 ]. 3.9. Food/Water Insecurity Drought may increase the difficulty that individuals face in obtaining adequate water. In the U.S. there have been reports of risk of drinking water shortages [ 173 ]. In Canada, there have been reports of costly water-related infrastructure such as water cisterns, septic tanks and wells damaged or at risk during drought [ 80 , 174 , 175 , 176 ]. The IPCC recognizes a broad range of impacts due to climate change and the altered frequencies of associated extreme events, such as drought and floods that can have negative effects on crop production [ 177 , 178 , 179 ]. Climate change can also alter the geographic distribution and productivity of crops and livestock and the distribution of pests and disease [ 180 , 181 , 182 ]. Food security can be affected through drought-related decreases in agricultural output, food shortages and higher food prices. Crop failures and reduced crop yields are of particular concern because they can have immediate and direct effects on both local and global food security [ 183 , 184 , 185 , 186 , 187 ].Climate change could lead to a decrease in the zinc and iron content of staple crops such as wheat [ 188 ]. Such a change in nutritional content could exacerbate iron deficiency, which occurs in both developed and developing countries [ 189 ]. Meanwhile, increasing homogeneity among crop species globally threatens food security with increased interdependence of food systems around the world and reduced genetic diversity [ 190 ].As a net exporter of food and the 9th largest exporter in the world [ 191 ], Canada is unlikely to experience the types of food shortages that could occur in other countries. However, food insecurity may be reflected in food price volatility caused by international market shortages [ 192 ], which could impact low income Canadians who find it difficult to pay higher prices for nutritious food. In addition, some individuals may be forced to use their resources to purchase water instead of food [ 46 , 49 ]. The media reported a surge in food bank users and pointed to high household spending on water in parts of California (U.S.) during their recent drought [ 193 ]. The widespread U.S. drought in 2012 was thought to be one of the factors contributing to an increase in food prices in countries, including Canada [ 194 , 195 ].In addition, threats from climate change to the food security of First Nations and Inuit populations in northern Canada are increasing [ 2 ]. While it appears to be complex, a recent Australian study suggests that drought may play a moderating role in the relationship between food insecurity and psychological distress, possibly increasing distress [ 196 ]. For those who rely on subsistence fishing, such as Aboriginal populations, health can be affected when drought affects the health of aquatic organisms [ 46 ]. Drought can increase the likelihood of disease spread and/or exposure to an accumulation of contaminants in the water for these aquatic organisms [ 46 ]. They can also be impacted by decreases in the quality and quantity of water during drought or through contamination introduced by the rains that follow drought [ 52 ]. 3.10. Socioeconomic Pathways Leading to Human Health Impacts Drought can affect social and economic wellbeing within a community through changes to water availability that can impact agriculture [ 176 ]. The Canadian agricultural sector produced crop and livestock products totaling over $41 billion in 2010 [ 197 ]. Costs associated with a drought can be significant. For example, total costs of drought across the Prairies in 1990 exceeded $800 million (equivalent 2011 Canadian dollars) [ 198 ]. The more recent drought of 2001–2002 was responsible for $3.6 billion in direct losses related to agricultural production [ 199 ]. Impacts on agriculture are passed on to the broader community and may lead to temporary unemployment [ 200 ] and increase stress levels [ 176 , 201 ]. The U.S. drought in the 1930s and its impact on agriculture contributed to the Great Depression; there was high unemployment, increased demands on government aid programs and community impacts as millions of people affected by the drought moved to other communities [ 202 ]. The impact of drought on different segments of a society can vary [ 203 ]. For example, the 2001–2002 drought brought significant impacts on agricultural production in Alberta's "Special Areas", but it was mainly an inconvenience for residents and businesses outside the agricultural sector [ 203 ]. Similarly, the 1987–1992 drought in California (U.S.) led to primarily behavioral rather than economic impacts for residents [ 200 ]. Loss of agricultural output in one area may drive up the demand and prices for farmers unaffected by the drought suggesting that some economic impacts may be highly localized [ 204 ]. Globally, the impact of climate change on agriculture in different regions and countries, such as other developed grain-exporting countries can affect the competitiveness of Canada's agricultural sector, particularly for wheat [ 174 ]. A widespread U.S. drought in 2012 increased crop prices, resulting in new record highs for Canadian farm incomes overall [ 205 ]. The international literature suggests that drought can impact communities by shifting resources from services such as public health to address more immediate needs (e.g., sanitation), as well as leading to conflict over water between groups and regions [ 2 , 42 , 59 , 90 , 206 , 207 , 208 ]. Recent media attention around conflicts over water in the western U.S. states among farmers, activists, developers and governments at state and local levels suggests that conflict over water in developed countries such as Canada could be a concern [ 209 ]. During periods of drought, water-related recreational activities may be cancelled or closed, such as during the 2001–2002 drought in the Prairies and the 1987–1992 drought in the U.S., which can have social impacts [ 174 , 176 , 200 ]. In the Canadian context, data from 2005 show that over 90% of the water withdrawn in the country was to support economic activity [ 12 ]. The impact of drought on economic activities overall, in particular through reductions in power generation capacity, could be considerable. For example, between 1999 (winter) and 2004 (fall), seasonal decreases in precipitation of up to 60% compared to normal conditions were observed in the Canadian cities of Calgary, Edmonton and Saskatoon [ 9 ]. Hydroelectricity generation accounts for 62% of electricity generation in Canada [ 12 ] (A large amount of hydroelectricity is generated in Canada and when in-stream use is taken into account, this industry accounts for the largest amount of water use in Canada [ 12 ]. The volume of water used for hydroelectricity generation was over 70 times the total of all other combined water use in Canada in 2005 [ 12 ]. Thermal (nuclear and fossil fuel) power generation and manufacturing are responsible for the largest and second largest total withdrawals of freshwater in Canada [ 12 , 334 ]). Shortfalls in this area could further amplify costs and stresses associated with drought, for example through the lack of revenue or the need to purchase alternative energy (see section 3.12.2 for additional discussion of the health impacts of these shortfalls). During the 2001 drought, hydroelectric power generation in Saskatchewan fell to 66% of the past four year average in the province resulting in power being purchased from other jurisdictions to meet electricity needs [ 174 ]. During the 1987–1992 drought in California (U.S.), hydroelectricity generation fell to under 60% of normal levels [ 200 ]. Additional energy was purchased from more expensive sources [ 200 ] and led to more than a $3 billion increase in electricity costs for consumers [ 204 ]. Drought may exacerbate the substantial impacts anticipated with climate change. The climate change-related decrease in water levels anticipated across the Great Lakes is expected to result in direct economic impacts between 2014 and 2030 of $9.61 billion (2012 USD) due to impacts on recreation, shipping and harbours, hydroelectricity generation, waterfront property values and groundwater use [ 210 ]. 3.11. Mental Health Drought and its associated financial burden can lead to stress [ 51 , 176 , 201 , 211 , 212 , 213 , 214 ] and is associated with physical and mental health problems, such as anxiety, emotional and psychological distress, loss and grief [ 46 , 215 , 216 , 217 , 218 ]. The pathways leading to these impacts are illustrated in Figure 2 . For farm families, financial stress [ 51 ] and drought [ 219 ] can amplify the other stressors they experience. For consumers with low socio-economic status, stress related to an increase in food costs or the decreased availability of food can also negatively impact mental health [ 46 ]. Figure 2 Pathways through which climate change-related disasters can affect mental health. Source: Reproduced from Berry et al. [ 220 ]. Australia has produced numerous studies regarding drought's relationship to mental health. Negative drought-related impacts on mental health have been identified for older women and Aboriginal groups in Australia [ 216 , 221 ], as well as for rural Australian youth [ 222 ], with emotional stress increasing with the length of the drought [ 217 ]. Dryland salinity, a phenomenon similar to drought, has also been associated with an increased risk of hospitalization for depression in Australia [ 223 ]. Although Stehlik et al. [ 214 ] found that drought-related stress negatively impacted relationships within farm families (Central Queensland, New South Wales, Australia), a later large-scale study across Australian agricultural communities did not find family relationships were negatively impacted by drought [ 224 ]. For drought that was both constant and long (although not for either alone), there was a difference in the distress levels in rural and urban populations in Australia, with relatively higher levels among rural populations [ 225 ]. The degradation of the physical environment associated with drying could also contribute to psychological distress for those who feel connected to the landscape (solastalgia) [ 226 ]. Research from New South Wales (Australia) found a statistically significant relationship between drought and suicide from 1964 to 2001 [ 227 ]. Limited studies suggest that drought is a contributing factor to suicide among men in rural areas in Australia [ 228 , 229 ]. However, Page and Fragar [ 230 ] found a greater rate of suicide between 1988 and 1997 among male farm managers and agricultural labourers than among the male Australian population or the rural population overall, independent of drought. 3.12. Populations Vulnerable to the Health Impacts of Drought Populations at higher risk to the health impacts of climate change have been identified in the literature. They include infants and children, elderly people, the socially and economically disadvantaged, pregnant women, people who spend time outdoors and those with chronic diseases and compromised immune systems ( Table 1 ) [ 1 , 59 , 231 , 232 , 233 ]. Overlap exists between populations vulnerable to the health impacts of drought and populations at risk of adverse impacts on health resulting from other extreme weather events, including those expected to increase in frequency and intensity with climate change [ 46 , 59 ]. For example, people of low-socioeconomic status have more difficulty taking protective actions during extreme heat events [ 234 ] and may also suffer greater impacts from drought [ 49 ]. As an important pathway for increased exposure to injuries, air pollution, food, water and vector-borne diseases as discussed above, drought is associated with many higher risk populations. The existence of common vulnerability factors to multiple climate-related vulnerabilities increases overall risk to public health, but also offers significant opportunities to take adaptive actions that convey benefits among a broad spectrum of risk areas, including drought. For example, Cheng and Berry [ 235 ] identify efforts to enhance social capital as a climate change adaptation that has multiple health co-benefits. ijerph-12-08359-t001_Table 1 Table 1 Vulnerability to climate-sensitive health outcomes by subpopulation [ 233 ]. Groups with Increased Vulnerability Climate-related Vulnerabilities Infants and children Heat stress, air pollution, water-borne/food-borne diseases, vector-borne diseases, malnutrition Pregnant women Heat stress, extreme weather events, water-borne/food-borne diseases, vector-borne diseases Elderly people and people with chronic medical conditions Heat stress, air pollution, extreme weather events, water-borne/food-borne diseases, vector-borne diseases Impoverished/low socioeconomic status Heat stress, air pollution, extreme weather events, water-borne/food-borne diseases, vector-borne diseases Outdoor workers Heat stress, air pollution, vector-borne diseases, ultraviolet light (UV) exposure The following sections describe how specific populations may be at increased vulnerability to the health impacts of drought. 3.12.1. Agricultural Workers and Rural Populations Agricultural workers and others who rely on water and/or rainfall for their livelihoods are at greater risk of experiencing the mental health impacts of drought [ 46 , 51 , 83 , 174 , 176 , 212 ]. Over time new agricultural practices and technologies have increased agricultural output, but have also led to increased costs and vulnerability for farmers during drought [ 83 ]. Given their occupational environment, agricultural workers are at increased risk of "silo-filler's disease" (NO 2 gas poisoning within silos) [ 158 ]. Horticulturalists and people working for nurseries, garden supply stores and recreational facilities have also been identified as being at higher risk of negative health outcomes [ 46 ]. Research in Australia suggests that drought-related stress on agricultural workers can be widespread. A survey of rural farmers in New South Wales, Australia found that 71.8% of rural farmers reported stress due to drought [ 213 ]. Limited evidence suggests that drought is a contributing factor to suicide among men in rural areas [ 228 , 229 ]. In some studies, younger farmers were found to experience higher levels of stress than their older counterparts between 55–64 years but not their oldest counterparts aged 65–74 years [ 212 , 236 ]. The mental health impacts of drought can extend beyond farmers themselves to their families [ 237 ]. Rural Australian youth have experienced negative mental health effects during prolonged drought [ 222 ], with their emotional stress increasing with the length of the drought [ 217 ]. Older women and Aboriginals have also been identified as being more vulnerable to these impacts [ 216 , 221 ]. In general, rural and urban populations can experience different levels of drought-related distress. For drought that was both constant and long (although not for either alone) in Australia, distress levels in rural populations were observed to be higher relative to urban populations [ 225 ]. 3.12.2. People with Chronic Health Conditions, Pregnant Women, Young Children and Older Adults People with chronic health conditions are at increased risk of health impacts associated with air quality compromised by drought including the effects of wildfires [ 46 , 47 , 49 , 83 , 154 , 155 ]. This group is also thought to be at greater risk of drought-related infectious diseases [ 49 ]. Pregnant and nursing women, dialysis patients and immune-compromised individuals are considered to be at increased risk from health impacts of drought and heat [ 47 ]. Pregnant women in their third trimester, diabetics and immunosuppressed individuals were thought to have increased risk of developing a disseminated case of Valley fever [ 238 ]. Young children and seniors are at increased risk from drought-related water-borne diseases [ 49 ]. In addition to the very young (particularly those who have received rotavirus vaccines) and the elderly, those who are immunocompromised are at greater risk of developing acute gastrointestinal illness from water-borne diseases [ 95 ]. Very young children and seniors overlap with those susceptible to heat as identified above and may also be at greater risk from decreases in water quality and availability [ 60 ]. Children under five years of age and adults over 50 years old are also at increased risk of developing a disseminated case of Valley fever [ 146 ]. Individuals under institutional care (e.g., in hospitals, nursing homes) who depend on electrical equipment may be impacted by hydroelectricity power shortages during a drought [ 46 ]. Drought can also lead to disruptions of heating and cooling systems that rely on water to function, and these disruptions may result in health impacts on individuals when it is not possible to maintain health-protective temperatures [ 46 ]. 3.12.3. People with Low Socioeconomic Status People with low socioeconomic status are at greater risk of health impacts from climate change [ 59 ]. For example, families who are already facing challenges meeting their needs or who engage in subsistence fishing or farming may be at risk of food insecurity with drought [ 46 , 49 ]. In Canada, 8.3% of households were food insecure in 2011–2012, with 36.7% of households in Nunavut food insecure during this period [ 30 ]. 3.12.4. People Engaged in Activities Where Risk May Increase with Drought Individuals who spend more time outdoors have increased exposure to certain drought-related risks including those for respiratory diseases resulting from exposures related to wildfires and increased dust (e.g., dust-borne Valley fever) [ 59 , 146 ]. Exposure to vector-borne diseases, such as WNV or Eastern Equine Encephalitis, that may be linked to drought is increased in people spending greater amounts of time outdoors for recreation and/or due to their occupation (e.g., agricultural workers, parks services), placing them at greater risk of contracting these diseases [ 47 , 76 ]. Although the overall risk is considered to be low, individuals with greater exposure to the droppings of infected rodents are at increased risk of contracting rodent-borne diseases (e.g., Hantavirus) [ 239 ]. Recreational water users are at greater risk of contracting water-borne diseases, especially from pathogens favored by warm, shallow waters that often accompany drought [ 46 , 49 , 75 ]. In addition to biological pathogens, they may also have greater exposure to chemical contaminants [ 46 ]. Spinal cord injuries related to diving into shallow waters largely occur in males between 15–25 years of age [ 170 ]. 3.12.1. Agricultural Workers and Rural Populations Agricultural workers and others who rely on water and/or rainfall for their livelihoods are at greater risk of experiencing the mental health impacts of drought [ 46 , 51 , 83 , 174 , 176 , 212 ]. Over time new agricultural practices and technologies have increased agricultural output, but have also led to increased costs and vulnerability for farmers during drought [ 83 ]. Given their occupational environment, agricultural workers are at increased risk of "silo-filler's disease" (NO 2 gas poisoning within silos) [ 158 ]. Horticulturalists and people working for nurseries, garden supply stores and recreational facilities have also been identified as being at higher risk of negative health outcomes [ 46 ]. Research in Australia suggests that drought-related stress on agricultural workers can be widespread. A survey of rural farmers in New South Wales, Australia found that 71.8% of rural farmers reported stress due to drought [ 213 ]. Limited evidence suggests that drought is a contributing factor to suicide among men in rural areas [ 228 , 229 ]. In some studies, younger farmers were found to experience higher levels of stress than their older counterparts between 55–64 years but not their oldest counterparts aged 65–74 years [ 212 , 236 ]. The mental health impacts of drought can extend beyond farmers themselves to their families [ 237 ]. Rural Australian youth have experienced negative mental health effects during prolonged drought [ 222 ], with their emotional stress increasing with the length of the drought [ 217 ]. Older women and Aboriginals have also been identified as being more vulnerable to these impacts [ 216 , 221 ]. In general, rural and urban populations can experience different levels of drought-related distress. For drought that was both constant and long (although not for either alone) in Australia, distress levels in rural populations were observed to be higher relative to urban populations [ 225 ]. 3.12.2. People with Chronic Health Conditions, Pregnant Women, Young Children and Older Adults People with chronic health conditions are at increased risk of health impacts associated with air quality compromised by drought including the effects of wildfires [ 46 , 47 , 49 , 83 , 154 , 155 ]. This group is also thought to be at greater risk of drought-related infectious diseases [ 49 ]. Pregnant and nursing women, dialysis patients and immune-compromised individuals are considered to be at increased risk from health impacts of drought and heat [ 47 ]. Pregnant women in their third trimester, diabetics and immunosuppressed individuals were thought to have increased risk of developing a disseminated case of Valley fever [ 238 ]. Young children and seniors are at increased risk from drought-related water-borne diseases [ 49 ]. In addition to the very young (particularly those who have received rotavirus vaccines) and the elderly, those who are immunocompromised are at greater risk of developing acute gastrointestinal illness from water-borne diseases [ 95 ]. Very young children and seniors overlap with those susceptible to heat as identified above and may also be at greater risk from decreases in water quality and availability [ 60 ]. Children under five years of age and adults over 50 years old are also at increased risk of developing a disseminated case of Valley fever [ 146 ]. Individuals under institutional care (e.g., in hospitals, nursing homes) who depend on electrical equipment may be impacted by hydroelectricity power shortages during a drought [ 46 ]. Drought can also lead to disruptions of heating and cooling systems that rely on water to function, and these disruptions may result in health impacts on individuals when it is not possible to maintain health-protective temperatures [ 46 ]. 3.12.3. People with Low Socioeconomic Status People with low socioeconomic status are at greater risk of health impacts from climate change [ 59 ]. For example, families who are already facing challenges meeting their needs or who engage in subsistence fishing or farming may be at risk of food insecurity with drought [ 46 , 49 ]. In Canada, 8.3% of households were food insecure in 2011–2012, with 36.7% of households in Nunavut food insecure during this period [ 30 ]. 3.12.4. People Engaged in Activities Where Risk May Increase with Drought Individuals who spend more time outdoors have increased exposure to certain drought-related risks including those for respiratory diseases resulting from exposures related to wildfires and increased dust (e.g., dust-borne Valley fever) [ 59 , 146 ]. Exposure to vector-borne diseases, such as WNV or Eastern Equine Encephalitis, that may be linked to drought is increased in people spending greater amounts of time outdoors for recreation and/or due to their occupation (e.g., agricultural workers, parks services), placing them at greater risk of contracting these diseases [ 47 , 76 ]. Although the overall risk is considered to be low, individuals with greater exposure to the droppings of infected rodents are at increased risk of contracting rodent-borne diseases (e.g., Hantavirus) [ 239 ]. Recreational water users are at greater risk of contracting water-borne diseases, especially from pathogens favored by warm, shallow waters that often accompany drought [ 46 , 49 , 75 ]. In addition to biological pathogens, they may also have greater exposure to chemical contaminants [ 46 ]. Spinal cord injuries related to diving into shallow waters largely occur in males between 15–25 years of age [ 170 ]. 3.13. Adaptation to the Human Health Impacts of Drought (Direct) A broad range of adaptation strategies are available that help reduce the health impacts of drought. They generally attempt to reduce the impacts directly or they modify the environmental services or socioeconomic pathways through which drought affects health ( Table 2 ). However, there has been little evaluation of the effectiveness of many of these adaptations [ 183 , 240 ]. Limited knowledge of the effectiveness of current measures to reduce health risks from drought increases the possibility of "maladaptation" as communities and individuals prepare for climate change. Ensuring that adaptation strategies are developed with evidence-based information and through consultation with a wide range of partners mitigates the risk of maladaptation. ijerph-12-08359-t002_Table 2 Table 2 Examples of adaptations to reduce the health impacts of drought. Adaptations to the human health impacts Water quality and water treatment monitoring Disease surveillance and monitoring Air quality monitoring and warnings Public education and outreach Mental health prevention and awareness programs Adaptations to address changes to environmental services or socioeconomic pathways Mainstreaming climate change into source water protection planning Enhancing water service delivery and watershed management Water capture and storage Increasing water use efficiency and water conservation Modifying or adopting new agricultural practices, such as increasing reliance on crop irrigation Introducing drought-resistant crops and diversifying crops and income sources Using seasonal climate outlooks for crop planning Developing watershed management and climate change adaptation plans Insurance and other financial assistance programs Irrigation and altering the timing and use of water 3.13.1. Monitoring Water Quality and Water Treatment In addition to monitoring indicators for drought itself [ 47 ] or drought-related drivers of health impacts [ 241 ], increased water quality testing and monitoring of drinking water and/or recreational water during droughts can help safeguard health [ 43 , 59 , 65 , 74 , 92 ]. For example, the province of Quebec has begun to adapt by increasing the monitoring of surface water and adopting more stringent standards for water quality surveillance. Water treatment is also an important adaptation to protect against water-borne disease [ 59 ] and boil-water advisories may be a supportive tool, although they are designed to address fecal and not algal or chemical contamination. During the 2001 drought within the Kanai Blood Indian Reserve in Alberta (Canada), there was an increase in the number of boil water advisories called to protect residents against potential contamination [ 174 ]. 3.13.2. Disease Surveillance and Monitoring Surveillance and monitoring of diseases sensitive to drought such as those caused by E. coli , C. perfringens and WNV are important adaptations [ 43 , 52 , 53 , 56 , 74 , 94 , 242 , 243 , 244 ]. Surveillance and monitoring can help to direct the provision of medical services and the development of needed vaccines to reduce risks to health [ 59 , 74 , 94 , 242 , 243 ]. Modeling could play a role alongside surveillance [ 125 ]. Control of disease vectors may also be implemented [ 59 ]. Primary care physicians can be part of monitoring efforts and help identify and address cases of malnutrition during drought episodes [ 207 ]. 3.13.3. Air Quality Monitoring and Warnings When drought threatens air quality, for example by increasing particulate matter, air quality monitoring and warnings for the public can help reduce health risks [ 51 ]. In over 60 communities and 10 provinces across Canada, the Air Quality Health Index (AQHI) provides individuals with information on the immediate health risks associated with local air quality [ 245 ]. The AQHI includes information on how to reduce exposure to air pollution that is tailored to the general public and to vulnerable populations (e.g., parents with children and infants, seniors and those with cardiovascular and respiratory diseases). It is regularly updated and accessible through the Internet [ 246 ]. However, it may be challenging to capture a quickly-developing event such as a forest fire smoke episode through the Index. Environment Canada informs the public of severe weather by providing special weather statements, advisories and watches [ 247 ]. In responding to wildfires in western Canada in 2014, Environment Canada issued alerts through wildfire smoke warnings and special air quality statements [ 248 ]. Provincial and community-level agencies further monitor events such as forest fires and provide information to those affected through a number of platforms, including websites and social media [ 249 ]. 3.13.4. Public Education and Outreach Public education is a key adaptation to protect health from drought. Personal protection measures that can be promoted by public health officials can include reducing water- and vector-borne disease exposures [ 59 ], wearing respiratory masks on days when ambient air quality could impact health [ 51 ], and the promotion of hand washing [ 74 ]. Education for travelers to and from Canada can raise awareness about diseases such as malaria [ 80 ], or tuberculosis, which may be increasingly transmitted under social conditions created by drought in developing countries [ 52 ]. To prevent spinal cord injuries resulting from diving in shallow waters, some jurisdictions have implemented public education, including the posting of notices where diving is not safe [ 170 ]. For example, brain and spinal cord injury prevention programs have been carried out in Saskatchewan and Ontario (Canada) and in Washington State (U.S.) [ 250 , 251 , 252 ]. However, evaluations of these programs suggest that program delivery needs to be well designed and robust to be effective. Although the initiative in Saskatchewan was found to change knowledge among students, in the Toronto (Ontario) case only 16% of students in schools where the program video was distributed viewed the information [ 250 ]. The one time intervention in Washington State, where students attended a session which included a video, lecture and personal testimony around traumatic brain and spinal cord injury, had little impact on knowledge or behavior [ 252 ]. Similarly, evidence suggests that boil water advisories may not be adhered to by everyone. During the gastroenteritis outbreak in Walkerton, Ontario (Canada), some residents under the boil water advisory continued to use and occasionally drink the water without boiling it [ 85 ]. The link between water and hygienic practices should be considered when developing drought adaptation measures. Conservation should take into consideration on-farm agricultural practices, regional water management and treatment and allow for and encourage individual hygienic practices such as hand washing [ 46 ]. In the later example, homes and businesses could also promote both water conservation and hand washing by installing low-flow faucet aerators [ 46 ]. Education of health care providers can also benefit efforts to address drought impacts on health [ 53 ]. 3.13.5. Addressing Mental Health Impacts Adaptive actions may be taken by individuals and professionals within primary care and community services settings to reduce the mental health impacts of drought. Personal hopefulness and positive attitudes towards the rural lifestyle have been highlighted as useful adaptations to drought [ 218 , 253 ]. According to Caldwell and Boyd [ 254 ] rural Australian families used a combination of both positive (e.g., problem-focused coping, optimism and positive appraisal) and less beneficial strategies (e.g., cognitive dissonance, denial and avoidance of negative social influences) in their response to drought. Primary care practitioners can provide support to reduce symptoms of mental illness exacerbated by drought and ineffective coping mechanisms such as substance abuse. They may also provide information about financial support resources or community drought support groups [ 207 , 255 ]. Mental health programs in schools and in occupational settings and suicide prevention and awareness programs have also been identified as possible adaptation measures [ 59 , 216 , 222 , 256 , 257 ]. Newspapers may be another pathway to communicate broadly about mental health during drought [ 258 ].In Australia rural support workers help traditional mental health professionals identify farmers experiencing mental distress and refer them to support services [ 259 ]. Training in mental health first aid can also increase the capacity of rural support workers to address mental health needs in drought-affected rural areas [ 260 ]. One study suggested that the Rural Adversity Mental Health Program (RAMHP), which is based on a community development model, is effective in building capacity and resiliency to cope with the impacts of drought in rural New South Wales (Australia) [ 261 ]. Linking the services and programs focused on mental health, community or social needs with those on practical assistance for farming communities is considered key for addressing mental health concerns associated with or exacerbated by drought [ 262 ]. Access to support services for older individuals is another consideration [ 256 ]. Community-building [ 263 ] and events which increase social capital can further help decrease psychological stress [ 254 ]. High levels of social capital in many Canadian (Prairie) rural communities [ 176 , 264 ], based upon high levels of trust, networking and participation in organizations help reduce stress associated with drought [ 176 , 264 ]. However, Australian farm families faced heavier workloads during the drought of 2002–2003, which left less time for community participation, volunteer activities and hence could negatively impact social capital and the community [ 201 ]. 3.13.1. Monitoring Water Quality and Water Treatment In addition to monitoring indicators for drought itself [ 47 ] or drought-related drivers of health impacts [ 241 ], increased water quality testing and monitoring of drinking water and/or recreational water during droughts can help safeguard health [ 43 , 59 , 65 , 74 , 92 ]. For example, the province of Quebec has begun to adapt by increasing the monitoring of surface water and adopting more stringent standards for water quality surveillance. Water treatment is also an important adaptation to protect against water-borne disease [ 59 ] and boil-water advisories may be a supportive tool, although they are designed to address fecal and not algal or chemical contamination. During the 2001 drought within the Kanai Blood Indian Reserve in Alberta (Canada), there was an increase in the number of boil water advisories called to protect residents against potential contamination [ 174 ]. 3.13.2. Disease Surveillance and Monitoring Surveillance and monitoring of diseases sensitive to drought such as those caused by E. coli , C. perfringens and WNV are important adaptations [ 43 , 52 , 53 , 56 , 74 , 94 , 242 , 243 , 244 ]. Surveillance and monitoring can help to direct the provision of medical services and the development of needed vaccines to reduce risks to health [ 59 , 74 , 94 , 242 , 243 ]. Modeling could play a role alongside surveillance [ 125 ]. Control of disease vectors may also be implemented [ 59 ]. Primary care physicians can be part of monitoring efforts and help identify and address cases of malnutrition during drought episodes [ 207 ]. 3.13.3. Air Quality Monitoring and Warnings When drought threatens air quality, for example by increasing particulate matter, air quality monitoring and warnings for the public can help reduce health risks [ 51 ]. In over 60 communities and 10 provinces across Canada, the Air Quality Health Index (AQHI) provides individuals with information on the immediate health risks associated with local air quality [ 245 ]. The AQHI includes information on how to reduce exposure to air pollution that is tailored to the general public and to vulnerable populations (e.g., parents with children and infants, seniors and those with cardiovascular and respiratory diseases). It is regularly updated and accessible through the Internet [ 246 ]. However, it may be challenging to capture a quickly-developing event such as a forest fire smoke episode through the Index. Environment Canada informs the public of severe weather by providing special weather statements, advisories and watches [ 247 ]. In responding to wildfires in western Canada in 2014, Environment Canada issued alerts through wildfire smoke warnings and special air quality statements [ 248 ]. Provincial and community-level agencies further monitor events such as forest fires and provide information to those affected through a number of platforms, including websites and social media [ 249 ]. 3.13.4. Public Education and Outreach Public education is a key adaptation to protect health from drought. Personal protection measures that can be promoted by public health officials can include reducing water- and vector-borne disease exposures [ 59 ], wearing respiratory masks on days when ambient air quality could impact health [ 51 ], and the promotion of hand washing [ 74 ]. Education for travelers to and from Canada can raise awareness about diseases such as malaria [ 80 ], or tuberculosis, which may be increasingly transmitted under social conditions created by drought in developing countries [ 52 ]. To prevent spinal cord injuries resulting from diving in shallow waters, some jurisdictions have implemented public education, including the posting of notices where diving is not safe [ 170 ]. For example, brain and spinal cord injury prevention programs have been carried out in Saskatchewan and Ontario (Canada) and in Washington State (U.S.) [ 250 , 251 , 252 ]. However, evaluations of these programs suggest that program delivery needs to be well designed and robust to be effective. Although the initiative in Saskatchewan was found to change knowledge among students, in the Toronto (Ontario) case only 16% of students in schools where the program video was distributed viewed the information [ 250 ]. The one time intervention in Washington State, where students attended a session which included a video, lecture and personal testimony around traumatic brain and spinal cord injury, had little impact on knowledge or behavior [ 252 ]. Similarly, evidence suggests that boil water advisories may not be adhered to by everyone. During the gastroenteritis outbreak in Walkerton, Ontario (Canada), some residents under the boil water advisory continued to use and occasionally drink the water without boiling it [ 85 ]. The link between water and hygienic practices should be considered when developing drought adaptation measures. Conservation should take into consideration on-farm agricultural practices, regional water management and treatment and allow for and encourage individual hygienic practices such as hand washing [ 46 ]. In the later example, homes and businesses could also promote both water conservation and hand washing by installing low-flow faucet aerators [ 46 ]. Education of health care providers can also benefit efforts to address drought impacts on health [ 53 ]. 3.13.5. Addressing Mental Health Impacts Adaptive actions may be taken by individuals and professionals within primary care and community services settings to reduce the mental health impacts of drought. Personal hopefulness and positive attitudes towards the rural lifestyle have been highlighted as useful adaptations to drought [ 218 , 253 ]. According to Caldwell and Boyd [ 254 ] rural Australian families used a combination of both positive (e.g., problem-focused coping, optimism and positive appraisal) and less beneficial strategies (e.g., cognitive dissonance, denial and avoidance of negative social influences) in their response to drought. Primary care practitioners can provide support to reduce symptoms of mental illness exacerbated by drought and ineffective coping mechanisms such as substance abuse. They may also provide information about financial support resources or community drought support groups [ 207 , 255 ]. Mental health programs in schools and in occupational settings and suicide prevention and awareness programs have also been identified as possible adaptation measures [ 59 , 216 , 222 , 256 , 257 ]. Newspapers may be another pathway to communicate broadly about mental health during drought [ 258 ].In Australia rural support workers help traditional mental health professionals identify farmers experiencing mental distress and refer them to support services [ 259 ]. Training in mental health first aid can also increase the capacity of rural support workers to address mental health needs in drought-affected rural areas [ 260 ]. One study suggested that the Rural Adversity Mental Health Program (RAMHP), which is based on a community development model, is effective in building capacity and resiliency to cope with the impacts of drought in rural New South Wales (Australia) [ 261 ]. Linking the services and programs focused on mental health, community or social needs with those on practical assistance for farming communities is considered key for addressing mental health concerns associated with or exacerbated by drought [ 262 ]. Access to support services for older individuals is another consideration [ 256 ]. Community-building [ 263 ] and events which increase social capital can further help decrease psychological stress [ 254 ]. High levels of social capital in many Canadian (Prairie) rural communities [ 176 , 264 ], based upon high levels of trust, networking and participation in organizations help reduce stress associated with drought [ 176 , 264 ]. However, Australian farm families faced heavier workloads during the drought of 2002–2003, which left less time for community participation, volunteer activities and hence could negatively impact social capital and the community [ 201 ]. 3.14. Adaptation to Address Changes to Environmental Services or Socioeconomic Pathways (Indirect) 3.14.1. Adaptation in Urban Environments Effective adaptation to climate change-related impacts, including drought, requires "mainstreaming" in which climate change information is routinely considered and incorporated into existing programs, policies and planning [ 2 , 62 ]. For example, in Ontario, de Loe and Berg [ 62 ] point to opportunities to "mainstream" climate change into source water protection planning within the Clean Water Act to help address anticipated changes to water supply and demand, including an increased frequency of drought [ 62 ]. Mainstreaming can be assisted by integrating learnings about effective risk management options and adaptations from experiences with current climate variability, particularly extreme weather events. However, mainstreaming also requires consideration and use of information about risks expected from future climate change in adaptation planning. A range of measures to adapt to changes in water quality and quantity in the context of climate change, such as leak detection, outreach, adjustments to the building code and management policies have been reported in the province of Quebec [ 49 ]. Many sectors will be affected by drought and play a role in adaptation to reduce risks to individuals and communities. In urban centers communities may adapt to drought by enhancing water delivery systems and through watershed management [ 59 , 74 , 204 ]. In Atlantic Canada, communities facing shortages or reduced water quality have made efforts to improve water treatment systems, conserve and protect surface water and develop new water sources through, for example, water capture [ 50 ]. Dams, desalination and water recycling are also being used to cope with the increased variability in water supply, although the cost and impacts of such actions are not well established [ 204 ]. The health sector must actively be engaged in drought management at all levels of government. For example, in Brazil actions to protect public health are embedded into all stages of drought-related disaster risk reduction [ 241 ]. Increasing water use efficiency is advocated as an important adaptation when water infrastructure already exists [ 204 ]. Water conservation has also been highlighted as an important measure for easing drought-related costs [ 50 , 51 , 204 ]. This may be done through voluntary or mandatory restrictions on water usage during vulnerable periods [ 43 ] or education campaigns to promote conservation [ 50 , 51 ]. During the 1987–1988 drought in California (U.S.) education was used in conjunction with mandatory and voluntary water conservation programs to target a 10%–25% reduction in water use [ 265 ]. As this drought stretched on, impacts spread from dryland agricultural areas to impacts on state level electrical power supply. Lessons learned in a study of this drought identified inadequacies in water resource planning, the existence of responsible agencies along political rather than affected area boundaries and a lack of inter-jurisdictional communication as key factors. Water shortages continue to occur in California despite national freshwater draws (per capita) having dropped below levels seen in 1975 [ 265 ], with the exception of municipal and industrial use which are also expected to decline as regulations and technologies for more efficient use are adopted. In the most recent 2014 drought in California the entire state was under moderate to exceptional drought and water customers in Santa Cruz were required to decrease their water consumption by 25% with financial penalties for non-compliance [ 266 ]. 3.14.2. Adaptation in the Agriculture Sector Farmers have a great deal of experience in coping with changes in conditions, including drought and are seen to be one population group with a high capacity to adapt [ 174 , 203 ]. In agricultural systems adaptation to drought occurs through the use of modified or new agricultural practices, such as increasing reliance on crop irrigation and the introduction of drought-resistant crops, as well as diversifying crops and income sources [ 61 , 244 , 254 , 267 , 268 , 269 , 270 , 271 , 272 ]. With the added challenge of climate change, seasonal climate outlooks are another tool that farmers in the U.S. are turning to in order to better determine which crops will be successful in a particular season [ 273 ]. The international literature also highlights the importance of drought forecasts and preparing for drought in advance [ 243 , 267 ]. The Southern African Development Communities Drought Monitoring Centre is an example of such an alerting body [ 274 ]. In Canada, agricultural adaptation over the past century has included the development of new agricultural practices and infrastructure to conserve and enhance water supplies and distribution and targeted institutional activities to improve water management [ 174 , 203 , 275 ]. For example, less frequent tilling and reductions of other field management practices that lowered soil moisture and increased erosion aided in lowering the vulnerability of farmers to the 2001–2002 drought relative to the drought of the 1930s [ 203 , 275 ]. Governments have also increased regional resilience by producing drought plans and enacting legislation to support coordinated water management. In England, water resource management plans have been incorporated into climate change adaptation plans [ 276 ]. In the U.S., drought plans have been produced by a number of states [ 277 ]. Water management and climate change adaptation plans are being developed by various watershed, municipal and regional authorities across Canada [ 278 , 279 ]. However, climate, human and other factors affecting water sources and reserves are extensive and rarely conform to the jurisdictional boundaries at which watershed management is required [ 280 , 281 ]. Programs to sustain farmers through times of very low production or crop failure due to natural hazards such as drought are expected to be important in the future [ 282 ], but developing proactive adaptation actions is key to increasing the resilience of existing agricultural systems to drought [ 283 , 284 , 285 , 286 ]. In fact, in the Prairies there are some areas in which conditions such as soil erosion and the accumulation of soluble salts in the root zone may warrant reclassification in use [ 287 ]. A media analysis of adaptations undertaken in Canada during the 2001–2002 drought showed that the most frequently mentioned actions included measures related to irrigation, conservation and the management of water resources, insurance and other financial assistance programs (e.g., Canadian Farm Income Program, Net Income Stabilization Account), as well as government based low water responses [ 288 ]. Irrigation in particular is one of the most broadly referenced agricultural adaptations to drought in the literature [ 289 ]. The use of irrigation systems is increasing across Canada and irrigation is becoming more and more common in regions that had previously relied on rain and subsurface water (Saskatchewan has an Irrigation Strategy to grow and extend the use of irrigation across the province as part of their Water Security Plan http://www.agriculture.gov.sk.ca/Irrigation-Strategy-2014 ; see also [ 335 ]). In Prince Edward Island the growing reliance on irrigation in potato farming and requests for new deep well licenses created serious concerns regarding the sustainability of groundwater supplies and whether or not a moratorium instituted over a decade earlier was still required [ 290 , 291 ]. In November, 2014, a decision was made to uphold the moratorium. In more arid regions where water supplies are finite, like the Okanagan region of southern British Columbia, water demands are managed through a combination of monitoring and management among users (e.g., managing water catchments, use of high elevation dams, reservoirs and instituting seasonal use restrictions). In Alberta's Oldman Dam water shortages have even been managed using a system of "sharing the shortage" in which users agree to share the reduced allocation of water due to drought [ 292 ]. In 2001, partial allotments of 60% were successfully used to ensure the delivery of water supplies to all licensees. Water consumption can also be reduced by managing the timing of irrigation so that it targets the sensitive stages of crop development. In the case of corn, for example, yield losses may be reduced significantly by preventing water deficits at critical times in development; notably during the reproductive stages [ 293 , 294 ]. Even when plants are subjected to drought conditions, the timing and duration of the event must also be considered since plant processes can often recover following a period of drought if critical management actions are taken at the correct time (e.g., irrigation during flowering). Water conservation strategies and the timing and use of water (water management and irrigation timing) represent simple solutions that can be managed in response to need, when and where required ( i.e. , the farm scale). In contrast, technological and transgenic approaches will take time to develop [ 295 ]. Moreover, since drought resistance is linked to overall plant performance, cultivars that are tolerant to extreme drought conditions may result in reduced production during less extreme or episodic drought conditions [ 296 , 297 , 298 ]. As such, farmers may need to balance trade-offs between production and the risk of drought in their choice of what to plant next season. Planting seeds representing a diversity of cultivar tolerance may prove beneficial in this case [ 299 ]. The combined use of existing drought tolerant cultivars with physiologically sensitive timing of irrigation can serve to reduce water demands during key stages of development. Furthermore, since drought stressed plants and livestock are also more vulnerable to diseases, competition from invasive species and pests, strategies that increase genetic and crop diversity across fields and the landscape may represent adaptations that will provide overall benefits to both health and production [ 288 , 300 ]. Looking to the future, against a global backdrop of diminishing lands suited to agriculture and multiple cropping due to climate change, Canada is expected to experience an increase in the area of land that is marginally or moderately suitable for agriculture in farther northern regions [ 301 ]. 3.14.3. Adaptation to Build Individual and Community Resilience At the individual level, personal savings may not be enough to support families during droughts, hence loans, savings strategies and insurance may be important adaptation options [ 267 , 270 , 302 ]. In the U.S., many families went into bankruptcy under the combined drought and Great Depression in the 1930s and had to draw on government aid for support [ 202 ]. Others moved to different communities to earn a living [ 202 ]. In 2012, following a particularly dry summer in Ontario, over 5000 crop damage reports were submitted by farmers [ 303 ]. Although drought-related crop losses in the Prairies from 2002 were offset by farm insurance [ 174 ], not all farmers are registered for insurance [ 303 , 304 ]. Moreover, while smaller farms tend to have more diversified crops which serve in part to mitigate risk ( i.e. , act as adaptation strategies; [ 305 , 306 ]), they may also be less likely to have insurance coverage. Another key adaptation to variability has been to increase income from off-farm sources [ 174 ] including from oil and gas leases [ 203 ]. The long-term viability of such tactics and their impact on social cohesion in communities is unclear. Socioeconomic challenges can impact the adaptive capacity of communities [ 307 ], but so can more external influences such as global scale changes in agricultural production, demand and markets [ 308 ]. As such, the presence of well-developed social institutions and access to highly qualified personnel and information are particularly important for remote rural communities [ 309 ]. In Canada, drought resiliency in some rural and agricultural communities is related to proximity to urban centers. A study by the International Institute for Sustainable Development and Agriculture and Agri-Food Canada [ 309 ] in a 550,000 square kilometer prairie region (comprising 53 Census Divisions across Alberta, Saskatchewan and Manitoba) found that the areas exhibiting the highest adaptive capacity index values were clustered within corridors between major urban centers. Proximity to urban centers is thought to bring benefits related to off-farm incomes, access to technology (including computers/computing technologies), information networks (email, Internet) and access to agricultural institutions. Conversely, the lowest adaptive capacity was observed in areas more remote from urban centers, which were situated in more northern zones of this prairie belt. The importance of social networks and connectivity may be more obvious during periods of crisis, such as during the 2002 drought on the Prairies when farmers in eastern Canada who were less affected by drought shipped 64,000 bales of hay to western Canada through "Hay West" to support the worst struck areas [ 175 ]. In 2012, some western Canadian farmers attempted to reciprocate with hay to eastern farmers [ 175 ]. Adaptations to address the impact of drought on mental health and social capital within the community are also discussed above. 3.14.1. Adaptation in Urban Environments Effective adaptation to climate change-related impacts, including drought, requires "mainstreaming" in which climate change information is routinely considered and incorporated into existing programs, policies and planning [ 2 , 62 ]. For example, in Ontario, de Loe and Berg [ 62 ] point to opportunities to "mainstream" climate change into source water protection planning within the Clean Water Act to help address anticipated changes to water supply and demand, including an increased frequency of drought [ 62 ]. Mainstreaming can be assisted by integrating learnings about effective risk management options and adaptations from experiences with current climate variability, particularly extreme weather events. However, mainstreaming also requires consideration and use of information about risks expected from future climate change in adaptation planning. A range of measures to adapt to changes in water quality and quantity in the context of climate change, such as leak detection, outreach, adjustments to the building code and management policies have been reported in the province of Quebec [ 49 ]. Many sectors will be affected by drought and play a role in adaptation to reduce risks to individuals and communities. In urban centers communities may adapt to drought by enhancing water delivery systems and through watershed management [ 59 , 74 , 204 ]. In Atlantic Canada, communities facing shortages or reduced water quality have made efforts to improve water treatment systems, conserve and protect surface water and develop new water sources through, for example, water capture [ 50 ]. Dams, desalination and water recycling are also being used to cope with the increased variability in water supply, although the cost and impacts of such actions are not well established [ 204 ]. The health sector must actively be engaged in drought management at all levels of government. For example, in Brazil actions to protect public health are embedded into all stages of drought-related disaster risk reduction [ 241 ]. Increasing water use efficiency is advocated as an important adaptation when water infrastructure already exists [ 204 ]. Water conservation has also been highlighted as an important measure for easing drought-related costs [ 50 , 51 , 204 ]. This may be done through voluntary or mandatory restrictions on water usage during vulnerable periods [ 43 ] or education campaigns to promote conservation [ 50 , 51 ]. During the 1987–1988 drought in California (U.S.) education was used in conjunction with mandatory and voluntary water conservation programs to target a 10%–25% reduction in water use [ 265 ]. As this drought stretched on, impacts spread from dryland agricultural areas to impacts on state level electrical power supply. Lessons learned in a study of this drought identified inadequacies in water resource planning, the existence of responsible agencies along political rather than affected area boundaries and a lack of inter-jurisdictional communication as key factors. Water shortages continue to occur in California despite national freshwater draws (per capita) having dropped below levels seen in 1975 [ 265 ], with the exception of municipal and industrial use which are also expected to decline as regulations and technologies for more efficient use are adopted. In the most recent 2014 drought in California the entire state was under moderate to exceptional drought and water customers in Santa Cruz were required to decrease their water consumption by 25% with financial penalties for non-compliance [ 266 ]. 3.14.2. Adaptation in the Agriculture Sector Farmers have a great deal of experience in coping with changes in conditions, including drought and are seen to be one population group with a high capacity to adapt [ 174 , 203 ]. In agricultural systems adaptation to drought occurs through the use of modified or new agricultural practices, such as increasing reliance on crop irrigation and the introduction of drought-resistant crops, as well as diversifying crops and income sources [ 61 , 244 , 254 , 267 , 268 , 269 , 270 , 271 , 272 ]. With the added challenge of climate change, seasonal climate outlooks are another tool that farmers in the U.S. are turning to in order to better determine which crops will be successful in a particular season [ 273 ]. The international literature also highlights the importance of drought forecasts and preparing for drought in advance [ 243 , 267 ]. The Southern African Development Communities Drought Monitoring Centre is an example of such an alerting body [ 274 ]. In Canada, agricultural adaptation over the past century has included the development of new agricultural practices and infrastructure to conserve and enhance water supplies and distribution and targeted institutional activities to improve water management [ 174 , 203 , 275 ]. For example, less frequent tilling and reductions of other field management practices that lowered soil moisture and increased erosion aided in lowering the vulnerability of farmers to the 2001–2002 drought relative to the drought of the 1930s [ 203 , 275 ]. Governments have also increased regional resilience by producing drought plans and enacting legislation to support coordinated water management. In England, water resource management plans have been incorporated into climate change adaptation plans [ 276 ]. In the U.S., drought plans have been produced by a number of states [ 277 ]. Water management and climate change adaptation plans are being developed by various watershed, municipal and regional authorities across Canada [ 278 , 279 ]. However, climate, human and other factors affecting water sources and reserves are extensive and rarely conform to the jurisdictional boundaries at which watershed management is required [ 280 , 281 ]. Programs to sustain farmers through times of very low production or crop failure due to natural hazards such as drought are expected to be important in the future [ 282 ], but developing proactive adaptation actions is key to increasing the resilience of existing agricultural systems to drought [ 283 , 284 , 285 , 286 ]. In fact, in the Prairies there are some areas in which conditions such as soil erosion and the accumulation of soluble salts in the root zone may warrant reclassification in use [ 287 ]. A media analysis of adaptations undertaken in Canada during the 2001–2002 drought showed that the most frequently mentioned actions included measures related to irrigation, conservation and the management of water resources, insurance and other financial assistance programs (e.g., Canadian Farm Income Program, Net Income Stabilization Account), as well as government based low water responses [ 288 ]. Irrigation in particular is one of the most broadly referenced agricultural adaptations to drought in the literature [ 289 ]. The use of irrigation systems is increasing across Canada and irrigation is becoming more and more common in regions that had previously relied on rain and subsurface water (Saskatchewan has an Irrigation Strategy to grow and extend the use of irrigation across the province as part of their Water Security Plan http://www.agriculture.gov.sk.ca/Irrigation-Strategy-2014 ; see also [ 335 ]). In Prince Edward Island the growing reliance on irrigation in potato farming and requests for new deep well licenses created serious concerns regarding the sustainability of groundwater supplies and whether or not a moratorium instituted over a decade earlier was still required [ 290 , 291 ]. In November, 2014, a decision was made to uphold the moratorium. In more arid regions where water supplies are finite, like the Okanagan region of southern British Columbia, water demands are managed through a combination of monitoring and management among users (e.g., managing water catchments, use of high elevation dams, reservoirs and instituting seasonal use restrictions). In Alberta's Oldman Dam water shortages have even been managed using a system of "sharing the shortage" in which users agree to share the reduced allocation of water due to drought [ 292 ]. In 2001, partial allotments of 60% were successfully used to ensure the delivery of water supplies to all licensees. Water consumption can also be reduced by managing the timing of irrigation so that it targets the sensitive stages of crop development. In the case of corn, for example, yield losses may be reduced significantly by preventing water deficits at critical times in development; notably during the reproductive stages [ 293 , 294 ]. Even when plants are subjected to drought conditions, the timing and duration of the event must also be considered since plant processes can often recover following a period of drought if critical management actions are taken at the correct time (e.g., irrigation during flowering). Water conservation strategies and the timing and use of water (water management and irrigation timing) represent simple solutions that can be managed in response to need, when and where required ( i.e. , the farm scale). In contrast, technological and transgenic approaches will take time to develop [ 295 ]. Moreover, since drought resistance is linked to overall plant performance, cultivars that are tolerant to extreme drought conditions may result in reduced production during less extreme or episodic drought conditions [ 296 , 297 , 298 ]. As such, farmers may need to balance trade-offs between production and the risk of drought in their choice of what to plant next season. Planting seeds representing a diversity of cultivar tolerance may prove beneficial in this case [ 299 ]. The combined use of existing drought tolerant cultivars with physiologically sensitive timing of irrigation can serve to reduce water demands during key stages of development. Furthermore, since drought stressed plants and livestock are also more vulnerable to diseases, competition from invasive species and pests, strategies that increase genetic and crop diversity across fields and the landscape may represent adaptations that will provide overall benefits to both health and production [ 288 , 300 ]. Looking to the future, against a global backdrop of diminishing lands suited to agriculture and multiple cropping due to climate change, Canada is expected to experience an increase in the area of land that is marginally or moderately suitable for agriculture in farther northern regions [ 301 ]. 3.14.3. Adaptation to Build Individual and Community Resilience At the individual level, personal savings may not be enough to support families during droughts, hence loans, savings strategies and insurance may be important adaptation options [ 267 , 270 , 302 ]. In the U.S., many families went into bankruptcy under the combined drought and Great Depression in the 1930s and had to draw on government aid for support [ 202 ]. Others moved to different communities to earn a living [ 202 ]. In 2012, following a particularly dry summer in Ontario, over 5000 crop damage reports were submitted by farmers [ 303 ]. Although drought-related crop losses in the Prairies from 2002 were offset by farm insurance [ 174 ], not all farmers are registered for insurance [ 303 , 304 ]. Moreover, while smaller farms tend to have more diversified crops which serve in part to mitigate risk ( i.e. , act as adaptation strategies; [ 305 , 306 ]), they may also be less likely to have insurance coverage. Another key adaptation to variability has been to increase income from off-farm sources [ 174 ] including from oil and gas leases [ 203 ]. The long-term viability of such tactics and their impact on social cohesion in communities is unclear. Socioeconomic challenges can impact the adaptive capacity of communities [ 307 ], but so can more external influences such as global scale changes in agricultural production, demand and markets [ 308 ]. As such, the presence of well-developed social institutions and access to highly qualified personnel and information are particularly important for remote rural communities [ 309 ]. In Canada, drought resiliency in some rural and agricultural communities is related to proximity to urban centers. A study by the International Institute for Sustainable Development and Agriculture and Agri-Food Canada [ 309 ] in a 550,000 square kilometer prairie region (comprising 53 Census Divisions across Alberta, Saskatchewan and Manitoba) found that the areas exhibiting the highest adaptive capacity index values were clustered within corridors between major urban centers. Proximity to urban centers is thought to bring benefits related to off-farm incomes, access to technology (including computers/computing technologies), information networks (email, Internet) and access to agricultural institutions. Conversely, the lowest adaptive capacity was observed in areas more remote from urban centers, which were situated in more northern zones of this prairie belt. The importance of social networks and connectivity may be more obvious during periods of crisis, such as during the 2002 drought on the Prairies when farmers in eastern Canada who were less affected by drought shipped 64,000 bales of hay to western Canada through "Hay West" to support the worst struck areas [ 175 ]. In 2012, some western Canadian farmers attempted to reciprocate with hay to eastern farmers [ 175 ]. Adaptations to address the impact of drought on mental health and social capital within the community are also discussed above. 3.15. Discussion 3.15.1. Knowledge of Current and Future Drought Impacts on Health In recent years information about the potential health impacts of drought has increased. Evidence pointing to health impacts that may be modified by climate change has continued to build since earlier reviews (e.g., by Smoyer-Tomic [ 51 ] in 2004). The number of Canadian-U.S. studies is increasing, with more than half of the records for this region published after 2004. However, Canadian studies continue to be focused on the Prairies. In addition to studies situated within the Canadian context broadly, more than half of the records identified focused specifically on western Canada and all but one targeted the Prairies specifically. Research linking mental health impacts with wildfire and drought has also expanded significantly. However, research around drought and health remains narrow. Approximately three-quarters of the primary international studies identified were focused on mental health and nearly all focus on Australia. In the Canadian-U.S. literature, more than one-third of the primary studies are focused on mosquito-borne disease, largely WNV and the related SLE. This also appears to be the area of current interest, with these records accounting for more than half of the primary Canadian-U.S. studies published after 2004. Limited data on actual health outcomes results in uncertainty about the severity or magnitude of these impacts. Consequently projections of climate change effects on drought-related health impacts remain difficult to make especially in the Canadian context. At the same time that more evidence is needed to understand the impact of drought on health outcomes, public health is challenged by the uncertainty surrounding the probable occurrence of future drought in Canada. There has been limited research directed toward the probability and assessment of future droughts specifically over Canada in the context of climate change and variability. Greater certainty around the characteristics of drought anticipated in the future and their linkages with specific health impacts could help adaptation planners better identify and prepare for key risks in their regions, including those related to human health. For example, the literature reviewed in this paper suggests that certain features of drought (e.g., subsequent heavy rainfall events) may significantly increase risks of some infectious diseases among Canadians. Knowledge of geographic areas likely to be affected by drought, as well as potential changes to frequency, duration and severity of events is critical for the development of adaptation strategies which can directly or indirectly reduce health impacts. Coordinated drought research is limited and fragmented in Canada, often carried out only in response to severe drought experience. Although most regions of Canada have experienced drought, the majority of studies have focused on the Prairies due to the greater frequency of droughts in this region. The variety in the forms of drought experienced in Canada needs to be taken into consideration when planning for possible health impacts. Given the complex interactions with water, a discussion of the impact of drought on soil through the health lens may add to the literature on water and water content alone. 3.15.2. Knowledge of Adaptation Options A recent 2013 systematic review of the effectiveness of public health interventions to reduce the health impacts of climate change highlighted a gap in the literature around adaptation to drought [ 310 ]. While there has been some effort to assess agricultural adaptation to drought [ 311 ] and public health promotion activities that touch on related impacts, such as spinal cord injury [ 250 , 251 ], none of the public health adaptations specific to drought identified in the international literature had been evaluated. Here, again, a lack of specificity around drought among other climate change-related extreme weather events limits the relevance of these discussions for public health planning. Potential adaptations to directly or indirectly reduce health risks from drought are plentiful, particularly those that address agricultural impacts of climate change and weather variability. Few adaptations are discussed specifically in the context of drought but are often put forward to address secondary concerns (e.g., vector-borne diseases). Factors influencing agriculture and agricultural practices are complex and sometimes contradictory, ranging from public level issues driven by policy, legislation and international agreements, to environmental and community issues relating to how land is used and the local impacts [ 287 ]. To a large extent adaptations to drought are emerging from regional risk and adaptation case studies [ 282 ]. There is an opportunity to address multiple climate change-related health impacts under broad public health adaptations. In addition, adaptation activities and programs in other sectors can contribute co-benefits to the agricultural sector. Programs that have been developed for other objectives, such as protecting wildlife or water, have been found to provide incidental benefits to the agricultural sector during drought periods. Such co-benefits are not well recognized, but constitute just short of two-thirds of activities captured in a national inventory of programs for managing drought and agricultural water in Canada [ 311 ]. 3.15.3. Challenges Adapting to Drought Measures taken to adapt to drought between 2000 and 2006 over the Canadian Prairies were expensive, their effectiveness varied and challenges were associated with their implementation [ 199 ]. Gaps in knowledge and misplaced assumptions have been identified as obstacles to drought adaptation. A number of barriers to agricultural adaptation to drought were highlighted during the drought of 2001–2002 in Canada, including the lack of research and a lack of funds, as well as problems relating to institutions and processes [ 288 ]. A study of water management systems in the North American Pacific Northwest found that water management decisions were centred around an assumption of stable conditions and this was compounded by a lack of access to regionally specific climate information [ 312 ]. Challenges have also been noted in simply obtaining high quality local hydrologic information from climate change datasets [ 184 , 185 , 313 ]. The paucity of such information makes it difficult for communities to develop more integrated climate adaptation plans which reduce drought risks. Severe drought can exceed the adaptive capacity of agricultural producers and communities. In their analysis of the 2001–2002 Canadian drought, Wheaton et al. [ 288 ] found that some adaptations were ineffective and that drought brought about negative impacts even when adaptation measures were put in place. Drought can be a prolonged event, which also suggests the importance of considering both long-term and short-term adaptations and approaches to mitigate any long-term health impacts. Despite rural areas having greater social capital than other parts of Canada [ 176 , 264 ], there is much reliance on technologies which may not be able to counter the full impacts of climate change on the production of key crops [ 178 ]. Adaptations derived from new technologies will take time to understand, develop and institute (e.g., transgenics or traditional breeding for drought tolerance) [ 314 ]. Furthermore, the resources required to generate necessary information about future climate and weather conditions and the required analyses may not be available to provide timely planning direction to farmers, municipalities and other local planners [ 286 , 315 ]. The literature reviewed in this paper highlighted that remote areas and communities may have significantly less adaptive capacity to reduce drought impacts than communities closer to urban centres. 3.15.4. Key Knowledge Gaps Key knowledge gaps that need to be addressed to inform drought adaptation include: Projections of drought for Canada (beyond western Canada) including the features of those droughts and regional susceptibility to increased frequency and/or intensity of major droughts; The impacts of drought on infectious diseases which represent a substantial burden of illness in Canada already (e.g., C. perfringens , Campylobacteriosis ), as well as the influence of climate change on these impacts; The impacts of drought on disease rates associated with fungi that are now established in Canada (e.g., Cryptococcus gattii and Blastomyces dermatidis ), as well as the influence of climate change on these impacts; The impact of drought on particle levels in already stressed airsheds and implications for health, as well as the influence of climate change on these impacts; The applicability of mental health impact studies of drought to the Canadian context; Impacts of drought on dust and wildfire-related health impacts; The effectiveness of adaptations to drought, including the impacts of non-health adaptations (e.g., for the agricultural sector) on reducing negative health outcomes. 3.15.1. Knowledge of Current and Future Drought Impacts on Health In recent years information about the potential health impacts of drought has increased. Evidence pointing to health impacts that may be modified by climate change has continued to build since earlier reviews (e.g., by Smoyer-Tomic [ 51 ] in 2004). The number of Canadian-U.S. studies is increasing, with more than half of the records for this region published after 2004. However, Canadian studies continue to be focused on the Prairies. In addition to studies situated within the Canadian context broadly, more than half of the records identified focused specifically on western Canada and all but one targeted the Prairies specifically. Research linking mental health impacts with wildfire and drought has also expanded significantly. However, research around drought and health remains narrow. Approximately three-quarters of the primary international studies identified were focused on mental health and nearly all focus on Australia. In the Canadian-U.S. literature, more than one-third of the primary studies are focused on mosquito-borne disease, largely WNV and the related SLE. This also appears to be the area of current interest, with these records accounting for more than half of the primary Canadian-U.S. studies published after 2004. Limited data on actual health outcomes results in uncertainty about the severity or magnitude of these impacts. Consequently projections of climate change effects on drought-related health impacts remain difficult to make especially in the Canadian context. At the same time that more evidence is needed to understand the impact of drought on health outcomes, public health is challenged by the uncertainty surrounding the probable occurrence of future drought in Canada. There has been limited research directed toward the probability and assessment of future droughts specifically over Canada in the context of climate change and variability. Greater certainty around the characteristics of drought anticipated in the future and their linkages with specific health impacts could help adaptation planners better identify and prepare for key risks in their regions, including those related to human health. For example, the literature reviewed in this paper suggests that certain features of drought (e.g., subsequent heavy rainfall events) may significantly increase risks of some infectious diseases among Canadians. Knowledge of geographic areas likely to be affected by drought, as well as potential changes to frequency, duration and severity of events is critical for the development of adaptation strategies which can directly or indirectly reduce health impacts. Coordinated drought research is limited and fragmented in Canada, often carried out only in response to severe drought experience. Although most regions of Canada have experienced drought, the majority of studies have focused on the Prairies due to the greater frequency of droughts in this region. The variety in the forms of drought experienced in Canada needs to be taken into consideration when planning for possible health impacts. Given the complex interactions with water, a discussion of the impact of drought on soil through the health lens may add to the literature on water and water content alone. 3.15.2. Knowledge of Adaptation Options A recent 2013 systematic review of the effectiveness of public health interventions to reduce the health impacts of climate change highlighted a gap in the literature around adaptation to drought [ 310 ]. While there has been some effort to assess agricultural adaptation to drought [ 311 ] and public health promotion activities that touch on related impacts, such as spinal cord injury [ 250 , 251 ], none of the public health adaptations specific to drought identified in the international literature had been evaluated. Here, again, a lack of specificity around drought among other climate change-related extreme weather events limits the relevance of these discussions for public health planning. Potential adaptations to directly or indirectly reduce health risks from drought are plentiful, particularly those that address agricultural impacts of climate change and weather variability. Few adaptations are discussed specifically in the context of drought but are often put forward to address secondary concerns (e.g., vector-borne diseases). Factors influencing agriculture and agricultural practices are complex and sometimes contradictory, ranging from public level issues driven by policy, legislation and international agreements, to environmental and community issues relating to how land is used and the local impacts [ 287 ]. To a large extent adaptations to drought are emerging from regional risk and adaptation case studies [ 282 ]. There is an opportunity to address multiple climate change-related health impacts under broad public health adaptations. In addition, adaptation activities and programs in other sectors can contribute co-benefits to the agricultural sector. Programs that have been developed for other objectives, such as protecting wildlife or water, have been found to provide incidental benefits to the agricultural sector during drought periods. Such co-benefits are not well recognized, but constitute just short of two-thirds of activities captured in a national inventory of programs for managing drought and agricultural water in Canada [ 311 ]. 3.15.3. Challenges Adapting to Drought Measures taken to adapt to drought between 2000 and 2006 over the Canadian Prairies were expensive, their effectiveness varied and challenges were associated with their implementation [ 199 ]. Gaps in knowledge and misplaced assumptions have been identified as obstacles to drought adaptation. A number of barriers to agricultural adaptation to drought were highlighted during the drought of 2001–2002 in Canada, including the lack of research and a lack of funds, as well as problems relating to institutions and processes [ 288 ]. A study of water management systems in the North American Pacific Northwest found that water management decisions were centred around an assumption of stable conditions and this was compounded by a lack of access to regionally specific climate information [ 312 ]. Challenges have also been noted in simply obtaining high quality local hydrologic information from climate change datasets [ 184 , 185 , 313 ]. The paucity of such information makes it difficult for communities to develop more integrated climate adaptation plans which reduce drought risks. Severe drought can exceed the adaptive capacity of agricultural producers and communities. In their analysis of the 2001–2002 Canadian drought, Wheaton et al. [ 288 ] found that some adaptations were ineffective and that drought brought about negative impacts even when adaptation measures were put in place. Drought can be a prolonged event, which also suggests the importance of considering both long-term and short-term adaptations and approaches to mitigate any long-term health impacts. Despite rural areas having greater social capital than other parts of Canada [ 176 , 264 ], there is much reliance on technologies which may not be able to counter the full impacts of climate change on the production of key crops [ 178 ]. Adaptations derived from new technologies will take time to understand, develop and institute (e.g., transgenics or traditional breeding for drought tolerance) [ 314 ]. Furthermore, the resources required to generate necessary information about future climate and weather conditions and the required analyses may not be available to provide timely planning direction to farmers, municipalities and other local planners [ 286 , 315 ]. The literature reviewed in this paper highlighted that remote areas and communities may have significantly less adaptive capacity to reduce drought impacts than communities closer to urban centres. 3.15.4. Key Knowledge Gaps Key knowledge gaps that need to be addressed to inform drought adaptation include: Projections of drought for Canada (beyond western Canada) including the features of those droughts and regional susceptibility to increased frequency and/or intensity of major droughts; The impacts of drought on infectious diseases which represent a substantial burden of illness in Canada already (e.g., C. perfringens , Campylobacteriosis ), as well as the influence of climate change on these impacts; The impacts of drought on disease rates associated with fungi that are now established in Canada (e.g., Cryptococcus gattii and Blastomyces dermatidis ), as well as the influence of climate change on these impacts; The impact of drought on particle levels in already stressed airsheds and implications for health, as well as the influence of climate change on these impacts; The applicability of mental health impact studies of drought to the Canadian context; Impacts of drought on dust and wildfire-related health impacts; The effectiveness of adaptations to drought, including the impacts of non-health adaptations (e.g., for the agricultural sector) on reducing negative health outcomes. 4. Conclusions Climate change is now considered to be one of the major threats facing public health in the coming decades [ 231 ]. Research on the health impacts of climate change (e.g., extreme heat events, vector-borne diseases) is increasing in Canada and internationally. However, health risks associated with drought and the implications of climate change have been less investigated. To our knowledge, this is the first comprehensive study of the health effects of drought on Canadians and of possible adaptation measures. The direct and indirect health impacts are potentially quite broad and serious given that drought is an important pathway for increased exposure to injuries, air pollution, food, water and vector-borne diseases. Evidence also suggests that a wide range of populations may be at higher risk for these impacts including infants and children, elderly people, the socially and economically disadvantaged, pregnant women, people who spend time outdoors and those with chronic diseases and compromised immune systems. The possible health impacts of drought on Canadians are of concern because drought is a regular occurrence in some communities. Evidence suggests that the incidences of drought increased over most of the country between 1950 and 2002. Several long duration and severe droughts have affected the southern Canadian Prairies and interior valleys of British Columbia over the last century. Available studies mainly suggest that future droughts will likely be more frequent, longer lasting and more severe in those regions that already experience these events ( i.e. , more southern and interior areas of the country). Increased adaptation to drought impacts will be needed at local and regional levels to protect health and to support vibrant and resilient communities. A range of measures exist to address the direct impacts of drought on health or that indirectly modify the ecological or socioeconomic pathways through which health is affected. The nature of individual droughts is expected to influence health impacts. However, there is a paucity of information concerning the effectiveness of many of the health adaptations. Efforts to protect the health of Canadians from future droughts will benefit by increased surveillance and monitoring of health impacts and greater investigation of health risks, vulnerable populations and measures to increase resilience at both the individual and community levels.The results of additional research to address current gaps and support future monitoring and surveillance will be key to inform adaptation and planning for future drought-related health impacts.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7095479/

Not so secure after all

How safe are our microbiology labs? Box 1: Laboratory lapses • 1998 Imperial College London fined for failure to follow health-and-safety rules in a study that created a chimaera of hepatitis C and dengue-fever viruses. • 2001 Anthrax spores sent maliciously through the post in the United States. A laboratory source for the bacterium was suspected. • 2003 Thomas Butler of Texas Tech University charged with infringing regulations on the handling of the plague bacterium, including bringing samples into the United States from Tanzania on a plane without declaring them to customs. • 2003/2004 SARS cases due to laboratory accidents in China, Taiwan and Singapore. • 2007 Texas A&M University work on 'select agents' shut down after failure to report a 2006 incident in which members of staff had been infected with Brucella and Coxiella . D. C. Related links Related links in Nature Research 10.1038/nj7063-1392a 10.1038/news050912-13 10.1038/436894a 10.1038/nbt0805-905 10.1038/437794a Related external links Biological and Toxin Weapons Convention (BTWC) website European Commission consultation paper (pdf) Related links in Nature Research 10.1038/nj7063-1392a 10.1038/news050912-13 10.1038/436894a 10.1038/nbt0805-905 10.1038/437794a Related external links Biological and Toxin Weapons Convention (BTWC) website European Commission consultation paper (pdf)

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3423999/

Recent Advances in Transition Metal-Catalyzed Glycosylation

Having access to mild and operationally simple techniques for attaining carbohydrate targets will be necessary to facilitate advancement in biological, medicinal, and pharmacological research. Even with the abundance of elegant reports for generating glycosidic linkages, stereoselective construction of α- and β-oligosaccharides and glycoconjugates is by no means trivial. In an era where expanded awareness of the impact we are having on the environment drives the state-of-the-art, synthetic chemists are tasked with developing cleaner and more efficient reactions for achieving their transformations. This movement imparts the value that prevention of waste is always superior to its treatment or cleanup. This review will highlight recent advancement in this regard by examining strategies that employ transition metal catalysis in the synthesis of oligosaccharides and glycoconjugates. These methods are mild and effective for constructing glycosidic bonds with reduced levels of waste through utilization of sub-stoichiometric amounts of transition metals to promote the glycosylation.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10418385/

Gating of β-Barrel Protein Pores, Porins, and Channels: An Old Problem with New Facets

β barrels are ubiquitous proteins in the outer membranes of mitochondria, chloroplasts, and Gram-negative bacteria. These transmembrane proteins (TMPs) execute a wide variety of tasks. For example, they can serve as transporters, receptors, membrane-bound enzymes, as well as adhesion, structural, and signaling elements. In addition, multimeric β barrels are common structural scaffolds among many pore-forming toxins. Significant progress has been made in understanding the functional, structural, biochemical, and biophysical features of these robust and versatile proteins. One frequently encountered fundamental trait of all β barrels is their voltage-dependent gating. This process consists of reversible or permanent conformational transitions between a large-conductance, highly permeable open state and a low-conductance, solute-restrictive closed state. Several intrinsic molecular mechanisms and environmental factors modulate this universal property of β barrels. This review article outlines the typical signatures of voltage-dependent gating. Moreover, we discuss recent developments leading to a better qualitative understanding of the closure dynamics of these TMPs. 1. The Structure and Composition of β Barrels Cellular and subcellular membranes include transmembrane proteins (TMPs) that facilitate solute transport and signaling. The first class of TMPs encompasses transmembrane α helices. These hydrophobic proteins are the most abundant among both prokaryotes and eukaryotes. The second class of TMPs includes β barrels, which are folded protein scaffolds made of anti-parallel β strands. A fundamental property of α-helical transmembrane proteins is that they feature continuous hydrophobic stretches of residues across the lipid membrane. In contrast, β barrels are made of polypeptide chains with alternating hydrophobic and hydrophilic residues. This way, a β-barrel structure is like a cylinder with an external hydrophobic interface oriented toward the lipid membrane and an internal hydrophilic surface surrounding an aqueous transmembrane channel. Hence, these TMPs may serve as conduits for transporting nutrients and small-molecule metabolites across membranes. Remarkably, the β strands are connected through a network of numerous hydrogen bonds between the amide and carbonyl groups of the polypeptide backbone. This dense hydrogen bonding distribution is the fundamental molecular mechanism by which a β-barrel scaffold attains its unusually high mechanical and thermodynamic stability [ 1 ]. In addition, β barrels have aromatic side chains at the water–membrane interface, thus forming stabilizing contacts with the polar headgroups and hydrophobic tails of surrounding lipids. Such TMPs are present in the outer membranes (OM) of mitochondria, chloroplasts, and Gram-negative bacteria [ 2 , 3 , 4 , 5 , 6 , 7 , 8 ]. Moreover, β-barrel structures are also formed by various pore-forming toxins (PFTs) [ 9 , 10 , 11 , 12 ]. In Gram-negative bacteria, the β strands are connected by short β turns (e.g., 4–6 residues in length) [ 13 ] on the periplasmic side and long flexible loops on the extracellular side. The size, flexibility, and conformation of various loops vary significantly in OM β-barrel proteins. They may have an important functional role, providing specificity to individual barrels. The loops can be oriented toward the extracellular side, or they can fold back into the pore lumen, drastically reducing the cross-sectional internal diameter of the hydrophilic channel. Hence, they can regulate small-molecule permeability and selectivity. Most OM proteins from Gram-negative bacteria form a β-barrel structure comprising 8–24 β strands [ 6 , 14 ] ( Figure 1 ; Table 1 ). Many β barrels exist as monomers (e.g., OmpA [ 15 ] and FhuA [ 16 , 17 ]). Yet, they can also oligomerize in various ways, to generate either multimeric structures of distinct β barrels (e.g., dimeric PapC [ 18 ] or trimeric OmpF [ 19 , 20 ] and OmpC [ 21 ]) or a single β barrel made of a few polypeptide chains (e.g., the trimeric TolC of Escherichia coli [ 22 , 23 ]). The narrowest monomeric OM β-barrel proteins formed by eight β strands include OmpA [ 15 ], OmpW [ 24 , 25 ], OmpX [ 26 ], and PagP [ 27 ]. OmpA [ 28 , 29 ], an essential virulence factor facilitating eukaryotic cell infection and antibiotic resistance, represents the most abundant OM β-barrel protein in E. coli . The monomeric OM proteins OmpT [ 30 ] and OmpG [ 31 , 32 , 33 ] of E. coli contain 10 and 14 β strands, respectively. The monomeric β-barrel protein ferric hydroxamate uptake component A (FhuA) [ 16 , 17 ] and the lipopolysaccharide (LPS) channel (LptD) [ 34 ] encompass 22 and 26 β strands, respectively. SprA, a protein-conducting translocon of the type 9 secretion system (T9SS), is a 36-stranded OM protein [ 35 ]. This is the widest single-polypeptide β-barrel known to date. The OM β-barrel proteins from Gram-negative bacteria execute various tasks, such as specific porins (e.g., OprD [ 36 ] and OpdK [ 37 ] of Pseudomonas aeruginosa ), passive-diffusion porins (e.g., OmpF [ 19 , 20 ] and OmpC [ 21 ]), enzymatic elements (e.g., the protease OmpT [ 30 ], lipase OMPLA [ 38 ], and acyltransferase PagP [ 27 ]), adhesin (e.g., OmpX [ 26 ]) and structural (e.g., OmpA [ 15 , 28 , 29 ]) proteins, secretion pathways (e.g., PapC [ 18 ]), efflux channels and pumps (e.g., TolC [ 22 , 23 ]), and active transporters (e.g., FhuA [ 16 , 17 ]). Below, Figure 1 , Figure 2 and Figure 3 and Table 1 , Table 2 and Table 3 are presented. Figure 1 β-barrel proteins of Gram-negative bacteria. ( a ) OmpA (PDB:1QJP; [ 15 ]). ( b ) OmpT (PDB:6EHD; [ 39 ]). ( c ) OprD (OccD1) (PDB:3SY7; [ 40 ]). ( d ) OmpG(PDB:2F1C; [ 31 ]). ( e ) OmpF(PDB:2ZFG; [ 41 ]). ( f ) FhuA(PDB:1BY3; [ 16 , 17 ]). ( g ) PapC(PDB:3FIP; [ 18 ]). ( h ) TolC(PDB:7NG9; [ 42 ]). Figure 2 Mitochondrial β-barrel proteins. ( a ) VDAC-1 (porin) from H. sapiens (PDB:6TIQ; [ 43 ]). ( b ) TOM complex from H. sapiens (PDB:7VD2; [ 44 ]). ( c ) FhaC from E. coli (PDB:4QKY; [ 45 , 46 , 47 ]). Figure 3 β-barrel pore-forming toxins. ( a ) α-hemolysin of S. aureus (PDB:4ANZ; [ 48 , 49 , 50 ]). ( b ) Anthrax toxin with lethal factor side and top view (PDB:6PSN; [ 51 ]). ( c ) γ–hemolysin from S. aureus (PDB:3B07; [ 9 ]). ( d ) Aerolysin prepore side and top view from A. hydrophila (PDB: 5JZH/5JZW; [ 12 ]). ( e ) MspA of M. smegmatis (PDB:1UUN; [ 52 ]). ijms-24-12095-t001_Table 1 Table 1 β-barrel proteins of outer membranes of Gram-negative bacteria. Molecular weights were determined using the UniProt profile from each protein database file. The PBD code for each protein is included. These codes were used to determine the oligomeric state. The average internal diameters do not include the side chains of the internal residues. Hence, these diameters are calculated from C α to C α atoms. The number of strands, average diameter, length of barrel, corks, and loops were all determined using the PyMOL Molecular Graphics System (Version 2.4.0; Schrödinger, LLC, New York, NY, USA). Proteins PDB Code Average Molecular Weight(kDa) Oligomeric State Number of β-Strands per Monomer Average Internal Diameter (à ) Length of Barrel (à ) Corks and Loops Citation OmpA 1QJP 37 Mono/Dimer 8 17 28 4 Loops [ 15 ] OmpW 2F1V/2F1T 21 Monomer 8 17 23 4 Loops [ 24 , 53 ] OprD 3SY7 48.4 Monomer 8 35 34 4 loops [ 40 ] OmpT 6EHD 40 Trimeric 10 33 42 8 Loops [ 39 ] OmpG 2F1C 35 Monomer 14 17 48 7 loops [ 31 ] OmpF 2ZFG 40 Trimer 16 33 42 8 Loops [ 41 ] OmpC 2J1N 40.4 Trimer 16 32 35 8 loops [ 21 ] PhoE 1PHO 39.5 Trimer 16 32 39 8 loops [ 54 ] Maltoporin 1AF6 49.9 Trimer 18 37 35 9 loops [ 55 ] FhuA 1BY3 82 Monomer 22 41 31 1 Cork & 11 loops [ 17 ] PapC 3FIP 91.5 Dimer 24 42 36 1 cork & 12 loops [ 18 ] TolC 7NG9 162 Trimer 6 32 128 6 loops [ 42 ] ijms-24-12095-t002_Table 2 Table 2 Mitochondrial β-barrel proteins. Molecular weights were determined using the UniProt profile from each protein database file. The PBD code for each protein is included. These codes were used to determine the oligomeric state. The average internal diameters do not include the side chains of the internal residues. Hence, these diameters are calculated from C α to C α atoms. The number of strands, average diameter, length of barrel, and loops were all determined using the PyMOL Molecular Graphics System (Version 2.4.0; Schrödinger, LLC). Proteins PDB Code Average Molecular Weight(kDa) Oligomeric State Number of β-Strands per Monomer Average Internal Diameter (à ) Length of Barrel (à ) Corks and Loops Citation VDAC 6TIQ 31 Dynamic (Dimer, Trimer, Tetramer) 19 33 37 9 loops [ 43 ] TOM complex 7VD2 38 Dimer 19 35 17 9 loops [ 44 ] Fhac/Sam50 4QKY 54.4 Hexamer 16 32 29 8 loops [ 45 , 46 , 47 ] ijms-24-12095-t003_Table 3 Table 3 β-barrel pore-forming toxins. Molecular weights were determined using the UniProt profile from each protein database file. The PBD code for each protein is included. These codes were used to determine the oligomeric state. The average internal diameters do not include the side chains of the internal residues. Hence, these diameters are calculated from C α to C α atoms. The number of strands, average diameter, and length of barrel were all determined using the PyMOL Molecular Graphics System (Version 2.4.0; Schrödinger, LLC). Toxins PDB Average Molecular Weight (kDa) Number of Chains Internal Diameter (à ) Length of Barrel (à ) Number of β-Strands Citation Cytolysin (Sticholysin II) 1GWY 19.3 1 14 23 10 [ 56 ] α-hemolysin 3ANZ 33 7 26 52 14 [ 48 , 49 , 50 ] γ-hemolysin 3B07 36.7 8 28 47 16 [ 9 ] Bi-component Toxin LukGH 4TW1 36.8 8 30 39 16 [ 57 ] Aerolysin 5JZH/5JZW 54.3 7 28 19 14 [ 12 ] Epsilon toxin 6RB9 36.3 7 30 68 14 [ 58 ] Anthrax Toxin 6PSN 90 7 26 93 14 [ 51 ] Lysenin 5GAQ 33.4 9 34 85 18 [ 59 ] MspA 1UUN 22.1 8 48 37 16 [ 52 ] In the OMs of human mitochondria, the most abundant β barrels are the 16-stranded Sam50 [ 45 , 46 ], the 19-stranded voltage-dependent anion-selective channel isoform 1 (VDAC1) [ 60 ], and the translocase complex of the OM (TOM) [ 44 ] ( Figure 2 ; Table 2 ). It should be noted that both VDAC1 and TOM are structurally distinctive from bacterial OM β barrels because of their odd number of strands. VDAC barrels include three isoforms that result from evolutionary processes and distinct adaptations to a diverse subset of functional roles and interactomes [ 61 , 62 , 63 , 64 , 65 , 66 ]. The mammalian β barrels feature various functionalities in cellular signaling and apoptosis. The OM of chloroplasts also contains various β-barrel proteins such as the outer envelope proteins 37 (OEP37 [ 67 , 68 ]). They function as transporters for small-molecule nutrients, peptides, and nucleic acids. Moreover, the OM includes the translocase of the outer chloroplasts envelope (Toc75) and the outer envelope protein 80 kD (OEP80), which have a focal role in the protein import into plastids [ 69 , 70 ]. Furthermore, many PFTs form β-barrel structures made from several protomers that assemble at the surface of a membrane for oligomerization and pore formation. The archetype of a homomeric β-barrel PFTs (β-PFTs) is the staphylococcal α-hemolysin, a heptameric protein of a known crystal structure [ 48 ]. This complex forms a mushroom-shaped assembly, and each protomer participates with two anti-parallel β strands to form a 14-stranded protein pore ( Figure 3 ; Table 3 ). The protective antigen channel (PA 63 ) of the anthrax toxin secreted by Bacillus anthracis is also a heptameric 14-stranded β barrel that facilitates the translocation of the edema factor and lethal factor proteins into the target cells [ 71 ]. In addition, β-PFTs can be formed by heteromeric complexes, such as bi-component toxins (e.g., leukocidins, γ-hemolysins, and Pantom-Valentine leukocidins (PVL)) [ 72 , 73 , 74 ]. Heteromeric β-PFTs require an interaction between the two distinct protomers [ 9 , 75 , 76 ]. In the past decade, new noncanonical β-barrel structures have been discovered for various transmembrane secretory systems. They include multimeric complexes made of a vast number of β strands mediating diverse protein secretion systems (e.g., CsgG [ 77 ] and secretins [ 78 , 79 ]). Another class of noncanonical β barrels is formed by cholesterol-dependent cytolysins that contain tens of protomers that co-participate in creating giant transmembrane protein pores with sizes between 50 and 200 antiparallel β strands [ 11 ]. 2. Early Observations of Voltage Gating of β Barrels Voltage gating is a biophysical process that implies the transient or permanent closure of a β-barrel protein pore, porin, or channel due to a transmembrane potential. As a result of this closure, the ionic flux is at least partly restricted. Early observations of voltage-dependent gating of various β barrels have been published by several groups, such as those of Rosenbusch [ 80 , 81 ], Lakey [ 82 , 83 ], Engelhardt [ 84 , 85 ], and Delcour [ 86 , 87 , 88 , 89 ]. An extensive amount of this pioneering work has been based on trimeric OM proteins OmpF [ 80 , 81 , 82 , 83 , 88 , 89 , 90 , 91 , 92 , 93 ], OmpC [ 86 , 87 , 94 , 95 ], and PhoE [ 81 , 82 , 88 , 96 ]. Single-molecule electrophysiology studies have reported voltage gating of porins under a broad range of experimental contexts that varied pH [ 97 ], salt concentration [ 98 , 99 ], membrane composition [ 100 , 101 ], method of channel reconstitution [ 89 ], electrostatic potential [ 102 ], and others. Different investigators observed diverse sensitivities to voltage gating in these early explorations. Yet, they agreed that different mechanisms mediate voltage gating. In addition, numerous native and mutated forms of β-barrel porins have been examined using single-channel electrical recordings [ 86 , 87 , 94 ] and full-atomistic molecular dynamics (MD) simulations [ 103 , 104 , 105 , 106 , 107 , 108 , 109 ] for determining biophysical properties such as ion permeation, unitary conductance, ionic selectivity, as well as the kinetics and dynamics of current gating fluctuations [ 110 , 111 , 112 , 113 ]. Several mechanisms of voltage-dependent gating were proposed, including the presence of charged residues within the constricted region of the pore interior and the motions of the long extracellular loops folding back into the pore lumen (e.g., L3 in OmpF). In addition, it was suggested that voltage gating occurs due to other intrinsic processes that correlate with the mechanical stability of β barrels. Later, it was identified that electrostatic effects [ 99 , 114 , 115 ] and pH [ 116 , 117 ] also play critical roles in the dynamics of current fluctuations and the overall stability of β barrels. In the following sections of this review article, we discuss the intrinsic spontaneous gating fluctuations of these TMPs and their regulatory mechanisms by specific environmental conditions. 3. Gating Activity Produced by Loops and Plugs Many β-barrel OM proteins exhibit steric restrictions for ionic flow, such as long extracellular loops that fold back into the pore lumen and internal plug domains. In addition, their conformational moieties may cause reversible current gating in these TMPs [ 118 , 119 ]. For example, in Pseudomonas aeruginosa , substrate-specific 18-stranded porins have large external loops L3, L4, and L7 that partition into the pore lumen [ 120 ], producing a very narrow pore eyelet. Hence, these β barrels exhibit a relatively low-conductance open-substate current decorated by frequent fluctuations [ 37 , 40 , 120 , 121 , 122 ]. Interestingly, deletions of the internal loops L3 in OprD (OccD1) [ 36 ] and L7 in OpdK (OccK1) [ 123 ] increased in their single-channel conductance. These findings suggest that these loops directly participate in the constricted regions of these porins. Moreover, significant progress has been made in understanding the gating activities of monomeric OM proteins because of their potential in biotechnology. The primary benefit of these barrels is the opportunity to redesign them as single-polypeptide chain protein nanopores for single-molecule stochastic sensing. Notably, such a strategy would circumvent the necessity for separating the desired protein pore from other products of the assembly reaction; otherwise, a tedious sample preparation is required for multimeric protein pores [ 4 , 124 , 125 , 126 , 127 ]. Below, Figure 4 and Figure 5 are presented. In the first example, OmpG [ 31 , 32 , 33 , 130 ], a monomeric 14-stranded β barrel that comprises seven extracellular loops, undergoes frequent current fluctuations around the open substate at neutral pH and an applied potential lower than 100 mV [ 112 , 113 ]. Protein engineering and MD simulations were utilized to produce a mutated OmpG with a 95% reduction in the gating activity [ 131 , 132 ]. This approach was achieved by the decrease in the moieties of the gating loop L6 [ 133 ] via (i) an exogenous disulfide bond engineered between strands β12 and β13, and (ii) the optimization of the β11-β12 inter-strand hydrogen bonding via an aspartic acid deletion. Thus, OmpG was the first monomeric β barrel engineered for acquiring a quiet open-substate for biosensing applications. These pioneering studies on the engineering of OmpG for biotechnological applications stimulated further developments for revealing mechanistic information about its gating activity. In a follow-up study, Zhuang and colleagues (2013) [ 128 ] used an innovative strategy for pinning individual extracellular loops of OmpG into the lipid bilayer ( Figure 4 a). This process has been conducted using the chemical modification of individual loops by long-hydrocarbon chain alkylation. This way, different OmpGs with one of the loops immobilized onto the lipid bilayer were systematically studied using NMR in detergent micelles and single-channel electrical recordings. Pinning loop L6 resulted in an OmpG protein pore with a highly reduced gating activity ( Figure 4 b). This discovery was in accordance with previous investigations that indicated the pivotal role of loop L6 in the voltage-gating function of OmpG [ 116 , 117 , 133 ]. Furthermore, this study provided key information about the structural and dynamic alterations of the neighboring and distant loops when one is immobilized onto the lipid bilayer. It also identified that in addition to loop L6, other extracellular loops contribute to channel closing and in different extents of cooperativity with L6. In an independently conducted study by Grosse and coworkers [ 134 ], two L6 deletion variants of OmpG showed a unitary conductance like the wild-type OmpG but significant reductions in the gating activity. Intriguingly, a large truncation mutant of OmpG that encompassed deletions of all loops still exhibited a fivefold decrease in the gating activity with respect to the native protein. Therefore, a residual voltage gating of OmpG was independent of loop L6 conformational changes within the pore lumen. This gating activity may be determined by global changes in barrel conformation, resulting in the transient reduction in transmembrane ionic flux. These reversible structural fluctuations of the porin might involve its barrel stretching and compression [ 135 ]. For example, extensive breathing motions in VDAC1 (VDAC1), a 19-stranded β barrel, were determined using MD simulations and solid-state NMR spectroscopy (see below) [ 136 ]. Recently, Sanganna Gari and colleagues (2021) used a high-resolution AFM-based spectroscopy approach to provide time-resolved conformational fluctuations of loop L6 within the pore lumen of OmpG [ 129 ]. This method, also called high-speed AFM high spectroscopy (HS-AFM-HS), was utilized to find correlations of the physical conformational dynamics of the sample in the vertical direction with 10-microsecond temporal resolution and at angstrom precision. Hence, they found correlations between the physical conformational dynamics of loop L6 probed by HS-AFM-HS and the kinetic details of voltage-dependent gating determined by single-channel electrical recordings on planar lipid bilayers ( Figure 5 ). These explorations were supplemented by MD simulations, which provided additional atomic details of the coexistence of the open and closed states made by fluctuations of loop L6. These studies aimed at a better understanding of the gating activity of OmpG, stimulated by the prospects of using this monomeric β barrel in biosensing applications [ 128 , 131 , 137 , 138 , 139 , 140 , 141 , 142 , 143 ]. In the second example, Ferric hydroxamate uptake component A (FhuA), a monomeric 22-stranded β-barrel of E. coli , was extensively engineered for a better understanding of the gating activity produced by its large extracellular loops and the N-terminal 160-residue plug domain [ 144 , 145 , 146 , 147 , 148 ]. These studies, which involved its functional reconstitution into lipid bilayers, revealed the complexity of different contributions of the cork and loops to the gating activity of this TMP. The primary function of FhuA is to mediate the active high-affinity Fe 3+ uptake into the cell [ 149 ]. A minimal 455-residue FhuA variant, which featured complete deletions of the plug domain and the large extracellular loops (L3, L4, L5, L10, L11), showed a quiet open-state conductance of ~1.6 nS in 300 KCl [ 150 ]. This relatively large single-channel conductance results from the passage of ions across an elliptical internal pore with sides of 2.6 × 3.6 nm. This FhuA variant was frequently employed for further developments in biosensing applications because of its monomeric nature, high thermodynamic stability, and relatively larger size [ 150 , 151 , 152 , 153 , 154 , 155 , 156 ]. Another example of a well-studied plug-containing OM protein is that of the usher pyelonephritis-associated pili C (PapC) [ 18 , 157 , 158 ]. This β barrel is a key element utilized by Gram-negative pathogenic bacteria (e.g., uropathogenic E. coli ) to produce and assemble extracellular pilous fibers. PapC is a large 24-stranded dimeric, twin β-barrel complex with each monomer containing five functional domains: a β-barrel translocation domain, a β-sandwich plug domain, an N-terminal periplasmic domain, and two C-terminal periplasmic domains [ 18 ]. The wild-type PapC is mainly closed due to the β-sandwich plug domain [ 159 ]. However, this closed state is accompanied by short-lived openings to various substates. Further, the opening probability of PapC was increased by subsequent deletions of the N- and C-terminal domains, suggesting that they participate in the functional gating activity. Yet, the deletion of the plug domain resulted in extremely large single-channel conductance openings of the pore of ~3 and 7.3 nS for the monomer and dimer in 1 M KCl, respectively. Frequent closures decorated these open substates. This finding is in accordance with the measured internal size of 4.5 × 2.5 nm for the plug-deleted PapC monomer. Later, antibiotic sensitivity and electrophysiology measurements were employed to determine that a single salt was required to stabilize the 76-residue plug domain within the pore lumen [ 160 ]. In addition, it was identified that the loop between strands β12 and β13 mediates the pore opening. Below, Figure 6 and Figure 7 are presented. 4. Gating Activity Modulated by the N-Terminal Tail Among mitochondrial OM proteins, the VDAC protein has drawn significant interest because of its critical regulatory implications in the metabolic operation of mitochondria under physiological and pathological conditions. The primary role of this multitasking β-barrel protein is to facilitate the exchange of ions, nucleotides, and metabolites between the mitochondrion and the cytosol [ 43 , 61 , 162 , 163 , 164 , 165 , 166 , 167 ]. Yet, its functional characteristics extend to that of a receptor for small molecules and proteins that regulate the overall OM permeability of mitochondria [ 165 , 167 , 168 , 169 , 170 , 171 ]. Human VDAC isoform 1 (hVDAC1) is the most abundant protein in mitochondrial membranes. The fundamental and translational implications of hVDAC1 in cell physiology and disease development have ignited numerous structural [ 63 , 135 , 172 , 173 , 174 , 175 , 176 ], biophysical [ 64 , 65 , 177 , 178 , 179 , 180 , 181 , 182 , 183 , 184 ], and functional [ 61 , 162 , 163 , 164 , 165 , 167 , 170 , 185 , 186 , 187 , 188 , 189 , 190 , 191 ] studies. Because of its pivotal role in interactions with apoptotic and anti-apoptotic proteins [ 165 , 168 , 169 , 190 ], hVDAC1 can potentially serve as a therapeutic target in diverse diseases, including several cancers, as well as cardiac and neurodegenerative pathologies [ 66 , 182 , 192 , 193 , 194 , 195 , 196 ]. The three-dimensional structure of VDAC1 has illuminated a 19-stranded β barrel with a partly α-helical N-terminal segment protruding into the pore lumen [ 60 , 173 , 197 , 198 , 199 ]. It is also worth mentioning that the first and last strands orient parallelly, making VDAC1 part of a unique subclass of β-barrel proteins. The channel exhibits a 4.1 nS conductance at transmembrane potentials lower than 30 mV and in 1 M KCl [ 200 ]. At elevated voltages greater than 30 mV, VDAC1 switches into a low-conductance closed state of ~2 nS. This process is symmetrical with respect to the polarity of the applied transmembrane potential. The transition from the large-conductance open state to the low-conductance close state also involves a drastic change in the ionic permeability from an anion- to a cation-selective pore [ 61 , 166 , 201 ]. Hence, it was postulated that extensive conformational changes of the N-terminal α-helix domain of the channel are responsible for its voltage-dependent alterations in the unitary conductance and ionic permeability. This interesting hypothesis stimulated further explorations of voltage sensing of this mitochondrial channel. To address this fundamental gap, Tejido and coworkers (2012) [ 202 ] engineered a double-cysteine mutant in murine VDAC1 (mVDAC1) to lock the N-terminal helix to the barrel wall. This was accomplished through L10C and A170C mutations on the N-terminal α helix and β strand 11, respectively, fixing the α helix to the pore wall. Surprisingly, the functional reconstitution of this mVDAC1 mutant, which encompasses a pore-lining N-terminal helix, did not reveal significant changes in the voltage-gating activity. This outcome suggested that the N-terminal helix remains linked to the pore wall while transitioning from the open to the closed state. In an independently conducted study, Zachariae and colleagues (2012) [ 136 ] utilized electrophysiology, MD simulations, and solid-state NMR spectroscopy to reveal that the absence of the N-terminal helix enhances the breathing conformational fluctuations of the barrel wall. Therefore, deleting the N-terminal helix of hVDAC1 catalyzes the transition of its open-state conformation to a partly collapsed, closed-state conformation. The rigid N-terminal helix, which is deeply located within the pore lumen, stabilizes the channel in the high-conductance state. In addition, they found that a transient dissociation of the N-terminal helix from the barrel wall is a mechanism for switching the channel into a partially collapsed state, explaining both the unitary conductance and ionic selectivity of the closed state ( Figure 6 ). Voltage-dependent gating activity of VDAC1 is also modulated by other environmental or physical factors, such as lateral membrane pressure [ 100 , 189 ], pH [ 203 ], and temperature [ 204 ]. Recently, substantial progress has been made in a better mechanistic understanding of the most sensitive site involved in the voltage-dependent gating activity of VDAC1. Noskov and colleagues (2022) employed full atomistic MD simulations, X-ray crystallography, and electrophysiology to determine that the K12 residue in the N-terminal helix is a focal point of the voltage gating of mVDAC1 [ 161 ]. This study revealed coordinated motions of internal charged residues with conformational alterations in the cross-sectional β-barrel geometry. K12 fluctuates between two distinct energetic substate minima. Its motions between the two substates amplify the barrel fluctuations, leading to channel gating. Remarkably, the K12E mutant exhibited a structure like that of the wild-type (WT) protein, yet with a restricted motion of E12 residue due to its interactions with adjacent side chains. In accordance with structural data, MD simulations suggested that this single-residue K12E substitution resulted in the stiffening of the channel wall, causing low-amplitude conformational fluctuations of the barrel. These fluctuations diminished barrel motions of the K12E mutant and prevented channel gating ( Figure 7 a). Further, multichannel electrophysiology experiments showed that the single-site mutations of K12 to alanine, serine, or glutamine significantly declined the gating activity of mVDAC1 ( Figure 7 b). 5. Modulation of the Voltage-Dependent Gating by Environmental Conditions 5.1. Effect of pH The impact of acidification on the gating activity has been observed by various groups using different β-barrel proteins, such as staphylococcal α-hemolysin [ 205 , 206 , 207 , 208 ], OmpC [ 87 ], OmpF [ 97 , 209 , 210 ], OmpG [ 116 , 117 , 129 , 133 ], and VDAC [ 203 , 210 , 211 ]. For example, the single-channel electrical signatures can span a wide range of gating fluctuations at an acidic pH, from an enhancement in the current noise [ 205 , 206 ] to large-amplitude current closures [ 207 , 208 ]. The acidification at pH values lower than the p K a (e.g., ~4.0, which is comparable to those of titrable Asp and Glu residues) potentially destabilizes stiffer barrel regions due to the disruption of salt bridges. If titrable salt bridges are in the voltage sensing domain of the pore, then their perturbation likely accelerates the switching of the pore from the open to the closed state. A systematical analysis of the pH dependence of the gating of mVDAC1 revealed an asymmetric effect of the acidification with a prominent effect on the cytosolic side and a modest impact on the mitochondrial intermembrane side [ 203 ]. The numerous stable salt bridges at the cytosolic side of mVDAC1 caused this asymmetric effect. Moreover, the acidification enhanced the single-channel conductance and anion selectivity because of the titrable negatively charged residues at very low pH values. Prior studies of the pH dependence of the voltage-dependent gating of OmpG [ 116 , 117 , 129 , 133 ] revealed loop L6 partitions into the pore lumen under acidic conditions, thus gating this monomeric porin. Min Chen and colleagues (2018) [ 212 ] determined that two charged patches within the pore lumen of OmpG are responsible for the increased gating activity of this OM protein at acidic pH values. Using a computational approach and single-molecule electrophysiology, they discovered that electrostatic interactions formed between loop L6 and charged residues on the barrel wall could be attractive or repulsive at a specified pH value. This way, a new strategy was developed for shifting the gating equilibrium by balanced protonation and deprotonation of essential histidine, aspartate, and glutamate side chains on the pore wall and loop L6. Recently, this approach inspired further engineering of OmpG for improved sensing capabilities at acidic pH [ 213 ]. Therefore, substituting charged residues on loop L6 with neutral side chains generated a relatively stable OmpG variant under a wider pH range. These redesign efforts show promise for developing and validating novel engineered OmpG nanopores for medical biotechnology. 5.2. Effect of Temperature A better understanding of the nature of conformational transitions of β barrels can be acquired by temperature dependence experiments. These studies illuminate thermostability features [ 214 ] of OM proteins when functionally reconstituted into lipid membranes, vesicles, or nanodiscs [ 215 , 216 ]. It should be mentioned that significant temperature alterations substantially affect various factors such as the ionic and sub-molecular diffusional mobilities, solvation layer, and unitary conductance. MD simulations and single-molecule electrophysiology provide information on the mechanisms of temperature-dependent changes in the unitary conductance of β-barrel proteins [ 217 , 218 , 219 ]. The difference between the unitary conductance of the barrel and its value corresponding to temperature-dependent solution conductivity is primarily accounted for by the significant interactions between translocating ions and surface charges on the pore wall. Electrophysiological measurements under different temperature conditions may potentially identify new open and/or closed substates where the barrel resides with the highest probability. For example, OmpA of E. coli undergoes interconvertible open states between small-conductance channels (e.g., 36–140 pS in 1 M KCl), between 15 and 37 °C, and large-conductance channels (e.g., 115–373 pS), between 15 and 37 °C [ 220 ]. At elevated temperatures, the ratio of the numbers of small- and large-conductance channels was altered, illustrating the dynamic coexistence of differently refolded barrels. A β barrel can also undergo significant changes in the equilibrium dynamics between its different substates. This situation occurs when the channel exhibits multiple substates resulting from complex interactions of different regions of the barrel. For example, OpdK of P. aeruginosa [ 37 ] shows three open substates, O 1 , O 2 , and O 3 [ 123 ]. At room temperature, the most probable substate is O 2 ( Figure 8 a). However, O 3 is the most probable substate at 4 °C. Temperature changes can reveal modifications in the activation free energy barriers required to transition from one substate to another ( Figure 8 b) [ 221 ]. The primary benefit of single-molecule electrophysiology is the ability to determine the average time constants corresponding to individual substates [ 222 ]. This way, the precise nature of the gating mechanism is uniquely determined by obtaining the enthalpic and entropic contributions to the kinetic and thermodynamic constants, revealing which process in the gating transition is dominant [ 221 , 223 , 224 , 225 , 226 , 227 ]. For example, these analyses reveal enthalpy- and entropy-driven conformational transitions [ 204 , 227 ]. In addition, they provide quantitative assessments of extensive entropic changes in the gating transitions that are compensated by large enthalpic alterations in the form of enthalpy–entropy compensation [ 221 , 228 , 229 ]. Finally, temperature scanning of β-barrel proteins also has practical importance for identifying their thermostability in applications of biosensor technologies [ 208 , 225 , 226 , 230 ]. Below, Figure 8 is presented. Figure 8 Temperature dependence of conductance substates of OpdK. ( a ) Single-channel electrical traces collected with the native OpdK at various temperatures. ( b ) A free energy landscape model illustrating the kinetic transitions among the O 1 , O 2 , and O 3 open substates. This model shows the activation free energies characterizing various kinetic transitions (Δ G O1→O2 ‡ , Δ G O2→O1 ‡ , Δ G O2→O3 ‡ , and Δ G O3→O2 ‡ ). This figure was adapted from Cheneke and coworkers (2015) [ 221 ]. 5.3. Effect of Lipid Composition and Bilayer Asymmetry As part of the surrounding environment, the membrane's composition may play a regulatory role in the voltage-dependent gating of a β-barrel protein through direct lipid–protein interactions [ 231 , 232 ]. Hence, lipid composition is an additional controlling factor, given its modulatory influence on mitochondrial and bacterial homeostasis. If both membrane monolayers consist of lipids with a relatively short spontaneous curvature, also called lamellar lipids [ 233 ], the elastic pressure within the hydrophobic core is modest. In this case, the lipid environment does not typically impact the gating dynamics of the channel because of its robust barrel scaffold. Yet, the presence of inverted hexagonal phase-forming lipids, also called nonlamellar lipids [ 234 , 235 , 236 ], exerts a substantial lateral packing pressure within the hydrophobic bilayer region, catalyzing ample thickness fluctuations of the membrane. Therefore, nonlamellar lipids significantly impact the gating dynamics of β barrels, despite their apparent mechanical robustness [ 237 ]. For example, phosphatidylethanolamine (PE) and cardiolipin (CL), two dominant lipids in the OMs of mitochondria, amplify the VDAC gating at negative applied potentials due to their high packing pressure [ 100 ]. Queralt-Martin and colleagues (2019) [ 189 ] systematically examined the influential role of the polar headgroup of membrane lipids on the voltage gating of mVDAC1. The charge of the phospholipid headgroup has a substantial effect on the channel gating as well. The positive charge of the headgroup amplifies gating, whereas the negative has a suppressing effect. This outcome reinforces the critical importance of the interfacial electrostatic forces between adjacent lipids and the membrane-solvated side of the mVDAC1 channel. Moreover, the same study clarified that the E73 residue, which faces the hydrophobic side of the channel, is not directly involved in the gating mechanism. Further, the interfacial electrostatic forces at the protein–lipid interface may also have a dominant role. The high local densities of acidic and basic side chains cluster on the hydrophobic side of the channel and in the proximity of the polar headgroup region of lipids, such as in the case of OmpF [ 238 ]. Because of their titrable nature, these local charge densities make the –lipidprotein binding mechanism pH sensitive. Therefore, further explorations are needed for a better understanding of the effect of lipid charges on critical functional aspects of Gram-negative bacteria, such as the antibiotic uptake through porin-facilitated routes. However, there is recent experimental evidence that the hydrocarbon tails of lipids also play an essential task in modulating the voltage-dependent gating of OmpF [ 239 ]. Such an influential role of the hydrophobic core is likely achieved through the reorganization of the OmpF trimer by adopting different local conformations of individual monomers. OMs of Gram-negative bacteria, mitochondria, and chloroplasts have an asymmetric distribution of lipid species in both leaflets. Hence, changes in the asymmetric composition, relative distribution, and physicochemical properties of lipid constituents across these membranes may impact the voltage-dependent gating dynamics of their β-barrel TMPs. Hwang and coworkers (2008) [ 240 ] used a droplet interface bilayer to produce the functional reconstitution of OmpG of E. coli into an asymmetric lipid bilayer with a positively charged monolayer opposing a negatively charged monolayer. Interestingly, they identified different gating signatures of OmpG that depended on the insertion leaflet of the asymmetric bilayer. 5.1. Effect of pH The impact of acidification on the gating activity has been observed by various groups using different β-barrel proteins, such as staphylococcal α-hemolysin [ 205 , 206 , 207 , 208 ], OmpC [ 87 ], OmpF [ 97 , 209 , 210 ], OmpG [ 116 , 117 , 129 , 133 ], and VDAC [ 203 , 210 , 211 ]. For example, the single-channel electrical signatures can span a wide range of gating fluctuations at an acidic pH, from an enhancement in the current noise [ 205 , 206 ] to large-amplitude current closures [ 207 , 208 ]. The acidification at pH values lower than the p K a (e.g., ~4.0, which is comparable to those of titrable Asp and Glu residues) potentially destabilizes stiffer barrel regions due to the disruption of salt bridges. If titrable salt bridges are in the voltage sensing domain of the pore, then their perturbation likely accelerates the switching of the pore from the open to the closed state. A systematical analysis of the pH dependence of the gating of mVDAC1 revealed an asymmetric effect of the acidification with a prominent effect on the cytosolic side and a modest impact on the mitochondrial intermembrane side [ 203 ]. The numerous stable salt bridges at the cytosolic side of mVDAC1 caused this asymmetric effect. Moreover, the acidification enhanced the single-channel conductance and anion selectivity because of the titrable negatively charged residues at very low pH values. Prior studies of the pH dependence of the voltage-dependent gating of OmpG [ 116 , 117 , 129 , 133 ] revealed loop L6 partitions into the pore lumen under acidic conditions, thus gating this monomeric porin. Min Chen and colleagues (2018) [ 212 ] determined that two charged patches within the pore lumen of OmpG are responsible for the increased gating activity of this OM protein at acidic pH values. Using a computational approach and single-molecule electrophysiology, they discovered that electrostatic interactions formed between loop L6 and charged residues on the barrel wall could be attractive or repulsive at a specified pH value. This way, a new strategy was developed for shifting the gating equilibrium by balanced protonation and deprotonation of essential histidine, aspartate, and glutamate side chains on the pore wall and loop L6. Recently, this approach inspired further engineering of OmpG for improved sensing capabilities at acidic pH [ 213 ]. Therefore, substituting charged residues on loop L6 with neutral side chains generated a relatively stable OmpG variant under a wider pH range. These redesign efforts show promise for developing and validating novel engineered OmpG nanopores for medical biotechnology. 5.2. Effect of Temperature A better understanding of the nature of conformational transitions of β barrels can be acquired by temperature dependence experiments. These studies illuminate thermostability features [ 214 ] of OM proteins when functionally reconstituted into lipid membranes, vesicles, or nanodiscs [ 215 , 216 ]. It should be mentioned that significant temperature alterations substantially affect various factors such as the ionic and sub-molecular diffusional mobilities, solvation layer, and unitary conductance. MD simulations and single-molecule electrophysiology provide information on the mechanisms of temperature-dependent changes in the unitary conductance of β-barrel proteins [ 217 , 218 , 219 ]. The difference between the unitary conductance of the barrel and its value corresponding to temperature-dependent solution conductivity is primarily accounted for by the significant interactions between translocating ions and surface charges on the pore wall. Electrophysiological measurements under different temperature conditions may potentially identify new open and/or closed substates where the barrel resides with the highest probability. For example, OmpA of E. coli undergoes interconvertible open states between small-conductance channels (e.g., 36–140 pS in 1 M KCl), between 15 and 37 °C, and large-conductance channels (e.g., 115–373 pS), between 15 and 37 °C [ 220 ]. At elevated temperatures, the ratio of the numbers of small- and large-conductance channels was altered, illustrating the dynamic coexistence of differently refolded barrels. A β barrel can also undergo significant changes in the equilibrium dynamics between its different substates. This situation occurs when the channel exhibits multiple substates resulting from complex interactions of different regions of the barrel. For example, OpdK of P. aeruginosa [ 37 ] shows three open substates, O 1 , O 2 , and O 3 [ 123 ]. At room temperature, the most probable substate is O 2 ( Figure 8 a). However, O 3 is the most probable substate at 4 °C. Temperature changes can reveal modifications in the activation free energy barriers required to transition from one substate to another ( Figure 8 b) [ 221 ]. The primary benefit of single-molecule electrophysiology is the ability to determine the average time constants corresponding to individual substates [ 222 ]. This way, the precise nature of the gating mechanism is uniquely determined by obtaining the enthalpic and entropic contributions to the kinetic and thermodynamic constants, revealing which process in the gating transition is dominant [ 221 , 223 , 224 , 225 , 226 , 227 ]. For example, these analyses reveal enthalpy- and entropy-driven conformational transitions [ 204 , 227 ]. In addition, they provide quantitative assessments of extensive entropic changes in the gating transitions that are compensated by large enthalpic alterations in the form of enthalpy–entropy compensation [ 221 , 228 , 229 ]. Finally, temperature scanning of β-barrel proteins also has practical importance for identifying their thermostability in applications of biosensor technologies [ 208 , 225 , 226 , 230 ]. Below, Figure 8 is presented. Figure 8 Temperature dependence of conductance substates of OpdK. ( a ) Single-channel electrical traces collected with the native OpdK at various temperatures. ( b ) A free energy landscape model illustrating the kinetic transitions among the O 1 , O 2 , and O 3 open substates. This model shows the activation free energies characterizing various kinetic transitions (Δ G O1→O2 ‡ , Δ G O2→O1 ‡ , Δ G O2→O3 ‡ , and Δ G O3→O2 ‡ ). This figure was adapted from Cheneke and coworkers (2015) [ 221 ]. 5.3. Effect of Lipid Composition and Bilayer Asymmetry As part of the surrounding environment, the membrane's composition may play a regulatory role in the voltage-dependent gating of a β-barrel protein through direct lipid–protein interactions [ 231 , 232 ]. Hence, lipid composition is an additional controlling factor, given its modulatory influence on mitochondrial and bacterial homeostasis. If both membrane monolayers consist of lipids with a relatively short spontaneous curvature, also called lamellar lipids [ 233 ], the elastic pressure within the hydrophobic core is modest. In this case, the lipid environment does not typically impact the gating dynamics of the channel because of its robust barrel scaffold. Yet, the presence of inverted hexagonal phase-forming lipids, also called nonlamellar lipids [ 234 , 235 , 236 ], exerts a substantial lateral packing pressure within the hydrophobic bilayer region, catalyzing ample thickness fluctuations of the membrane. Therefore, nonlamellar lipids significantly impact the gating dynamics of β barrels, despite their apparent mechanical robustness [ 237 ]. For example, phosphatidylethanolamine (PE) and cardiolipin (CL), two dominant lipids in the OMs of mitochondria, amplify the VDAC gating at negative applied potentials due to their high packing pressure [ 100 ]. Queralt-Martin and colleagues (2019) [ 189 ] systematically examined the influential role of the polar headgroup of membrane lipids on the voltage gating of mVDAC1. The charge of the phospholipid headgroup has a substantial effect on the channel gating as well. The positive charge of the headgroup amplifies gating, whereas the negative has a suppressing effect. This outcome reinforces the critical importance of the interfacial electrostatic forces between adjacent lipids and the membrane-solvated side of the mVDAC1 channel. Moreover, the same study clarified that the E73 residue, which faces the hydrophobic side of the channel, is not directly involved in the gating mechanism. Further, the interfacial electrostatic forces at the protein–lipid interface may also have a dominant role. The high local densities of acidic and basic side chains cluster on the hydrophobic side of the channel and in the proximity of the polar headgroup region of lipids, such as in the case of OmpF [ 238 ]. Because of their titrable nature, these local charge densities make the –lipidprotein binding mechanism pH sensitive. Therefore, further explorations are needed for a better understanding of the effect of lipid charges on critical functional aspects of Gram-negative bacteria, such as the antibiotic uptake through porin-facilitated routes. However, there is recent experimental evidence that the hydrocarbon tails of lipids also play an essential task in modulating the voltage-dependent gating of OmpF [ 239 ]. Such an influential role of the hydrophobic core is likely achieved through the reorganization of the OmpF trimer by adopting different local conformations of individual monomers. OMs of Gram-negative bacteria, mitochondria, and chloroplasts have an asymmetric distribution of lipid species in both leaflets. Hence, changes in the asymmetric composition, relative distribution, and physicochemical properties of lipid constituents across these membranes may impact the voltage-dependent gating dynamics of their β-barrel TMPs. Hwang and coworkers (2008) [ 240 ] used a droplet interface bilayer to produce the functional reconstitution of OmpG of E. coli into an asymmetric lipid bilayer with a positively charged monolayer opposing a negatively charged monolayer. Interestingly, they identified different gating signatures of OmpG that depended on the insertion leaflet of the asymmetric bilayer. 6. Applications in Biotechnology This review article is focused on the biophysical mechanisms of voltage gating of β-barrel proteins. As mentioned above, many of these studies have been stimulated by prospects of employing these protein scaffolds in applied areas of biosensing and medical biotechnologies. The structural integrity and high thermodynamic stability of β barrels make them robust and versatile nanostructures, while the current modulation produced by their interaction with other molecules provides a sensitive readout [ 126 ]. These two properties create opportunities to develop powerful single-molecule sensors for various applications in molecular biomedical diagnostics and environmental monitoring [ 127 , 241 , 242 , 243 ]. For example, β barrels are utilized in DNA sequencing [ 244 ]. In addition, they are employed in the detection [ 245 , 246 , 247 ], chemical modification [ 248 ], and sequencing [ 249 , 250 ] of proteins. Specifically, MspA [ 251 ] and α-hemolysin [ 252 ] are examples of β barrels optimized for DNA sequencing, while proteins like FhuA [ 208 ] and OmpG [ 131 ] have been engineered to detect numerous target proteins. As basic research rapidly progresses, more β-barrel proteins are redesigned to address persistent demands and technical shortcomings in nanobiotechnology. 7. Concluding Remarks In this review article, we briefly recapitulate elements concerning the structure and composition of these β-barrel protein pores, porins, and channels. The primary aim is to critically discuss the mechanisms of intrinsic voltage-dependent gating of these TMPs. Further protein engineering of barrel proteins will likely generate novel redesigned scaffolds for medical biotechnology. Moreover, an enormous body of literature concerns voltage gating of VDAC1 due to its regulatory mechanisms in mitochondria under physiological and pathological conditions. Several voltage-gating issues remain unresolved, so more developments and efforts are necessitated for their comprehensive and quantitative understanding. These fundamental gaps will likely be addressed in the future by utilizing high-resolution technologies both in a cell-free environment and in living cells.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1904290/

Lateral flow microarrays: a novel platform for rapid nucleic acid detection based on miniaturized lateral flow chromatography

Widely used nucleic acid assays are poorly suited for field deployment where access to laboratory instrumentation is limited or unavailable. The need for field deployable nucleic acid detection demands inexpensive, facile systems without sacrificing information capacity or sensitivity. Here we describe a novel microarray platform capable of rapid, sensitive nucleic acid detection without specialized instrumentation. The approach is based on a miniaturized lateral flow device that makes use of hybridization-mediated target capture. The miniaturization of lateral flow nucleic acid detection provides multiple advantages over traditional lateral flow devices. Ten-microliter sample volumes reduce reagent consumption and yield analyte detection times, excluding sample preparation and amplification, of 0 amol dnaR89 controls ( P 0 amol dnaR89 controls ( P < 0.05, n = 6). By the same criterion, 1 fmol dnaR89 detection limits were obtained for both R-24-43 and R-36-55 ( P < 0.05, n = 6). Figure 5 B depicts the performance of LFM detection over 0 to 2500 amol dnaR89 range using the R-77-96/R-57-76-3TN capture/detection probes. LFM detection exhibited excellent linearity, R 2 = 0.989, over this 10-fold range of target molecules. While capture probe R-24-43 exhibited less sensitivity than R-77-96, this capture probe displayed excellent signal linearity between 2.5 and 100 fmol dnaR89, R 2 = 0.968 ( Figure 5 C). These findings demonstrate that the LFM capacity to display multiple capture sequences can be used to simultaneously provide sensitive detection and extend assay linearity through the use of capture probes with differing hybridization characteristics. LFM detection time course studies The small sample volumes used for LFM detection and the reduced surface area traversed during capillary lateral flow significantly reduces detection times for the LFM relative to traditional lateral flow devices. To quantitatively present the speed of LFM nucleic acid hybridization-based detection, we used digital video to follow hybridization sandwich assay-mediated detection of synthetic target molecule dnaR89. Individual frames were isolated from video datasets and quantified for relative signal intensity over the course of capillary lateral flow across the LFM substrate. The resulting signal data was plotted versus time in seconds as shown in Figure 6 . For time measurements, t 0 was defined as the time when the sample front reached the first row of LFM features. Signal was detectable for 1000-fmol target in 2 s following sample transport across R-77-96 capture elements. Within 4 s 100 fmol dnaR89 was detectable while 10 fmol was clearly detectable by 30 s as defined by the earliest time point at which 90% of the pixels composing the R-77-96 microarray features were greater than one standard deviation above background. Lateral flow transport of the 10-μl sample was complete by 120 s. Figure 6. Time course of LFM detection: 10-μl samples containing either 1000 fmol (circle), 100 fmol (square) or 10 fmol (diamond) dnaR89 were run on appropriately patterned LFMs. Video data were collected and colorimetric signal intensity measured from video frames at R-77-96 capture features. Capillary transport of the 10-μl sample was complete by 120 s. Lines represent logarithmic curve fits to the data. Allele-specific Isothermal Amplification and LFM-based Product Detection Prior reports have described a single nucleotide polymorphism (SNP) present in B. anthracis but not close phylogenetic near neighbors including B. cereus and B. thuringiensis ( 29 , 30 ). This SNP has been used as the basis for a sensitive and highly discriminatory real-time PCR assay for B. anthracis ( 30 ). To determine the utility of LFM technology for detecting minority nucleic acids in complex samples, NASBA primers were designed to amplify the plcR allele of B. anthracis . P1 and P2 primer sequences, plc-P1 and plc-P2, used for NASBA amplification are provided in Table 1 and their binding positions illustrated in Figure 1 . Varying amounts of total cellular RNA isolated from B. anthracis or 2 ng of B. thuringiensis HD 621 RNA as a negative control were introduced to 1 μg of total human cellular RNA isolated from HeLa S3 cells. The resulting mixtures were subjected to NASBA amplification using plc-P1 and plc-P2 primers. Human RNA was included in NASBA amplification reactions to approximate the nucleic acid complexity expected in human diagnostic specimens. An aliquot of 2 μl of NASBA reaction mixture was removed after a 60-min incubation at 41°C, mixed with 8 μl of LFM running buffer and assayed for plcR amplicon by LFM. Dyed microspheres cross-linked to R-57-76-3TN were used for detection of NASBA amplicons captured on LFMs carrying R-77-96. Data from these studies are presented in Figure 7 . Following 60 min of NASBA amplification, as little as 0.5 pg for total cellular B. anthracis RNA could be detected in a background matrix of 1 μg of human total RNA. These studies closely approximate the conditions expected for complex human diagnostic samples and reveal the capacity of the LFM platform to specifically detect NASBA reaction products generated from mixed samples where the target sequence is a minority species. While the number of plcR mRNA copies in a B. anthracis cell has not been determined, an estimate of LFM assay sensitivity, in terms of B. anthracis cells, can be calculated based on total RNA yields. Total RNA yields from vegetative B. anthracis were in the range of ∼167–250 fg RNA/cell. Using this value, an estimate of LFM sensitivity corresponds to the detection of approximately to 2–3 B. anthracis cells. Figure 7. ( A ) Indicated amounts of total cellular RNA from B. anthracis Sterne strain 7702 or, as a negative control, 2 ng B. thuringiensis strain HD 621 RNA (0 fg panel) were introduced to 1 μg of total human cellular RNA isolated from HeLa S3 cells. RNA mixtures were subjected to NASBA amplification for 60 min after which 2 μl aliquots of the NASBA reactions were mixed with 8 μl of LFM running buffer and introduced to LFMs. Enlarged LFM sub-regions are shown following cropping, grayscale conversion and Auto Contrast adjustment in Photoshop. The legend indicates microarray element identities: (+) dnaR89 as a positive hybridization control, (–) R-57-76-3N as negative hybridization control, ( 24–43 ) capture probe R-24-43, ( 36–55 ) capture probe R-36–55, (77–96) capture probe R-77-96. ( B ) Graph of quantified signals from B. anthracis and B. thuringiensis challenged LFMs with linear regression line ( R 2 = 0.970). 0 fg B. anthracis total cellular RNA data point contains 2 ng B. thuringiensis total cellular RNA in addition to 1 μg human total cellular RNA. Error bars depict measurement SD (three determinations). DISCUSSION Lateral flow detection of DNA or RNA amplification reaction products provides one means of simplifying nucleic acid detection. Indeed, a lateral flow platform may offer many significant advantages for employing nucleic acid assays under conditions where a fully equipped molecular biology laboratory infrastructure is not available or desirable. Such situations would include resource poor settings, point-of-care, battlefield deployments and scenarios where first responders must quickly determine the threat presented by an unknown substance. To date, however, lateral flow devices have predominantly been fabricated using one or a few capture lines thus limiting the information capacity of the device to one or a few analytes ( 14 , 27 ). As a step toward higher information-content lateral flow nucleic acid detection, we have developed nitrocellulose-patterning methods that enable microarray feature density to be attained on lateral flow compatible substrates. Making use of a non-contact peizo actuated picoliter deposition system, we have patterned lateral flow compatible nitrocellulose membranes with features similar in size and spacing to those typically found on spotted glass microarrays. The sensitivity of lateral flow nucleic acid detection methods previously reported in the literature has been in the order of 1 fmol ( 25 ). We find that the LFM platform provides rapid detection of as little as 250 amol of target using a low-cost and widely available flatbed scanner, a standard personal computer system and a commercially available microarray data extraction suit or free image analysis software. This detection limit is similar to the sensitivity reported for fluorescence and chemiluminescence microarray detection strategies ( 9 , 46 ). While the LFM implementation reported here exhibits excellent linearity ( R 2 = 0.989), the linear dynamic range is less than that commonly associated with fluorescence-based detection. Alternative LFM detection schemes and the use of capture probes of differing hybridization characteristics should enable greater dynamic range while retaining the simplicity of the LFM approach. Indeed, examining the signals generated by R-77-96 and by R-24-43, the effective linear range of the LFM assay extends over a 400-fold range of target from 250 to 100 fmol ( Figure 5 B and C). The information density of the LFM offers the capacity for additional capture probes of varying hybridization potential to be included that should allow this dynamic range to be extended further. The uniformity of sample flow exhibited by the LFM suggests that larger capture probe sets can be accommodated without complications arising from physical factors. For example, concentrations of analyte 40-fold above the linear range of R-77-96 did not adversely impact the linearity of R-24-43 signal at LFM elements situated directly downstream (with respect to sample flow) of R-77-96 capture features ( Figure 5 B and C). Only at artificially high microsphere capture densities, such as those produced by the positive control hybridizations in Figure 4 , are signal gradients observed as a function of physical location on the LFM, presumably due to physical occlusion of membrane pores by high local accumulations of microspheres. LFMs offer several advantages arising directly from the miniaturization of the system without sacrificing detection sensitivity. While traditional lateral flow assays make use of sample volumes in the order of hundreds of microliters to milliliters, the miniaturization approach we have developed reduces sample volume to 10 μl. This reduced sample volume significantly decreases the consumption of reagents required for amplification. Here we have made use of 2 μl of a NASBA reaction diluted to 10 µl in running buffer. By reducing standard NASBA reaction volumes from 20 to 2 μl, a one order of magnitude reduction in enzyme consumption is realized. It should also be noted that other amplification schemes, such as those that make use of microfluidic systems or lab-on-a-chip technologies, could be integrated with a miniaturized lateral flow-based detection system to provide a rapid and cost-effective means of detecting analytes. A further benefit of miniaturization is the time required to detect analyte following introduction of amplified material to the LFM. While the procedures used here employed NASBA amplification and traditional RNA isolation protocols requiring ∼90 min to complete, more recent advances in nucleic acid preparation and amplification have reported significant reduction in sample processing times ( 47 ). As amplification protocols become more rapid, the speed with which amplicons can be detected, without complex optical systems and fluorescent detection, becomes critical to realizing the potential of these technologies. The LFM methods described here detect nucleic acid analytes in less than 2 min. Given that 250 amol is equivalent to 1.5 × 10 8 molecules, efficient amplification methods that offer 10 9 -fold amplification, widely cited amplification levels for PCR- and NASBA-based techniques ( 22 , 48 ), would theoretically enable the detection of single-copy targets by LFM following amplification. Future systems that couple advanced amplification technologies and compatible streamlined nucleic acid preparation modalities with rapid LFM detection will allow significant decreases in sample-to-answer times without costly or complex instrumentation.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9462723/

“ What can we do to actually reach all these animals ?” Evaluating approaches to improving working equid welfare

NGOs utilise a range of programming to bring about desired changes. This study examines the advantages and limitations of the range of approaches used by one particular NGO sector: working equid welfare organisations, from the perspective of NGO staff. NGO staff hold a wealth of knowledge regarding the design and implementation of welfare improvement initiatives; however this knowledge is rarely systematically documented. Through 32 semi-structured interviews the experiences of staff across multiple organisations were explored. The nine most commonly utilised approaches within equid focussed NGO programming were discussed with participants. Many themes resulting from the interviews centred around the strengths and disadvantages of these individual approaches, implemented both historically and presently by different NGOs. The influences of the context in which initiatives took place and barriers to welfare improvement that existed across approaches were also examined. Striking a balance came through as a strong theme at many levels. Balancing top-down versus bottom-up approaches was a frequently voiced concern, as was balancing the utility of certain approaches against associated factors warranting caution. Combinations of approaches that were felt to be complementary provided a balance that drew on the strengths and mitigated for the weaknesses of different approaches. The need to tailor approaches to individual contexts was also raised and is especially relevant to INGOs working across multiple countries, cultures and political structures. The study provides an informed insight into potential factors to consider when designing future welfare initiatives. The collaborative pooling of experience across different NGOs could help make welfare initiatives more effective and provide a framework for NGOs in other fields to learn from each other's collective knowledge. Introduction The approaches utilised by non-governmental organisations (NGOS) to bring about change vary based on factors such as the organisation's size, mission and structure [ 1 ]. Organisations can be direct implementers (such as delivering aid in disaster affected areas or implementing humanitarian activities), purely advocacy-based (influencing policy and governmental decision making) or a hybrid of both [ 2 ]. Larger international NGOs likely employ multiple types of approaches whilst smaller specialist NGOs may focus solely on delivering one type of dedicated programming. Despite the variety of different subject areas that NGOs encompass, parallels can be drawn between the approaches used and programming decisions made across sectors. This study focuses on animal welfare NGOs, specifically those working to improve global welfare standards of working equids (donkeys, mules and horses). The diverse range of approaches used by these organisations to try and achieve welfare improvement has raised questions regarding the efficacy of these different models in achieving welfare change [ 3 , 4 ]. However there is little research evidence available to determine whether the effects of welfare initiatives endure in the long-term after programming ends [ 5 ]. Implementing an inappropriate type of initiative can lead to not only a lack of improvement in equid welfare but the loss of NGO resources and time, unmet expectations of target communities and potential disengagement with future welfare efforts [ 6 ]. Taking into consideration the existing debates regarding programming approaches, as well as the lack of published research evaluating the success of different types of welfare improvement initiatives [ 5 – 8 ]; the selection of an appropriate approach to implement can be difficult for practitioners. NGO staff hold a wealth of knowledge regarding the design and implementation of welfare improvement initiatives, including the contextual factors and practical barriers that impact success. However, this knowledge is rarely systematically documented. This study examines the advantages and limitations of the range of approaches used by one particular NGO sector: working equid welfare organisations, from the perspective of NGO staff across a range of positions. To the authors' knowledge, this is the first study in the sector to explore the experiences of staff across multiple organisations, it is hoped that the collective experiences documented can aid all organisations when considering the design of future welfare initiatives. Working equids are vital both economically and socially to millions of people worldwide. Contrary to the belief that with the increase of mechanisation working animals are declining in importance, recent figures show a rise in the number of working donkeys used worldwide, and a particular increase across sub-Saharan Africa [ 9 ]. Equids fulfil diverse roles, they offer a form of both traction and transport and are utilised across multiple types of industry [ 4 ]. They are of particular importance to marginalised sections of society and those in extreme poverty, sometimes representing a household's only form of income [ 10 ]. Yet working equids commonly suffer from a high prevalence of welfare problems [ 11 ] and are largely absent from animal health and agricultural policy [ 10 ]. As a consequence, a range of organisations run programmes designed to improve the welfare of working equids and support the livelihoods of people relying on them. Originally, the provision of free or subsidised veterinary care was the most commonly utilised approach, however it is becoming increasingly acknowledged that the influences determining working equid welfare are complex and interlinked, with studies suggesting that socioeconomic and environmental factors both play important roles [ 11 – 13 ]. As a result, many NGOs are broadening the scope of their programming to target human factors and more actively engage with the people responsible for determining animal welfare. Approaches include introducing knowledge and skills to communities and helping to develop local infrastructure in order to support sustainable welfare improvements [ 14 ]. Increasing owner knowledge about animal welfare and sentience through targeted education programmes is also an avenue utilised for changing human attitudes towards animals [ 15 ]. In working with communities, the implementation of an initiative aiming to improve equid welfare relies entirely on community engagement and participation, therefore the inclusion of local perspectives and the addressing of socio-behavioural factors is crucial to ensuring initiative success [ 16 ]. Many NGOs are now using community focussed participatory exercise approaches to prioritise community views on welfare issues and use these perspectives to guide the direction of future programming [ 17 ]. At national and international levels, organisations are also engaging with the people responsible for creating animal welfare legislation and policy to create top-down changes that promote positive welfare. Methods Semi-structured, online interviews were carried out with employees of animal welfare organisations, experienced in the development and implementation of equid welfare improvement initiatives. Interviews ran from February to July 2021 and a total of 32 individuals participated from 8 animal welfare organisations of varying sizes (INGO to national), participants were based in 13 countries. A representative range of opinions was sought within organisations (from office-based roles in headquarters to field staff working within equid-owning communities) and participants held positions including Directorial roles, Regional Managers and Co-ordinators, Veterinarians, Researchers and Welfare Officers. All individuals were involved with welfare approaches in some capacity, from those responsible for choosing the direction of programming through to those facilitating initiatives at ground level. Researchers initially contacted two known staff members from two international working equid NGOS. These key contacts identified potential participants within their organisation and invitations to participate were sent to these individuals by the research team. The key contacts also identified key contacts in other relevant national and international NGOs who then facilitated participant recruitment within their organisation. Through this form of snowball sampling a volunteer sample of participants was recruited, all people who expressed their willingness to participate were interviewed. Interviews were pre-arranged, written consent was obtained and subsequently interviews took place online using Zoom video conferencing software. For subsequent transcription, all interviews were recorded and the average interview length was 55 minutes (30-108 minute range). Participants received a document detailing the interview themes before interviews were conducted ( S1 File .) and a document listing the 9 most common approaches utilised by NGOs to improve working equid welfare ( S2 File ). The list provided was designed to be as exhaustive as possible and was informed by current NGO programming at the time; the option was also given for participants to discuss any non-listed approaches that they had encountered. Participants were asked about the approaches most often employed by their organisation (including any not listed) and whether they had noticed any changes in programming direction over time. They were also asked about their own past experiences carrying out welfare initiatives, particularly regarding instances which had been notably successful or unsuccessful, potential influencing factors and the context in which the initiative had taken place. A thematic analysis was conducted in order to explore similarities or differences in themes emerging from different approaches, organisations and staff positions. Themes were generated based on discussions of each approach separately. The 9 approaches in the list discussed during interviews were considered and reclassified during thematic analysis according to the manner in which they were most often grouped or referred to by participants. Although related, skills initiatives and education were frequently discussed by participants as separate programmes and with many points raised concerning each, it was decided to split these topics and present them separately. Conversely the topic 'entertainment with a welfare message' was mentioned most frequently in reference to education. As such, content relating to this topic is presented incorporated into the education section. Two non-listed approaches were discussed by participants, non-equid livelihoods-based programmes and disease surveillance programmes, discussion of these approaches has been integrated into the umbrella topic under which they fit. Themes related to common challenges which occurred across approaches were also noted and analysed. The University of Portsmouth's Ethics Committee for the Faculty of Science and Health reviewed and approved the study (reference SHFEC 2020 – 087). Results and discussion Through the process of thematic analysis major themes emerged from the data. Many of these themes centred around the strengths and weaknesses of the individual approaches implemented both historically and presently by different NGOs. The influences of the context in which initiatives took place and barriers to welfare improvement that exist across approaches were also examined. Each of these areas will be explored and prominent subthemes highlighted. Programmatic approaches to improving welfare The nine most commonly utilised approaches within equid focussed NGO programming were explicitly discussed with participants. Approaches were considered in turn and the associated advantages, limitations and potential consequences of each were considered. Advocacy It was apparent that there has been a relatively recent shift placing advocacy as a strong focus across virtually all organisations. This is not unique to the working equid sector; the growth of advocacy across both generalist and specialist UK based NGOs has been documented [ 18 ]. A larger focus on advocacy was perceived as having both benefits and disadvantages. Some participants felt that advocating for policy change was the best way to enhance the lives of large numbers of animals, many more than could be reached through more practical, on the ground initiatives. "I actually think targeting governments , and getting them to implement the [WOAH] chapters and getting them to enforce good welfare is probably your best way of making a huge difference , and having a huge reach . " P14. It was recognised that advocacy can help change systems in a strategic way that aims to address root causes of poor welfare, rather than addressing the symptoms of individual animals. "going for the right level would get better spread , better results , I think , in the end but it's more of a long-term plan…firefighting is never going to get you a result at the end of the day . It's not a strategic way of thinking . " P14. Recent advocacy efforts have seen equid welfare NGOs working together to form groups such as the International Coalition for Working Equids to implement the World Organisation for Animal Health (WOAH) Welfare of Working Equids Terrestrial Code Chapter 7.12 [ 19 ]. Other groups such as the Equid Power Network [ 20 ] have been involved in promoting educational events for the public and professionals across disciplines. It was acknowledged that for certain large-scale international problems (such as the trade in donkey skins [ 21 ]), correspondingly large-scale advocacy was likely to be the only effective solution. However, it was felt by some participants that the current prominence of advocacy on NGO agendas came at the expense of other approaches. "I think , particularly nowadays , every charity organisation is actually advocating advocacy… It seems as if the practical work on the ground has been neglected" P11. It was also expressed that advocacy could be essentially risky in terms of its rewards compared to the effort put in to achieve changes. Dependent upon the political climate, policy changes can be difficult to achieve and the process was described as being both lengthy and often frustrating for NGO staff. "the least successful approach…(is) the engagement with the policymakers because these policymakers they are bureaucrats and the government systems are very slow" P28. A second highlighted issue was that even if an advocacy strategy was successful in creating a policy change, this was no guarantee that the effect of the change would be felt by animals and communities. "So that [law] existed when I first started working with the project and I don't think I ever saw it enforced and I was there for 2 and a half years . So like although it was written down it didn't make any difference on the ground as such . That's just one of those things isn't it ? " P2. A gap due to lack of enforcement of welfare legislation was often discussed as an obstacle to achieving welfare change. "There is some policy on document , this policy has not been implemented , you know , these are really challenges for us . " P25. A number of reasons were suggested for failure to enforce regulations including a lack of staff capacity, lack of resources at the disposal of the enforcing body and a lack of education of enforcers. "policeman for example . Suddenly he is told that if he sees someone beating a donkey or carrying more than 300 kilos weight that man is… breaking the law . So he sees somebody carrying 200 kilos in the cart with 2 flat tyres , a bad harness and you know dogs biting the donkeys and everything else . And he's not broken the rules so he lets him go , that's the first problem . That the people that are enforcing the rules don't understand what they are enforcing…without that knowledge they can't make a judgement . " P4. These factors are often systematic within enforcement entities (most commonly police or government enforcement bodies) which presents another challenge for NGOs who can have little control over how the policy is eventually enforced. This led to discussions that some political infrastructures are less likely to successfully support effective policy changes. "lobbying and influencing future policy regarding equids can have greater and lasting impacts on welfare…mainly when the country is … resourceful enough to enforce it . " P29. Although acknowledging that it was resource intensive, some participants described an ideal balance needed of top down (advocacy based) and bottom-up (practical field based) programming. They felt that working from both directions was the most effective in terms of advocacy enabling higher up changes that supported an environment where good welfare could be created and sustained through on the ground programming. "in an ideal world , you need to be doing advocacy , potentially , on a regional and international level . And then on all the subsequent levels down , right to the community . But you also need to be doing that community level work , and all the way up and having the combination of the two . " P27. Continuing work on the ground alongside advocating enables immediate animal welfare needs to be addressed whilst informing advocacy efforts through having a thorough understanding of community level needs. "I think it's great that there's a lot of organisations , I think , now moving to this more advocacy work , but I think , unless you know , the situation on the ground , I think it's very difficult to advocate for those people and those horses , because you don't know the context . " P13. Research Organisations differed in the level of research that they carried out. Some considered research a priority and employed dedicated research teams to support their projects. "I would focus on truly understanding the situation before you try to intervene" P5. The investment of NGOs in research can be seen across sectors including animal welfare, international development and the environment, where research outputs are used to guide programming development [ 22 – 24 ]. The need for a thorough understanding of the community context, priorities and issues were underscored when planning a programme, especially if entering a new community. "I really like how we got some decent research at the beginning of our project… we had an understanding of those places and that really helped to then design a better project from the word go . " P2. Research was described as important both in informing the direction of programming by gathering contextual information about target communities and later ensuring that the approaches utilised were creating the desired changes. Although participants acknowledged that research could be a useful tool, some voiced feasibility concerns over the level of time and resources that research requires. "I think it can be , it can be quite time consuming . And it can be quite expensive , as we all know , with research . But you know that investment is really worthwhile for that long term and sort of wider impact . " P13. Size of organisation was also a factor that affected research engagement, with small national NGOs having limited ability to carry out their own research projects. "Yeah , we're not really big research people . I think it's more capacity and time" P23. Where research was being used as a tool to identify prevalent welfare problems in an area, some participants expressed frustration. "The issues for most of the working equids are pretty simple , so they are overworked , underfed and beaten . It doesn't take a lot of research to see that . " P4. This mainly centred on the time taken to conduct research on a subject that participants either felt was obvious to identify or could feasibly start to be addressed before research results were formalised. " you know from the people or from what you see on the ground from the animals , what exactly the problem is . I think you have to start while … you are prioritising , while you are actually identifying the problems" P11. Another issue identified by participants was not to do with the research process itself but the way in which research outputs were able to be used. "actually I think that what is missing in research is a link between the university or the academic setting and the farm setting , the real life" P1. Some felt that despite research being carried out, the findings were not routinely put into practice. "I think the charity organisations they do one thing very nicely but that stops there when the research towards papers has been published , it goes there on the shelf . " P11. Planning for utilisation of research was described as important and some organisations only carried out research that they felt would be of direct practical value within their projects. Many participants described gaps where they felt further research was needed and across organisations three common topics were identified. Firstly, there was a call for research that provided further evidence on the socioeconomic contribution of equids to livelihoods. This was particularly mentioned as important for advocating for equids at a policy level. "the socioeconomic value of the donkey and the donkey-owning people , what is the contribution of these people to the GDP of the country because it goes unnoticed . Of course imagine if the brick stops ? Imagine all the houses , real-estate would collapse but people don't know the bricks are made on the donkey back" P8. Within the topic of livelihoods, it was also felt that evidence directly linking improvements in human livelihoods to animal welfare was needed. "we have been promoting the livelihood of the communities , but we don't have any evidence , scientific evidence to support that whether the increase in the livelihood improves the animal welfare or vice versa" P28. Secondly, further information was required on infectious diseases of equids, a topic that has major welfare implications but was not felt to receive enough research attention. "so many nasty diseases actually really hampering the health and welfare of equids in general in most developing countries" P11. Within the topic of infectious diseases, zoonoses were included, an area likely to receive increased research attention in the wake of the COVID-19 pandemic. The final topic that was raised by multiple participants was the need for information or guidelines for the amount of weight animals should be carrying or pulling. "how much weight can a donkey carry , that's a question that comes up a lot . " P9. Veterinary clinics The provision of free (or subsidised) veterinary care was the approach that generated the largest amount of discussion. "I think it is one of those things that divides opinions the most I'm going to say running free vet clinics . " P2. Whilst historically the most commonly used approach, a shift away from the provision of free veterinary clinics has been seen in recent years across many NGOs. A spectrum of opinions was expressed with large differences seen between the philosophies of individual organisations. The main concern regarding the provision of free veterinary clinics was that of sustainability. Criticisms both relating to the amount of resources necessary to run clinics and questions over the creation of lasting welfare change were expressed, mirroring debates in the literature [ 8 , 25 ]. "[our organisation] cannot make sustainable scale level impact by veterinary service provision directly . There are too many donkeys and not enough vets and we wouldn't be able to afford all the vets even if there were enough . " P7. Creation of dependency on the service was a barrier to independent long-term welfare changes when NGO services are likely to, at some point, be withdrawn, enabling organisations to move to another location [ 7 ]. "I think that's always the difficulty when you go in , provide that free stuff , or subsidises , it's much then harder to not provide it because it becomes an expectation , and I think we have to be aware of that" P13. It was also expressed that free clinics, as an intensive approach, reached a limited number of animals in comparison to those able to be reached by other types of initiative. "for years and years we may have worked in an area where the 200 donkeys in this immediate area are greatly benefitting from our weekly clinic but the other 2 million donkeys in this country never see us so what can we do to actually reach all these animals ? " P10. Although free clinics are one of the most immediate ways to relieve animal suffering, participants highlighted that they address the symptoms rather than the root causes of welfare problems leading to lack of long-term change. "if you're not addressing the root causes , then you're not really going to be achieving , you know essentially being a band aid kind of project . " P15. This view supports previous research advocating for a move to more prevention focussed programming [ 3 ]. Having a free service available to address animal health problems also led to questions over owner motivation to prevent those problems occurring. "the feeling of some of the staff were that the same horses just kept coming back … they felt that they were very much treating symptoms rather than addressing the causes . " P5. The lack of cost associated with veterinary treatment provided by NGOs had, in some instances, led to owner delays in treatment seeking for serious welfare issues. This potential to cause harm to the animals that were meant to benefit from treatment provision was described by one participant as a significant factor in the organisation's decision to stop providing a free clinic treatment model. "with the mobile clinics , sometimes , because people would know that it would be free , people would wait a month to seek treatment because they'd be like , well , the [NGO] van will be here in a month , and then it'll be free . So they had evidence , not just that it wasn't being successful , but that potentially , there was actually harm . " P27. There were also concerns that an unintended consequence of providing free treatment was a negative impact on local veterinary professionals through loss of business, a concern echoed in other publications [ 4 ]. "I was in a brick kiln in [country] working with one of the private para-vets…and I asked him what his biggest challenges were in his work . And he said , 'it was when you guys used to turn up and give free treatments , and nobody would pay for any of my treatment . " P27. Parallels can also be drawn with the human primary healthcare sector where NGO healthcare provision has been criticised for undermining and fragmenting local health systems [ 26 ]. However, in areas where there is no established veterinary infrastructure, NGO clinics may provide the only treatment that animals are able to access. "just recognising that there are welfare needs , and how do you address them right now , if you don't have provision of free care ? " P31. Whilst concerns regarding sustainability existed, addressing poor health and injury is the quickest way to alleviate suffering and all participants acknowledged that veterinary treatment was likely to make a large positive impact on the quality of life of the animals receiving treatment. "who can deny that veterinary clinics are good for the animals that get to see them ! " P5. In areas where health problems were severe, it was felt that efforts to address other areas of welfare ahead of or instead of veterinary treatment would achieve limited success. "saying let's sit down and talk about aspects of welfare is great but you need to reach a certain level of veterinary support before that tends to happen in my opinion . " P6. Some organisations used veterinary clinics as an initial approach when entering a new community in order to establish a positive relationship and demonstrate the professionalism and trustworthiness of the organisation. "providing those services to the communities is , is a good introduction method for your ability . So I think that's a very important aspect to include in a community . " P21. This was especially pertinent in areas where equids are not considered a priority species for receiving treatment. "So free treatments is always used as a gateway to create a win and show them that actually treating the animals improves the body condition , improves the life expectancy , improves efficiency and work capacity" P11. Employing a free clinic approach to treat the worst welfare problems in these situations can be used as an avenue to engage with owners about other aspects of welfare and lead on to discussions about preventative care. "So once we treat and we've made the donkey's life better , the livelihood is saved and the community engagement gets even better . They get really engaged with us because they know the value 'these people came here all to help our livelihoods , our donkeys so I think we should listen to them about the better harness to prevent wounds , I should listen to them about better hoof care'" P8. Veterinary clinics can also support other domains such as training and research. Clinics provide an opportunity for vet students to learn skills in recognising and treating common equine health problems and positive handling methods. This was highlighted as being especially important in countries where equines are absent from the veterinary curriculum. Clinics also offer the opportunity for research to be conducted which can form an important part of disease surveillance and prevention. "the other important thing in veterinary clinics is actually the charities they do support the student research schemes to do surveillance programmes rather than doing it by themselves in most of the cases in collaboration with the government . " P11. Acknowledging the balance of positive and negative aspects surrounding the provision of veterinary care, some participants suggested that careful and strategic planning and clear communication can minimise some concerns. "that's something that you have to be really careful when working with community members and people , unfounded expectations . And you have to be very clear , and therefore you need to have a strategy saying that , okay , we're going to be doing this for this amount of time" P21. Others preferred to use clinics only under certain circumstances or in combination with other approaches. "I don't think that long-term that is purely the best way forward . In the right time and the right place . " P2. There was also a distinction made between short-term or one-off clinics run by vets flying in from overseas and clinics that build local capacity. "it would depend how it was provided… if you're building skills in somebody , in a locally trained vet , then that is slightly different . Obviously , then , hopefully , you're creating a more sustainable , long-term solution and that you've got locally skilled people who understand local context and can provide a locally appropriate service . " P32. Para-vets (community animal health workers) The role of para-vets (sometimes called community animal health workers (CAHWs) or community based welfare advisors) was felt by many participants to be important in contributing to the sustainability of local scale health and welfare provision. "about 1000 para-vets a year I think were being trained there and they would go back to their communities and they would take all of their knowledge with them" P4. Accessibility of para-vet services was highlighted as a big advantage to communities. As community-based practitioners, para-vets are available locally and at short notice. "they are… from within the communities . So , they can help the communities in need especially in the night hours when all the hospitals and medical schools are closed . " P28. Whereas particularly in rural and remote areas, accessing more formal veterinary services at clinics or health posts can mean a long journey. "And if a community is lucky , he has to travel with the sick animal at times , over two kilometres minimum , it is very difficult . " P29. It was also mentioned that nomadic communities who are reliant on animal breeding for their livelihood, but often fall outside of the reach of veterinary services, can benefit from para-vet training. The affordability of para-vet services was felt to be an important factor in the success of para-vet programmes. "if you can make some basic first aids available at the place where they are , that is something that could be affordable . " P8. The flexibility of payment that para-vets can offer was preferred by some communities as it was more practically feasible for their livelihoods. "They can pay anytime , the government service provider has to accept the money now , you know , they don't have cash economy . Farmers can get money when they sell some of their produce during market days . They may not have money essentially . But the CAHWs can treat them readily and accepts the money next month or next week . " P29. Although limited in scope in comparison to a veterinary professional, para-vets can help reduce the pressure on veterinary services by treating more minor health issues. " They took the pressure off the team because instead of having to go out , keep getting calls from you know 'my donkey's got a cut' that could be treated by the CAHW . If the donkey had something more serious they would still go to the CAHW , he would assess it and then say 'yeah I can do that' or say 'this is beyond me' and he would then call the team" P4. Para-vets can also then act as bridge between communities and veterinarians or other equine service providers to promote preventative healthcare and positive welfare. "So apart from this , we link them with the existing services that we are supporting , like vaccination" P29. In a human health parallel, there has been a rekindling of interest in the use of community health workers to help meet child survival goals, particularly in areas of the world where there is low coverage of primary healthcare professionals [ 27 ]. Despite the positive aspects that para-vets can bring, some participants mentioned concerns about regulation of para-vet services, this especially applied to areas where para-vets had greater scope to provide treatments such as administration of drugs by injection. "some states are giving them three months of training and some states are giving them two years training , three years training . So it's very , it is a huge variation among different states . " P28. Ethical concerns were also voiced that some individuals profit from a business opportunity and as a consequence the quality of care given can be questionable. "They end up in under dosing the animals , and not giving proper doses…even though it is so crucial , it has to be strictly followed that actually they properly operate in particular areas . " P11. Incorrect dosing of animals, leading to potential increases in multidrug resistance, has also been documented in drug retail outlet workers [ 28 ]. Regulation of para-vet training, monitoring of service quality and inclusion of veterinary ethics in training were suggested as avenues to ensure that service providers were having the desired positive impact on welfare. General welfare messaging Although discussed less frequently than some of the other approaches, general welfare messaging had been utilised more frequently recently due to the COVID-19 pandemic impacting NGO's ability to provide a physical presence in communities. "So we've been trying to send messages to them , educational messages…making sure that they understand that even if we're not there physically , that we're there for them as support . " P24. Amid concerns that welfare may regress during this time, participants described welfare messaging via mobile phones as a way to keep in touch with project participants. "we send out bulk SMSs . So if we need to let them know anything important , or we want to give any tips…like with COVID , we kept on sending information about how to keep themselves safe" P23. It was described that general welfare messages were a good way to reach a large amount of people at a time in a format that was easily accessible and were used in particular to cascade preventative care messages to communities. "We're able to reach about 70 to 80% of our target group , with this approach , we also understand that we can cover also communities that are out of our intervention area . " P30. However one disadvantage was that there was little avenue for following up what the impact of the information was. "If you had done a radio announcement or broadcast…you had reached like maybe 1000 people but you had no real clear idea of what effect you had on those people" P4. In the humanitarian sector issues have been raised regarding the utility and function of general messaging such as programme signage, with concerns that increasing individual NGO visibility in order to account to donors may be prioritised over messaging efficacy [ 29 ]. Monitoring and evaluation of this type of indirect method is needed to ensure that the desired impact is being created. Microcredit The formation of microcredit groups was one of the least utilised approaches discussed despite being successfully and widely used across other NGO sectors [ 30 , 31 ]. Some participants expressed their reservations at the prospect of being responsible for others' finances. "We've done a bit looking at the microcredit but I mean , it's such a minefield for me" P13. Some felt that such approaches would not be suitable for the cultural context or communities that they worked in. "we don't get involved in their financial… it's too problematic for us to try and have like saving schemes and hold people's money" P23. Success was suggested to be contingent on a cohesive community and was described as more likely to be successful in certain countries in which community savings schemes such as Iddir in Ethiopia are common practice [ 32 ]. Those who had undertaken a microcredit approach described the benefits that could be realised when a group of people worked together to invest in improving welfare. Collectively groups could afford items that they never would have been able to purchase alone. "we had lead people in each little community group we were working with , it actually came from their idea… so that was great for trust reasons to let them lead it . And when they had enough money then , they didn't wait until everyone had enough , just each month when they grouped together they could buy a bit of a harness so they decided who would get the harness that month and they all kept going until everyone got one . " P2. Despite the potential that microcredit schemes offer, one participant observed a reluctance among donors to offer credit related types of programme. "A lot of donors are really skittish about that . Because , you know , giving people money – what is that ?! We are so afraid of giving people money because we think they are going to be , that they are going to misuse them" P21. However, it was pointed out that without access to funding, (such as that provided by seed fund schemes) equine service providers were unlikely to be able to establish themselves as viable, self-sufficient businesses. "you know , people tend to think 'but we are giving them training , that should be enough for them to , you know , create a large business and work' ! No… I think , the most important aspects for creating businesses are , you know , the accessibility of funding . " P21. Equine saving and credit groups, in certain contexts, had far wider societal implications. "two things which is good . One is the equid owners , the equid family have a direct access on that front . The other thing is that it's decreased dependency on the local moneylenders" P25. It was described that before the existence of the NGO saving and credit group, borrowing of money for emergency expenses was only available from high interest rate local moneylenders. When the equid owners were unable to pay, they entered a cycle of bonded labour; the role of moneylending in the persistence of the poverty cycle has also been highlighted in other NGO sectors [ 33 ]. The avoidance of bonded labour was mentioned as the primary reason to consider starting a microcredit scheme by an NGO that worked mainly with owners in the brick kiln industry. "So the brick kiln owner exploits the people who want money for emergencies , it might be a child's medical emergency or a marriage or a death in the family… the small little group can help them in crisis . That is something we should also try I think . I have heard about and there's a possibility we should bring it into the programme , I feel strongly" P8. Education Across organisations, a common theme in successful educational approaches was interactive, practical learning. Whether it was in the context of training vet students, government officials, owners at veterinary clinics or staff at behavioural workshops, it was felt that if learners could carry out practical activities themselves, the educational message would be better remembered and overall learning would be more effective. "personally I would always go for practical training , I don't like chalk and talk in a classroom , it is all about getting out and seeing it on the ground" P3. It has been documented that practical demonstrations of novel management techniques can accomplish recognition of the technique's benefits within the owner population [ 34 ]. Hands on contact and demonstrations with animals were seen as being particularly memorable. "the one that was most successful , I think was where we were able to do a practical session and go out and actually handle working equids and show them . " P14. This also applied to training in new techniques for members of NGO staff. "there have been animals I guess that have changed the way a whole team might think . " P6. Practical participation when introducing owners to basic management practices such as wound cleaning can be especially important as individuals need to be able to repeat the techniques unsupervised outside of a clinical setting. "And when you do the veterinary care , okay , that's when I started introducing the method of , I do it for you halfway , and I tell you what to do , and I sit there and watch you , and you do it yourself . " P22. Practical participation was also mentioned as being useful as low levels of literacy in some equid owning communities mean that other commonly used methods of communication such as written welfare messaging are unsuitable. One participant also described a higher level of community engagement for practical tasks. "not just words , because they're very much you know , tired , bored by the extension messages coming from government and talks . " P29. Interactive exercises with owners were also discussed as a method of encouraging owners to be able to visualise things from the perspective of their equid. "we prepare a sack full of sand… and then make to carry the owners of the donkey and then feel how heavy . See look . This is what you're actually loading your donkey . Can you feel how heavy it is ? " P11. Empathy has been directly implicated as an influence on animal welfare [ 35 ] and studies have demonstrated relationships between higher levels of empathy towards working horses and more accurate perception of their pain [ 13 ]. It has been suggested that in decisions relating to welfare, owner empathy may play a bigger role than socio-economic status [ 36 ] therefore the fostering of empathy in owners has been recommended as a method for the improvement of equid welfare [ 13 ]. The development of empathy was also a large focus of school directed education programmes. NGOs varied in their level of school education offered; one charity had a dedicated education team specifically to develop humane education programmes delivered across their projects. Other NGOs described running after school welfare clubs where fun activities such as singing, art and drama conveyed equine welfare messages. It was felt that children who were aware of animal welfare needs would grow up to be more empathetic owners, taking better care of their animals. "One key approach that I think should be adopted is focusing on child education , you know , from the early years on animal welfare , because these are the future generation . " P26. At a young age it was felt that children were better able to take on board new perspectives than adults who may be more entrenched in their beliefs. "it is fundamental , working with children… you know , because children really , really can change their mindset . " P21. It was described that the reach of children's education programmes extended beyond just the children themselves, an effect also seen in the field of child-centred environmental education [ 37 ]. The information given was disseminated by children to their parents and friends and generated awareness of future NGO events and initiatives. "not only will they then grow up to be members of the community , but they're also going home and talking about animal welfare to their parents . And so also potentially talking about [our organisation] if [we are] coming or… other organisations that have been coming into the communities so that we get better uptake of community training initiatives that we run . " P31. Many of the children involved in schools programmes already have roles in taking care of their family's equid and so the information given can be of direct benefit. "their neighbours , their families own an animal , and then they can influence back their parents to , to at least provide the necessary needs for animals . And these kids , you know , they provide some water to the animals , they groom them . " P29. One difficulty of school education was being able to measure the impact that programmes are having. "we have quite a large school education programme… it would be really great to be able to demonstrate the benefits that aren't just anecdotal or theoretical . But yeah again , that's a really difficult thing" P31. The lack of direct evidence from monitoring and evaluation of school education initiatives was, for some, a discouraging factor for investing in child focussed programming. "I think there are some things that take absolutely ages to be of benefit like the education of kids in schools and stuff , that sounds like something that is really really useful to do but in terms of success…you've got to be watching that project for what 15 odd years before you get an idea really of how successful it is . " P2. Community participatory exercises A focus on the human drivers of welfare had seen an increase in the use of community participatory exercises such as participatory rural appraisal (PRA) by working equid NGOs. PRA uses a variety of participatory tools such as transect walks, scoring and ranking exercises, informal mapping and the creation of diagrams to enable local people to share their perspectives and knowledge [ 38 ]. PRA approaches are well established in epidemiological research, international development and environmental education and are now becoming more frequently used in animal welfare initiatives [ 39 – 42 ]. Exercises are often pictorially based in order to ensure that group members with low levels of literacy are not excluded from participating. Participants described participatory approaches as an effective way of engaging individuals from the very beginning of the initiative process. Exercises were especially useful for identifying community views on priority welfare issues. Upjohn et al. [ 34 ] highlighted that priority welfare issues for people in target communities are not necessarily the same as those identified by an NGO, and this was echoed by participants. "actually the owners are not prioritising the kinds of issues that are implicitly assumed to be priority issues within those interventions" P32. Understanding these differences in perception can avoid previously seen scenarios where a mismatch in priorities has led to the failure of an implemented initiative through lack of community engagement. "when we started training , we see the impact is very limited , because we didn't involve the communities" P16. It was described that approaches such as PRA have enabled a move away from previously used techniques where information was imparted to the target community (with purely an educational focus) but local people did not participate in the knowledge generation, share their insights or have any ownership over the learning process and outcomes generated [ 38 ]. Higher levels of engagement and investment in initiatives was one of the biggest advantages discussed regarding the use of participatory approaches. "they feel included and more invested" P21. Participatory exercises are designed to allow individuals to share their local knowledge, perspectives and opinions. "you manage to have them on board so you are not the teacher teaching them , but having them collaborating with you" P1. Utilising these opinions to come to a shared agreement regarding the direction of welfare focussed efforts enhanced investment in the subsequent initiative. "if someone comes up with an idea they always try harder to make it work because it is their idea and they take it forward more . " P2. Enabling communities to develop their own strategies was felt to empower communities and ensure that future actions were appropriate and tailored to the local context. "instead of us telling them what to do , we try to get them to facilitate each other in terms of recognising what are the issues with animal welfare and what they can do about it . So we are empowering them and the answers came from them . " P19. Another advantage of participatory exercises discussed was their ability to identify the good practices that communities already have established. Highlighting these practices not only reinforced positive welfare aspects within sessions but allowed these practices to be built on to facilitate further welfare improvement. " in that workshop , they got the communities to identify what they already did that was really good . And that they wanted to kind of keep… . And then they got the communities to identify what they then wanted to change " P27. The long-term impact of initiatives born though participatory techniques was also felt by participants to be better. "it creates greater and more sustainable behaviour change when you do it in partnership with people . " P7. This was also linked to the ability to use participatory exercises to identify and discuss the root causes of welfare problems. "I think that that's very important to find out , you know , what are the issues and not just the symptoms but actually delving down into what are the causes of something" P15. Being able to then address the root causes of welfare problems represented an avenue for sustainable welfare improvement. Skills initiatives These initiatives aim to train individuals in specialist (non-veterinary) skills such as farriery, saddlery and harnessing in order to create businesses providing those services to the community. Participants described: "education and skills initiative…this is a very crucial action in creating awareness and equipping the communities…with basic knowledge and skills" P11. Studies have reported as many as 78% of equids in a population affected by working equipment related injuries, and these skills initiatives can be especially useful in such areas [ 43 ]. "they taught the para-vets how to make basic harness , just pack saddles and very basic harness but that's had a massive effect . The knock-on effect from that was tremendous" P4. With huge variability seen in the standards of local service providers [ 14 ], creating a network of highly trained providers can ensure that both animals and owners are receiving high quality services, helping to maintain welfare within the community. "where we were based…pretty much all of the owners would use one of the farriers that we trained . I think that was so successful" P2. Most initiatives aim to train members of a community so that they can become full-time service providers within their community. "you want for them to work with you to be viable local businesses . " P21. However for this to happen, individuals need to have a financially feasible business which some participants described as difficult to establish. "People were really keen to learn how to make stuff but trying to then sell it was quite hard because it was obviously way more expensive than like old clothing or the rags people were using" P2. In order for initiatives to be successful, there needed to be a demand within the community for the service that was being offered. "just because you create a service doesn't necessarily mean that you fixed the problem" P32. It was felt that some previously run initiatives had been unsuccessful because the skills imparted by NGOs, and services subsequently offered, were not targeting priority issues for local people and resulted in a lack of investment. Differences in welfare priorities between communities and NGOs have been previously noted [ 34 ]. Due to this need for service demand, some organisations only focussed on upskilling existing service providers within communities. "we don't want to start from zero we just want to start where there are already service providers . " P26. In this scenario individuals are already established which removes the barrier of creating both service demand and local infrastructure. "there was already this setup to use that service so it wasn't like this big stepping stone it was just improving a service that was already there . " P2. Relatively small changes to the operations of these individuals can then have widespread effects within a community. Another challenge discussed was dropout of individuals enrolled in training courses. "you're taking 10 people… at the end of the year you are left with six people . So my concern is those four people why are they leaving ? " P22. As courses are intensive to teach, a substantial amount of resources and time are invested in candidates and participant dropout was a significant loss. Trained individuals subsequently leaving their communities was another area of concern discussed which was felt to hinder the success of initiatives looking to build networks of local service providers within an area. Participants highlighted the need for continued support for skills initiatives, particularly for ensuring the longevity of service provision. Previously used models for the delivery of skills initiatives saw UK based professionals teaching courses for short time periods in target countries. This was not felt to be adequate for the long-term maintenance of these skills within communities. "If you do it in a way that's shipping them in and then inject some knowledge and go away again , I have found that those are very unsuccessful although they have a lot of excitement at the time . " P5. Many NGOs have now shifted to in-country delivery of skills training programmes which allows courses to be taught by local professionals, tailored to the local context and provides increased scope for ongoing support. Barriers to welfare improvement Although participants discussed the strengths and weaknesses of the different initiatives used, similarities emerged in factors that impacted NGO programming regardless of the approach taken. These inherent challenges when trying to engage in welfare improvement initiatives cross-cut initiative categories and were widely described by participants working across countries. Equid status One such barrier, described across many institutions, was people's attitude towards equids, both at the community and policymaking level. This was especially apparent in the case of donkeys who, across countries, are frequently associated with low status. " even now in [this country] donkeys are not considered great… most of the people , not even the veterinarians want to go and treat the donkey , thinking it is a very neglected abused animal . " P8. It was felt that donkeys' association with menial or household work and comparatively lower sale price influenced opinion of their 'worth'; this has also been documented in other research studies [ 12 , 44 ]. "Donkeys are generally not taken care of well by the communities because they are cheaper and they are let loose for grazing" P28. This results in a systematic lack of basic care for donkeys, which is not reflected for other equine species. The perception within communities that donkeys are low maintenance and 'take care of themselves' was frequently mentioned. "malnutrition is one of the problems because as I said , donkeys their priority is so down some of them don't even feed them . Because donkeys normally… get whatever they get from roaming that is how they live" P11. Due to donkeys' low status, encouraging consideration of their welfare needs can be met with incredulity and sometimes derision. "apparently lots of them turned up just because they couldn't believe there was going to be a project about donkeys and they just came to find out , because that was the ultimate craziest thing [a project about] being nice to a donkey . " P5. The association of equids with low status can also extend to professions related to equids such as farriery. This perception and a lack of respect associated with the profession can prevent the recruitment and retention of individuals to farriery training programmes. "it's not a very respected profession . So there's not much pride in being a farrier . Often , the animal's feet are described as dirty , farriery is therefore described as a dirty job" P27. The cultural status of equids in some areas also affected their treatment by vets. "so getting them to engage with the idea of equids as a species you want to be a vet for is often challenging . " P32. The potential reasons for this were discussed and included: a lack of equid inclusion in veterinary curricula, wider societal views that do not prioritise equids and a lack of uptake of services as some owners are reluctant to pay for treatment for equids. Despite their importance to livelihoods and domestic tasks, the value of equids and the activities they perform are not always recognised by owners. "I would definitely hear owners saying , there's no point , you know , there's no point putting money into donkeys because they die anyway . " P32. Other livestock (those that provide a more direct benefit such as milk or meat) are often considered to be more important and as a consequence receive more care and veterinary attention in situations where financial resources are scarce [ 45 ]. This situation is also mirrored in the advocacy struggles encountered by NGO staff when trying to influence policy changes in favour of working equids. "the invisible equine within the development agenda is a problem and then animal welfare is also a bit of a blind spot . " P15. A lack of recognition of the value of equids by governments and policymakers both at local and national levels was described. "And when you actually look at the population statistics with FAO , some countries don't even register that they've actually got working equids . " P14. It was felt that research demonstrating the contribution that equids make could favourably influence their status both in the eyes of the public and policymakers. "we need to bring a change in the attitude of the general public of how good and innocent and hard-working an animal it is . How is their contribution to the economy" P8. There was understanding that other important societal issues will compete for government funds and legislative action but issues around working equids appeared to be low on policymakers' list of priorities. "In Africa you've got poverty , unemployment , HIV/AIDS . Huge , these are huge cross-cutting issues . They equally demand funding and skills" P15. The indirect benefits that working equids provide make it difficult to quantify their value to a country's economy [ 4 ]; and in countries where working equid populations are decreasing (as mechanisation becomes more widespread), there were concerns that their position is viewed as irrelevant to a countries' future growth. "the population of equids as per the government census has decreased significantly … so the government is not focusing on it…the equids are not a priority" P28. Working equids' exclusion from livestock policies was a barrier that was discussed across organisations and NGOs, and advocating for their inclusion was a focus of their programming. Successful advocacy resulted in the acknowledgement of equids as 'working livestock' contributing to food security by the UN Committee of Food Security in 2016 [ 46 ]. Recognising working equids as livestock covered within national livestock policy definitions can have a range of beneficial effects [ 47 ]. These include access to national animal health strategies and services such as vaccination campaigns, disease surveillance and emergency relief schemes [ 45 ]. One participant highlighted the need for equid inclusion in current livestock disease control strategies, particularly in the case of zoonotic diseases that pose a serious risk to human health such as anthrax. "it affects cattle , it affects sheep , goats , horse , donkey , human , and other animals too . The funny thing is the government actually do the vaccine against cattle , but they don't vaccinate against donkeys… they're actually opening a pocket , the problem is still there" P11. Leaving equids unvaccinated can create reservoirs of disease in populations that can then aid transmission to other animals or potentially humans. Access Accessibility was raised as a barrier to the success of welfare initiatives, both regarding access to communities and access to all individuals responsible for maintaining equid welfare. It was discussed that some equid owning communities live in very remote rural areas where physical access is limited by the environment, such as mountain communities. By virtue of their position, these are often the communities that rely heavily on working equids for many roles including the transport of people and goods across difficult to navigate terrain where the use of a vehicle is not possible [ 48 ]. But nevertheless, for NGOs attempting to reach these communities, access is a challenge. "I think it's more their location makes it quite difficult for even our teams that are even based near them to even be able to access them , to bring the stuff that they need to kind of do the work . And it's usually the worst welfare conditions are in the areas that are the most remote . " P17. The COVID-19 pandemic has further exacerbated issues surrounding community access, a finding that has also been documented across NGO sectors [ 49 ]. It was described that remote communities also have problems accessing veterinary services due to their distance. This was especially mentioned in the case of brick kilns which, due to pollution regulations, are situated away from urban areas. "even if they want to call veterinary services , people don't come because of the distance and the remoteness of the place . " P8. Access to the full spectrum of individuals involved in the day-to-day care of equids was also another challenge highlighted by professionals. This is especially pertinent to cultures where the gender division of labour can mean that the individual working with the animal is not necessarily the person who undertakes routine equid management or attends an animal clinic. " So for example in [country] you often have children bring the donkey to the clinic for treatment and I remember on a particular occasion a young girl had brought a donkey , it had abscesses in both front feet and the vet said 'I thought I told you , you weren't to work the donkey' , clearly by the foot pads the donkey had been worked in between the treatments and the girl's reply through translation was 'I know what you said but how do I tell my father ? '" P6. Many participants described the need for programmatic activities such as meetings, workshops and educational events to include all individuals responsible for equid care. However, this was sometimes challenging; cultural norms in some areas can prohibit mixed-gender community participation and different time constraints across genders can mean that certain individuals are not available to take part in NGO programmes. In these situations NGOs can struggle to include the views of, or impart knowledge to all of the people necessary to make improving welfare successful. A lack of equid trained vets A frequently highlighted barrier to the success of equid welfare projects was the absence of training related to equines in the curriculum of veterinarians across countries. This has resulted in local veterinary service providers who do not have the specialist knowledge or skills to treat equine disease or injury. As many communities do not have the option to seek treatment for their animals elsewhere, this lack of access to appropriate treatment further compounds the poor welfare state of working equids. "If they manage to get somebody there for the veterinary services , the person who deals with the case aren't trained or have the skills to deal with donkeys and mules . " P8. While production animals appear to be the focus of much of the veterinary curriculum, equids in some areas are not commonly encountered by vets. "5th year vet students and they've not touched a horse , they're scared of it ! " P13. In some areas, even if there is provision made for the veterinary treatment of horses, knowledge of donkey treatment is severely lacking. It is common that no differentiation is made between treatments for different equid species. "Another problem is that the majority of people think that donkeys are just a small horse and they have the same needs" P1. Across NGOs there was a programming focus on working in partnership with veterinary universities to bring elements of equine training to the curriculum. "So we've tried to do a 101 working equid curriculum . That might be ideally , teachers that will take this on board and make them their own and listen to our recordings . And we're building slowly , sort of a database that they can have . " P31. Programmes where vet students can gain experience in the handling and treatment of equids were also run by NGOs with the hope of ensuring a better coverage of equid trained vets returning to their regions. "when the vets come and work with us , they can come with us in the field as part of the internship programme , start touching the donkeys and injecting donkeys , in our presence so it builds their confidence . " P8. Equid trained para-vets can also help to fill this gap in provision. Monitoring and evaluation challenges Many participants described an increased focus on monitoring and evaluation (M&E) of initiatives. This focus parallels changes seen far earlier in the field of international development where the need for accountability in the use of governmental and institutional funding resulted in more rigorous reporting and evaluation systems [ 50 ]. M&E was not perceived negatively by participants, all participants were in favour of tracking the progress of initiatives and felt that M&E was important to know how much of an impact their projects were making. However comprehensive and structured M&E processes were described across virtually all organisations as a relatively new introduction and rigorous M&E across the different approaches now utilised by organisations was highlighted as a challenge. "I would say this is relatively new to the programme so…it is a little bit trial and error . And it's also quite new to our teams . So there's a lot of work in that" P13. Apprehension was expressed that the increased need for formal evaluation has put extra pressure and workload onto those collecting M&E data at ground level. "So there's a lot of paperwork which the field team hates ! " P23. It also carried the risk that the time spent evaluating projects would result in less time devoted to carrying out the activities that would create desired change, a focus on upwards accountability seen in other NGO sectors [ 29 ]. "suddenly we went from actually going out there helping donkeys to spending all our time trying to prove we were helping donkeys" P4. With the implementation of new frameworks came concern that the expectations were not compatible with the practicalities of situations on the ground. "it was a very sort of tricky conversation to say , you know , we are not sure if this is realistic" P32. Participants spoke about the need to use indicators that are 'meaningful'. Historically numbers of animals reached by NGO programmes were the most common indicator of success measured [ 5 ]. However, this did not capture the impact that initiatives had on those animals reached and whether this led to any type of long-term welfare benefit. "Instead of just saying we treated , I think it used to be like 360 , 000 donkeys a year at that time , as [a colleague] quite rightly said 'that could be one donkey 360 , 000 times because you treated it really badly and haven't done anything for it ! ' What does it mean ? It doesn't mean anything" P4. It was also raised that what is deemed a meaningful indicator may differ by area. "So we also have the balance between indicators that Western developed countries think of as indicators versus what are really useful indicators here in more developing countries . " P19. With the focus shifting from numbers of animals reached to longer-term success, interviewees expressed difficulties in finding the most effective indicators to measure complex outcomes that can be influenced by multiple factors. This was especially true of multidisciplinary initiatives which seek to improve equid welfare indirectly through livelihood enhancement. "I think the biggest challenge with that is that it is successful , but it's been hard to measure how… how can you then know that that's directly improving equine welfare , if you're improving someone's livelihood" P20. It was also a concern that external factors could lead to an unrepresentative negative evaluation. "actually measuring real impact , particularly when…in the middle of your three years , there's a massive drought and people lose their livestock and body condition goes right down . You've been doing these beautiful welfare assessments and then suddenly… there's so much else that how do you then have all this data and have the thousand caveats that come with it ? " P13. However one participant suggested that the perceived difficulties surrounding comprehensive M&E can be unfounded. " one of the myths is that M&E is expensive and complicated and it certainly doesn't have to be . You can do it at a lot of different levels and I think that often the baby is thrown out with the bath water because people think you have to do so much when really you can still gain a lot of information from a smaller but dedicated and strategic approach ." P5. It was suggested that smaller changes, more reflective of the incremental nature of changes made by welfare initiatives could be a more suitable avenue than trying to immediately link initiatives to direct welfare improvement that may take far longer to be realised. "I think there is not enough emphasis placed on smaller indicators and looking at success in other ways… . So for example provision of water could be a small one , you could say 100% of people weren't providing water all the time that their animals were stabled and now 80% are , and although that might not lead all the way through to animal-based indicators of improvements in dehydration say , you've got those human behaviour indicators of what they're doing" P5. Programmatic approaches to improving welfare The nine most commonly utilised approaches within equid focussed NGO programming were explicitly discussed with participants. Approaches were considered in turn and the associated advantages, limitations and potential consequences of each were considered. Advocacy It was apparent that there has been a relatively recent shift placing advocacy as a strong focus across virtually all organisations. This is not unique to the working equid sector; the growth of advocacy across both generalist and specialist UK based NGOs has been documented [ 18 ]. A larger focus on advocacy was perceived as having both benefits and disadvantages. Some participants felt that advocating for policy change was the best way to enhance the lives of large numbers of animals, many more than could be reached through more practical, on the ground initiatives. "I actually think targeting governments , and getting them to implement the [WOAH] chapters and getting them to enforce good welfare is probably your best way of making a huge difference , and having a huge reach . " P14. It was recognised that advocacy can help change systems in a strategic way that aims to address root causes of poor welfare, rather than addressing the symptoms of individual animals. "going for the right level would get better spread , better results , I think , in the end but it's more of a long-term plan…firefighting is never going to get you a result at the end of the day . It's not a strategic way of thinking . " P14. Recent advocacy efforts have seen equid welfare NGOs working together to form groups such as the International Coalition for Working Equids to implement the World Organisation for Animal Health (WOAH) Welfare of Working Equids Terrestrial Code Chapter 7.12 [ 19 ]. Other groups such as the Equid Power Network [ 20 ] have been involved in promoting educational events for the public and professionals across disciplines. It was acknowledged that for certain large-scale international problems (such as the trade in donkey skins [ 21 ]), correspondingly large-scale advocacy was likely to be the only effective solution. However, it was felt by some participants that the current prominence of advocacy on NGO agendas came at the expense of other approaches. "I think , particularly nowadays , every charity organisation is actually advocating advocacy… It seems as if the practical work on the ground has been neglected" P11. It was also expressed that advocacy could be essentially risky in terms of its rewards compared to the effort put in to achieve changes. Dependent upon the political climate, policy changes can be difficult to achieve and the process was described as being both lengthy and often frustrating for NGO staff. "the least successful approach…(is) the engagement with the policymakers because these policymakers they are bureaucrats and the government systems are very slow" P28. A second highlighted issue was that even if an advocacy strategy was successful in creating a policy change, this was no guarantee that the effect of the change would be felt by animals and communities. "So that [law] existed when I first started working with the project and I don't think I ever saw it enforced and I was there for 2 and a half years . So like although it was written down it didn't make any difference on the ground as such . That's just one of those things isn't it ? " P2. A gap due to lack of enforcement of welfare legislation was often discussed as an obstacle to achieving welfare change. "There is some policy on document , this policy has not been implemented , you know , these are really challenges for us . " P25. A number of reasons were suggested for failure to enforce regulations including a lack of staff capacity, lack of resources at the disposal of the enforcing body and a lack of education of enforcers. "policeman for example . Suddenly he is told that if he sees someone beating a donkey or carrying more than 300 kilos weight that man is… breaking the law . So he sees somebody carrying 200 kilos in the cart with 2 flat tyres , a bad harness and you know dogs biting the donkeys and everything else . And he's not broken the rules so he lets him go , that's the first problem . That the people that are enforcing the rules don't understand what they are enforcing…without that knowledge they can't make a judgement . " P4. These factors are often systematic within enforcement entities (most commonly police or government enforcement bodies) which presents another challenge for NGOs who can have little control over how the policy is eventually enforced. This led to discussions that some political infrastructures are less likely to successfully support effective policy changes. "lobbying and influencing future policy regarding equids can have greater and lasting impacts on welfare…mainly when the country is … resourceful enough to enforce it . " P29. Although acknowledging that it was resource intensive, some participants described an ideal balance needed of top down (advocacy based) and bottom-up (practical field based) programming. They felt that working from both directions was the most effective in terms of advocacy enabling higher up changes that supported an environment where good welfare could be created and sustained through on the ground programming. "in an ideal world , you need to be doing advocacy , potentially , on a regional and international level . And then on all the subsequent levels down , right to the community . But you also need to be doing that community level work , and all the way up and having the combination of the two . " P27. Continuing work on the ground alongside advocating enables immediate animal welfare needs to be addressed whilst informing advocacy efforts through having a thorough understanding of community level needs. "I think it's great that there's a lot of organisations , I think , now moving to this more advocacy work , but I think , unless you know , the situation on the ground , I think it's very difficult to advocate for those people and those horses , because you don't know the context . " P13. Research Organisations differed in the level of research that they carried out. Some considered research a priority and employed dedicated research teams to support their projects. "I would focus on truly understanding the situation before you try to intervene" P5. The investment of NGOs in research can be seen across sectors including animal welfare, international development and the environment, where research outputs are used to guide programming development [ 22 – 24 ]. The need for a thorough understanding of the community context, priorities and issues were underscored when planning a programme, especially if entering a new community. "I really like how we got some decent research at the beginning of our project… we had an understanding of those places and that really helped to then design a better project from the word go . " P2. Research was described as important both in informing the direction of programming by gathering contextual information about target communities and later ensuring that the approaches utilised were creating the desired changes. Although participants acknowledged that research could be a useful tool, some voiced feasibility concerns over the level of time and resources that research requires. "I think it can be , it can be quite time consuming . And it can be quite expensive , as we all know , with research . But you know that investment is really worthwhile for that long term and sort of wider impact . " P13. Size of organisation was also a factor that affected research engagement, with small national NGOs having limited ability to carry out their own research projects. "Yeah , we're not really big research people . I think it's more capacity and time" P23. Where research was being used as a tool to identify prevalent welfare problems in an area, some participants expressed frustration. "The issues for most of the working equids are pretty simple , so they are overworked , underfed and beaten . It doesn't take a lot of research to see that . " P4. This mainly centred on the time taken to conduct research on a subject that participants either felt was obvious to identify or could feasibly start to be addressed before research results were formalised. " you know from the people or from what you see on the ground from the animals , what exactly the problem is . I think you have to start while … you are prioritising , while you are actually identifying the problems" P11. Another issue identified by participants was not to do with the research process itself but the way in which research outputs were able to be used. "actually I think that what is missing in research is a link between the university or the academic setting and the farm setting , the real life" P1. Some felt that despite research being carried out, the findings were not routinely put into practice. "I think the charity organisations they do one thing very nicely but that stops there when the research towards papers has been published , it goes there on the shelf . " P11. Planning for utilisation of research was described as important and some organisations only carried out research that they felt would be of direct practical value within their projects. Many participants described gaps where they felt further research was needed and across organisations three common topics were identified. Firstly, there was a call for research that provided further evidence on the socioeconomic contribution of equids to livelihoods. This was particularly mentioned as important for advocating for equids at a policy level. "the socioeconomic value of the donkey and the donkey-owning people , what is the contribution of these people to the GDP of the country because it goes unnoticed . Of course imagine if the brick stops ? Imagine all the houses , real-estate would collapse but people don't know the bricks are made on the donkey back" P8. Within the topic of livelihoods, it was also felt that evidence directly linking improvements in human livelihoods to animal welfare was needed. "we have been promoting the livelihood of the communities , but we don't have any evidence , scientific evidence to support that whether the increase in the livelihood improves the animal welfare or vice versa" P28. Secondly, further information was required on infectious diseases of equids, a topic that has major welfare implications but was not felt to receive enough research attention. "so many nasty diseases actually really hampering the health and welfare of equids in general in most developing countries" P11. Within the topic of infectious diseases, zoonoses were included, an area likely to receive increased research attention in the wake of the COVID-19 pandemic. The final topic that was raised by multiple participants was the need for information or guidelines for the amount of weight animals should be carrying or pulling. "how much weight can a donkey carry , that's a question that comes up a lot . " P9. Veterinary clinics The provision of free (or subsidised) veterinary care was the approach that generated the largest amount of discussion. "I think it is one of those things that divides opinions the most I'm going to say running free vet clinics . " P2. Whilst historically the most commonly used approach, a shift away from the provision of free veterinary clinics has been seen in recent years across many NGOs. A spectrum of opinions was expressed with large differences seen between the philosophies of individual organisations. The main concern regarding the provision of free veterinary clinics was that of sustainability. Criticisms both relating to the amount of resources necessary to run clinics and questions over the creation of lasting welfare change were expressed, mirroring debates in the literature [ 8 , 25 ]. "[our organisation] cannot make sustainable scale level impact by veterinary service provision directly . There are too many donkeys and not enough vets and we wouldn't be able to afford all the vets even if there were enough . " P7. Creation of dependency on the service was a barrier to independent long-term welfare changes when NGO services are likely to, at some point, be withdrawn, enabling organisations to move to another location [ 7 ]. "I think that's always the difficulty when you go in , provide that free stuff , or subsidises , it's much then harder to not provide it because it becomes an expectation , and I think we have to be aware of that" P13. It was also expressed that free clinics, as an intensive approach, reached a limited number of animals in comparison to those able to be reached by other types of initiative. "for years and years we may have worked in an area where the 200 donkeys in this immediate area are greatly benefitting from our weekly clinic but the other 2 million donkeys in this country never see us so what can we do to actually reach all these animals ? " P10. Although free clinics are one of the most immediate ways to relieve animal suffering, participants highlighted that they address the symptoms rather than the root causes of welfare problems leading to lack of long-term change. "if you're not addressing the root causes , then you're not really going to be achieving , you know essentially being a band aid kind of project . " P15. This view supports previous research advocating for a move to more prevention focussed programming [ 3 ]. Having a free service available to address animal health problems also led to questions over owner motivation to prevent those problems occurring. "the feeling of some of the staff were that the same horses just kept coming back … they felt that they were very much treating symptoms rather than addressing the causes . " P5. The lack of cost associated with veterinary treatment provided by NGOs had, in some instances, led to owner delays in treatment seeking for serious welfare issues. This potential to cause harm to the animals that were meant to benefit from treatment provision was described by one participant as a significant factor in the organisation's decision to stop providing a free clinic treatment model. "with the mobile clinics , sometimes , because people would know that it would be free , people would wait a month to seek treatment because they'd be like , well , the [NGO] van will be here in a month , and then it'll be free . So they had evidence , not just that it wasn't being successful , but that potentially , there was actually harm . " P27. There were also concerns that an unintended consequence of providing free treatment was a negative impact on local veterinary professionals through loss of business, a concern echoed in other publications [ 4 ]. "I was in a brick kiln in [country] working with one of the private para-vets…and I asked him what his biggest challenges were in his work . And he said , 'it was when you guys used to turn up and give free treatments , and nobody would pay for any of my treatment . " P27. Parallels can also be drawn with the human primary healthcare sector where NGO healthcare provision has been criticised for undermining and fragmenting local health systems [ 26 ]. However, in areas where there is no established veterinary infrastructure, NGO clinics may provide the only treatment that animals are able to access. "just recognising that there are welfare needs , and how do you address them right now , if you don't have provision of free care ? " P31. Whilst concerns regarding sustainability existed, addressing poor health and injury is the quickest way to alleviate suffering and all participants acknowledged that veterinary treatment was likely to make a large positive impact on the quality of life of the animals receiving treatment. "who can deny that veterinary clinics are good for the animals that get to see them ! " P5. In areas where health problems were severe, it was felt that efforts to address other areas of welfare ahead of or instead of veterinary treatment would achieve limited success. "saying let's sit down and talk about aspects of welfare is great but you need to reach a certain level of veterinary support before that tends to happen in my opinion . " P6. Some organisations used veterinary clinics as an initial approach when entering a new community in order to establish a positive relationship and demonstrate the professionalism and trustworthiness of the organisation. "providing those services to the communities is , is a good introduction method for your ability . So I think that's a very important aspect to include in a community . " P21. This was especially pertinent in areas where equids are not considered a priority species for receiving treatment. "So free treatments is always used as a gateway to create a win and show them that actually treating the animals improves the body condition , improves the life expectancy , improves efficiency and work capacity" P11. Employing a free clinic approach to treat the worst welfare problems in these situations can be used as an avenue to engage with owners about other aspects of welfare and lead on to discussions about preventative care. "So once we treat and we've made the donkey's life better , the livelihood is saved and the community engagement gets even better . They get really engaged with us because they know the value 'these people came here all to help our livelihoods , our donkeys so I think we should listen to them about the better harness to prevent wounds , I should listen to them about better hoof care'" P8. Veterinary clinics can also support other domains such as training and research. Clinics provide an opportunity for vet students to learn skills in recognising and treating common equine health problems and positive handling methods. This was highlighted as being especially important in countries where equines are absent from the veterinary curriculum. Clinics also offer the opportunity for research to be conducted which can form an important part of disease surveillance and prevention. "the other important thing in veterinary clinics is actually the charities they do support the student research schemes to do surveillance programmes rather than doing it by themselves in most of the cases in collaboration with the government . " P11. Acknowledging the balance of positive and negative aspects surrounding the provision of veterinary care, some participants suggested that careful and strategic planning and clear communication can minimise some concerns. "that's something that you have to be really careful when working with community members and people , unfounded expectations . And you have to be very clear , and therefore you need to have a strategy saying that , okay , we're going to be doing this for this amount of time" P21. Others preferred to use clinics only under certain circumstances or in combination with other approaches. "I don't think that long-term that is purely the best way forward . In the right time and the right place . " P2. There was also a distinction made between short-term or one-off clinics run by vets flying in from overseas and clinics that build local capacity. "it would depend how it was provided… if you're building skills in somebody , in a locally trained vet , then that is slightly different . Obviously , then , hopefully , you're creating a more sustainable , long-term solution and that you've got locally skilled people who understand local context and can provide a locally appropriate service . " P32. Para-vets (community animal health workers) The role of para-vets (sometimes called community animal health workers (CAHWs) or community based welfare advisors) was felt by many participants to be important in contributing to the sustainability of local scale health and welfare provision. "about 1000 para-vets a year I think were being trained there and they would go back to their communities and they would take all of their knowledge with them" P4. Accessibility of para-vet services was highlighted as a big advantage to communities. As community-based practitioners, para-vets are available locally and at short notice. "they are… from within the communities . So , they can help the communities in need especially in the night hours when all the hospitals and medical schools are closed . " P28. Whereas particularly in rural and remote areas, accessing more formal veterinary services at clinics or health posts can mean a long journey. "And if a community is lucky , he has to travel with the sick animal at times , over two kilometres minimum , it is very difficult . " P29. It was also mentioned that nomadic communities who are reliant on animal breeding for their livelihood, but often fall outside of the reach of veterinary services, can benefit from para-vet training. The affordability of para-vet services was felt to be an important factor in the success of para-vet programmes. "if you can make some basic first aids available at the place where they are , that is something that could be affordable . " P8. The flexibility of payment that para-vets can offer was preferred by some communities as it was more practically feasible for their livelihoods. "They can pay anytime , the government service provider has to accept the money now , you know , they don't have cash economy . Farmers can get money when they sell some of their produce during market days . They may not have money essentially . But the CAHWs can treat them readily and accepts the money next month or next week . " P29. Although limited in scope in comparison to a veterinary professional, para-vets can help reduce the pressure on veterinary services by treating more minor health issues. " They took the pressure off the team because instead of having to go out , keep getting calls from you know 'my donkey's got a cut' that could be treated by the CAHW . If the donkey had something more serious they would still go to the CAHW , he would assess it and then say 'yeah I can do that' or say 'this is beyond me' and he would then call the team" P4. Para-vets can also then act as bridge between communities and veterinarians or other equine service providers to promote preventative healthcare and positive welfare. "So apart from this , we link them with the existing services that we are supporting , like vaccination" P29. In a human health parallel, there has been a rekindling of interest in the use of community health workers to help meet child survival goals, particularly in areas of the world where there is low coverage of primary healthcare professionals [ 27 ]. Despite the positive aspects that para-vets can bring, some participants mentioned concerns about regulation of para-vet services, this especially applied to areas where para-vets had greater scope to provide treatments such as administration of drugs by injection. "some states are giving them three months of training and some states are giving them two years training , three years training . So it's very , it is a huge variation among different states . " P28. Ethical concerns were also voiced that some individuals profit from a business opportunity and as a consequence the quality of care given can be questionable. "They end up in under dosing the animals , and not giving proper doses…even though it is so crucial , it has to be strictly followed that actually they properly operate in particular areas . " P11. Incorrect dosing of animals, leading to potential increases in multidrug resistance, has also been documented in drug retail outlet workers [ 28 ]. Regulation of para-vet training, monitoring of service quality and inclusion of veterinary ethics in training were suggested as avenues to ensure that service providers were having the desired positive impact on welfare. General welfare messaging Although discussed less frequently than some of the other approaches, general welfare messaging had been utilised more frequently recently due to the COVID-19 pandemic impacting NGO's ability to provide a physical presence in communities. "So we've been trying to send messages to them , educational messages…making sure that they understand that even if we're not there physically , that we're there for them as support . " P24. Amid concerns that welfare may regress during this time, participants described welfare messaging via mobile phones as a way to keep in touch with project participants. "we send out bulk SMSs . So if we need to let them know anything important , or we want to give any tips…like with COVID , we kept on sending information about how to keep themselves safe" P23. It was described that general welfare messages were a good way to reach a large amount of people at a time in a format that was easily accessible and were used in particular to cascade preventative care messages to communities. "We're able to reach about 70 to 80% of our target group , with this approach , we also understand that we can cover also communities that are out of our intervention area . " P30. However one disadvantage was that there was little avenue for following up what the impact of the information was. "If you had done a radio announcement or broadcast…you had reached like maybe 1000 people but you had no real clear idea of what effect you had on those people" P4. In the humanitarian sector issues have been raised regarding the utility and function of general messaging such as programme signage, with concerns that increasing individual NGO visibility in order to account to donors may be prioritised over messaging efficacy [ 29 ]. Monitoring and evaluation of this type of indirect method is needed to ensure that the desired impact is being created. Microcredit The formation of microcredit groups was one of the least utilised approaches discussed despite being successfully and widely used across other NGO sectors [ 30 , 31 ]. Some participants expressed their reservations at the prospect of being responsible for others' finances. "We've done a bit looking at the microcredit but I mean , it's such a minefield for me" P13. Some felt that such approaches would not be suitable for the cultural context or communities that they worked in. "we don't get involved in their financial… it's too problematic for us to try and have like saving schemes and hold people's money" P23. Success was suggested to be contingent on a cohesive community and was described as more likely to be successful in certain countries in which community savings schemes such as Iddir in Ethiopia are common practice [ 32 ]. Those who had undertaken a microcredit approach described the benefits that could be realised when a group of people worked together to invest in improving welfare. Collectively groups could afford items that they never would have been able to purchase alone. "we had lead people in each little community group we were working with , it actually came from their idea… so that was great for trust reasons to let them lead it . And when they had enough money then , they didn't wait until everyone had enough , just each month when they grouped together they could buy a bit of a harness so they decided who would get the harness that month and they all kept going until everyone got one . " P2. Despite the potential that microcredit schemes offer, one participant observed a reluctance among donors to offer credit related types of programme. "A lot of donors are really skittish about that . Because , you know , giving people money – what is that ?! We are so afraid of giving people money because we think they are going to be , that they are going to misuse them" P21. However, it was pointed out that without access to funding, (such as that provided by seed fund schemes) equine service providers were unlikely to be able to establish themselves as viable, self-sufficient businesses. "you know , people tend to think 'but we are giving them training , that should be enough for them to , you know , create a large business and work' ! No… I think , the most important aspects for creating businesses are , you know , the accessibility of funding . " P21. Equine saving and credit groups, in certain contexts, had far wider societal implications. "two things which is good . One is the equid owners , the equid family have a direct access on that front . The other thing is that it's decreased dependency on the local moneylenders" P25. It was described that before the existence of the NGO saving and credit group, borrowing of money for emergency expenses was only available from high interest rate local moneylenders. When the equid owners were unable to pay, they entered a cycle of bonded labour; the role of moneylending in the persistence of the poverty cycle has also been highlighted in other NGO sectors [ 33 ]. The avoidance of bonded labour was mentioned as the primary reason to consider starting a microcredit scheme by an NGO that worked mainly with owners in the brick kiln industry. "So the brick kiln owner exploits the people who want money for emergencies , it might be a child's medical emergency or a marriage or a death in the family… the small little group can help them in crisis . That is something we should also try I think . I have heard about and there's a possibility we should bring it into the programme , I feel strongly" P8. Education Across organisations, a common theme in successful educational approaches was interactive, practical learning. Whether it was in the context of training vet students, government officials, owners at veterinary clinics or staff at behavioural workshops, it was felt that if learners could carry out practical activities themselves, the educational message would be better remembered and overall learning would be more effective. "personally I would always go for practical training , I don't like chalk and talk in a classroom , it is all about getting out and seeing it on the ground" P3. It has been documented that practical demonstrations of novel management techniques can accomplish recognition of the technique's benefits within the owner population [ 34 ]. Hands on contact and demonstrations with animals were seen as being particularly memorable. "the one that was most successful , I think was where we were able to do a practical session and go out and actually handle working equids and show them . " P14. This also applied to training in new techniques for members of NGO staff. "there have been animals I guess that have changed the way a whole team might think . " P6. Practical participation when introducing owners to basic management practices such as wound cleaning can be especially important as individuals need to be able to repeat the techniques unsupervised outside of a clinical setting. "And when you do the veterinary care , okay , that's when I started introducing the method of , I do it for you halfway , and I tell you what to do , and I sit there and watch you , and you do it yourself . " P22. Practical participation was also mentioned as being useful as low levels of literacy in some equid owning communities mean that other commonly used methods of communication such as written welfare messaging are unsuitable. One participant also described a higher level of community engagement for practical tasks. "not just words , because they're very much you know , tired , bored by the extension messages coming from government and talks . " P29. Interactive exercises with owners were also discussed as a method of encouraging owners to be able to visualise things from the perspective of their equid. "we prepare a sack full of sand… and then make to carry the owners of the donkey and then feel how heavy . See look . This is what you're actually loading your donkey . Can you feel how heavy it is ? " P11. Empathy has been directly implicated as an influence on animal welfare [ 35 ] and studies have demonstrated relationships between higher levels of empathy towards working horses and more accurate perception of their pain [ 13 ]. It has been suggested that in decisions relating to welfare, owner empathy may play a bigger role than socio-economic status [ 36 ] therefore the fostering of empathy in owners has been recommended as a method for the improvement of equid welfare [ 13 ]. The development of empathy was also a large focus of school directed education programmes. NGOs varied in their level of school education offered; one charity had a dedicated education team specifically to develop humane education programmes delivered across their projects. Other NGOs described running after school welfare clubs where fun activities such as singing, art and drama conveyed equine welfare messages. It was felt that children who were aware of animal welfare needs would grow up to be more empathetic owners, taking better care of their animals. "One key approach that I think should be adopted is focusing on child education , you know , from the early years on animal welfare , because these are the future generation . " P26. At a young age it was felt that children were better able to take on board new perspectives than adults who may be more entrenched in their beliefs. "it is fundamental , working with children… you know , because children really , really can change their mindset . " P21. It was described that the reach of children's education programmes extended beyond just the children themselves, an effect also seen in the field of child-centred environmental education [ 37 ]. The information given was disseminated by children to their parents and friends and generated awareness of future NGO events and initiatives. "not only will they then grow up to be members of the community , but they're also going home and talking about animal welfare to their parents . And so also potentially talking about [our organisation] if [we are] coming or… other organisations that have been coming into the communities so that we get better uptake of community training initiatives that we run . " P31. Many of the children involved in schools programmes already have roles in taking care of their family's equid and so the information given can be of direct benefit. "their neighbours , their families own an animal , and then they can influence back their parents to , to at least provide the necessary needs for animals . And these kids , you know , they provide some water to the animals , they groom them . " P29. One difficulty of school education was being able to measure the impact that programmes are having. "we have quite a large school education programme… it would be really great to be able to demonstrate the benefits that aren't just anecdotal or theoretical . But yeah again , that's a really difficult thing" P31. The lack of direct evidence from monitoring and evaluation of school education initiatives was, for some, a discouraging factor for investing in child focussed programming. "I think there are some things that take absolutely ages to be of benefit like the education of kids in schools and stuff , that sounds like something that is really really useful to do but in terms of success…you've got to be watching that project for what 15 odd years before you get an idea really of how successful it is . " P2. Community participatory exercises A focus on the human drivers of welfare had seen an increase in the use of community participatory exercises such as participatory rural appraisal (PRA) by working equid NGOs. PRA uses a variety of participatory tools such as transect walks, scoring and ranking exercises, informal mapping and the creation of diagrams to enable local people to share their perspectives and knowledge [ 38 ]. PRA approaches are well established in epidemiological research, international development and environmental education and are now becoming more frequently used in animal welfare initiatives [ 39 – 42 ]. Exercises are often pictorially based in order to ensure that group members with low levels of literacy are not excluded from participating. Participants described participatory approaches as an effective way of engaging individuals from the very beginning of the initiative process. Exercises were especially useful for identifying community views on priority welfare issues. Upjohn et al. [ 34 ] highlighted that priority welfare issues for people in target communities are not necessarily the same as those identified by an NGO, and this was echoed by participants. "actually the owners are not prioritising the kinds of issues that are implicitly assumed to be priority issues within those interventions" P32. Understanding these differences in perception can avoid previously seen scenarios where a mismatch in priorities has led to the failure of an implemented initiative through lack of community engagement. "when we started training , we see the impact is very limited , because we didn't involve the communities" P16. It was described that approaches such as PRA have enabled a move away from previously used techniques where information was imparted to the target community (with purely an educational focus) but local people did not participate in the knowledge generation, share their insights or have any ownership over the learning process and outcomes generated [ 38 ]. Higher levels of engagement and investment in initiatives was one of the biggest advantages discussed regarding the use of participatory approaches. "they feel included and more invested" P21. Participatory exercises are designed to allow individuals to share their local knowledge, perspectives and opinions. "you manage to have them on board so you are not the teacher teaching them , but having them collaborating with you" P1. Utilising these opinions to come to a shared agreement regarding the direction of welfare focussed efforts enhanced investment in the subsequent initiative. "if someone comes up with an idea they always try harder to make it work because it is their idea and they take it forward more . " P2. Enabling communities to develop their own strategies was felt to empower communities and ensure that future actions were appropriate and tailored to the local context. "instead of us telling them what to do , we try to get them to facilitate each other in terms of recognising what are the issues with animal welfare and what they can do about it . So we are empowering them and the answers came from them . " P19. Another advantage of participatory exercises discussed was their ability to identify the good practices that communities already have established. Highlighting these practices not only reinforced positive welfare aspects within sessions but allowed these practices to be built on to facilitate further welfare improvement. " in that workshop , they got the communities to identify what they already did that was really good . And that they wanted to kind of keep… . And then they got the communities to identify what they then wanted to change " P27. The long-term impact of initiatives born though participatory techniques was also felt by participants to be better. "it creates greater and more sustainable behaviour change when you do it in partnership with people . " P7. This was also linked to the ability to use participatory exercises to identify and discuss the root causes of welfare problems. "I think that that's very important to find out , you know , what are the issues and not just the symptoms but actually delving down into what are the causes of something" P15. Being able to then address the root causes of welfare problems represented an avenue for sustainable welfare improvement. Skills initiatives These initiatives aim to train individuals in specialist (non-veterinary) skills such as farriery, saddlery and harnessing in order to create businesses providing those services to the community. Participants described: "education and skills initiative…this is a very crucial action in creating awareness and equipping the communities…with basic knowledge and skills" P11. Studies have reported as many as 78% of equids in a population affected by working equipment related injuries, and these skills initiatives can be especially useful in such areas [ 43 ]. "they taught the para-vets how to make basic harness , just pack saddles and very basic harness but that's had a massive effect . The knock-on effect from that was tremendous" P4. With huge variability seen in the standards of local service providers [ 14 ], creating a network of highly trained providers can ensure that both animals and owners are receiving high quality services, helping to maintain welfare within the community. "where we were based…pretty much all of the owners would use one of the farriers that we trained . I think that was so successful" P2. Most initiatives aim to train members of a community so that they can become full-time service providers within their community. "you want for them to work with you to be viable local businesses . " P21. However for this to happen, individuals need to have a financially feasible business which some participants described as difficult to establish. "People were really keen to learn how to make stuff but trying to then sell it was quite hard because it was obviously way more expensive than like old clothing or the rags people were using" P2. In order for initiatives to be successful, there needed to be a demand within the community for the service that was being offered. "just because you create a service doesn't necessarily mean that you fixed the problem" P32. It was felt that some previously run initiatives had been unsuccessful because the skills imparted by NGOs, and services subsequently offered, were not targeting priority issues for local people and resulted in a lack of investment. Differences in welfare priorities between communities and NGOs have been previously noted [ 34 ]. Due to this need for service demand, some organisations only focussed on upskilling existing service providers within communities. "we don't want to start from zero we just want to start where there are already service providers . " P26. In this scenario individuals are already established which removes the barrier of creating both service demand and local infrastructure. "there was already this setup to use that service so it wasn't like this big stepping stone it was just improving a service that was already there . " P2. Relatively small changes to the operations of these individuals can then have widespread effects within a community. Another challenge discussed was dropout of individuals enrolled in training courses. "you're taking 10 people… at the end of the year you are left with six people . So my concern is those four people why are they leaving ? " P22. As courses are intensive to teach, a substantial amount of resources and time are invested in candidates and participant dropout was a significant loss. Trained individuals subsequently leaving their communities was another area of concern discussed which was felt to hinder the success of initiatives looking to build networks of local service providers within an area. Participants highlighted the need for continued support for skills initiatives, particularly for ensuring the longevity of service provision. Previously used models for the delivery of skills initiatives saw UK based professionals teaching courses for short time periods in target countries. This was not felt to be adequate for the long-term maintenance of these skills within communities. "If you do it in a way that's shipping them in and then inject some knowledge and go away again , I have found that those are very unsuccessful although they have a lot of excitement at the time . " P5. Many NGOs have now shifted to in-country delivery of skills training programmes which allows courses to be taught by local professionals, tailored to the local context and provides increased scope for ongoing support. Barriers to welfare improvement Although participants discussed the strengths and weaknesses of the different initiatives used, similarities emerged in factors that impacted NGO programming regardless of the approach taken. These inherent challenges when trying to engage in welfare improvement initiatives cross-cut initiative categories and were widely described by participants working across countries. Equid status One such barrier, described across many institutions, was people's attitude towards equids, both at the community and policymaking level. This was especially apparent in the case of donkeys who, across countries, are frequently associated with low status. " even now in [this country] donkeys are not considered great… most of the people , not even the veterinarians want to go and treat the donkey , thinking it is a very neglected abused animal . " P8. It was felt that donkeys' association with menial or household work and comparatively lower sale price influenced opinion of their 'worth'; this has also been documented in other research studies [ 12 , 44 ]. "Donkeys are generally not taken care of well by the communities because they are cheaper and they are let loose for grazing" P28. This results in a systematic lack of basic care for donkeys, which is not reflected for other equine species. The perception within communities that donkeys are low maintenance and 'take care of themselves' was frequently mentioned. "malnutrition is one of the problems because as I said , donkeys their priority is so down some of them don't even feed them . Because donkeys normally… get whatever they get from roaming that is how they live" P11. Due to donkeys' low status, encouraging consideration of their welfare needs can be met with incredulity and sometimes derision. "apparently lots of them turned up just because they couldn't believe there was going to be a project about donkeys and they just came to find out , because that was the ultimate craziest thing [a project about] being nice to a donkey . " P5. The association of equids with low status can also extend to professions related to equids such as farriery. This perception and a lack of respect associated with the profession can prevent the recruitment and retention of individuals to farriery training programmes. "it's not a very respected profession . So there's not much pride in being a farrier . Often , the animal's feet are described as dirty , farriery is therefore described as a dirty job" P27. The cultural status of equids in some areas also affected their treatment by vets. "so getting them to engage with the idea of equids as a species you want to be a vet for is often challenging . " P32. The potential reasons for this were discussed and included: a lack of equid inclusion in veterinary curricula, wider societal views that do not prioritise equids and a lack of uptake of services as some owners are reluctant to pay for treatment for equids. Despite their importance to livelihoods and domestic tasks, the value of equids and the activities they perform are not always recognised by owners. "I would definitely hear owners saying , there's no point , you know , there's no point putting money into donkeys because they die anyway . " P32. Other livestock (those that provide a more direct benefit such as milk or meat) are often considered to be more important and as a consequence receive more care and veterinary attention in situations where financial resources are scarce [ 45 ]. This situation is also mirrored in the advocacy struggles encountered by NGO staff when trying to influence policy changes in favour of working equids. "the invisible equine within the development agenda is a problem and then animal welfare is also a bit of a blind spot . " P15. A lack of recognition of the value of equids by governments and policymakers both at local and national levels was described. "And when you actually look at the population statistics with FAO , some countries don't even register that they've actually got working equids . " P14. It was felt that research demonstrating the contribution that equids make could favourably influence their status both in the eyes of the public and policymakers. "we need to bring a change in the attitude of the general public of how good and innocent and hard-working an animal it is . How is their contribution to the economy" P8. There was understanding that other important societal issues will compete for government funds and legislative action but issues around working equids appeared to be low on policymakers' list of priorities. "In Africa you've got poverty , unemployment , HIV/AIDS . Huge , these are huge cross-cutting issues . They equally demand funding and skills" P15. The indirect benefits that working equids provide make it difficult to quantify their value to a country's economy [ 4 ]; and in countries where working equid populations are decreasing (as mechanisation becomes more widespread), there were concerns that their position is viewed as irrelevant to a countries' future growth. "the population of equids as per the government census has decreased significantly … so the government is not focusing on it…the equids are not a priority" P28. Working equids' exclusion from livestock policies was a barrier that was discussed across organisations and NGOs, and advocating for their inclusion was a focus of their programming. Successful advocacy resulted in the acknowledgement of equids as 'working livestock' contributing to food security by the UN Committee of Food Security in 2016 [ 46 ]. Recognising working equids as livestock covered within national livestock policy definitions can have a range of beneficial effects [ 47 ]. These include access to national animal health strategies and services such as vaccination campaigns, disease surveillance and emergency relief schemes [ 45 ]. One participant highlighted the need for equid inclusion in current livestock disease control strategies, particularly in the case of zoonotic diseases that pose a serious risk to human health such as anthrax. "it affects cattle , it affects sheep , goats , horse , donkey , human , and other animals too . The funny thing is the government actually do the vaccine against cattle , but they don't vaccinate against donkeys… they're actually opening a pocket , the problem is still there" P11. Leaving equids unvaccinated can create reservoirs of disease in populations that can then aid transmission to other animals or potentially humans. Access Accessibility was raised as a barrier to the success of welfare initiatives, both regarding access to communities and access to all individuals responsible for maintaining equid welfare. It was discussed that some equid owning communities live in very remote rural areas where physical access is limited by the environment, such as mountain communities. By virtue of their position, these are often the communities that rely heavily on working equids for many roles including the transport of people and goods across difficult to navigate terrain where the use of a vehicle is not possible [ 48 ]. But nevertheless, for NGOs attempting to reach these communities, access is a challenge. "I think it's more their location makes it quite difficult for even our teams that are even based near them to even be able to access them , to bring the stuff that they need to kind of do the work . And it's usually the worst welfare conditions are in the areas that are the most remote . " P17. The COVID-19 pandemic has further exacerbated issues surrounding community access, a finding that has also been documented across NGO sectors [ 49 ]. It was described that remote communities also have problems accessing veterinary services due to their distance. This was especially mentioned in the case of brick kilns which, due to pollution regulations, are situated away from urban areas. "even if they want to call veterinary services , people don't come because of the distance and the remoteness of the place . " P8. Access to the full spectrum of individuals involved in the day-to-day care of equids was also another challenge highlighted by professionals. This is especially pertinent to cultures where the gender division of labour can mean that the individual working with the animal is not necessarily the person who undertakes routine equid management or attends an animal clinic. " So for example in [country] you often have children bring the donkey to the clinic for treatment and I remember on a particular occasion a young girl had brought a donkey , it had abscesses in both front feet and the vet said 'I thought I told you , you weren't to work the donkey' , clearly by the foot pads the donkey had been worked in between the treatments and the girl's reply through translation was 'I know what you said but how do I tell my father ? '" P6. Many participants described the need for programmatic activities such as meetings, workshops and educational events to include all individuals responsible for equid care. However, this was sometimes challenging; cultural norms in some areas can prohibit mixed-gender community participation and different time constraints across genders can mean that certain individuals are not available to take part in NGO programmes. In these situations NGOs can struggle to include the views of, or impart knowledge to all of the people necessary to make improving welfare successful. A lack of equid trained vets A frequently highlighted barrier to the success of equid welfare projects was the absence of training related to equines in the curriculum of veterinarians across countries. This has resulted in local veterinary service providers who do not have the specialist knowledge or skills to treat equine disease or injury. As many communities do not have the option to seek treatment for their animals elsewhere, this lack of access to appropriate treatment further compounds the poor welfare state of working equids. "If they manage to get somebody there for the veterinary services , the person who deals with the case aren't trained or have the skills to deal with donkeys and mules . " P8. While production animals appear to be the focus of much of the veterinary curriculum, equids in some areas are not commonly encountered by vets. "5th year vet students and they've not touched a horse , they're scared of it ! " P13. In some areas, even if there is provision made for the veterinary treatment of horses, knowledge of donkey treatment is severely lacking. It is common that no differentiation is made between treatments for different equid species. "Another problem is that the majority of people think that donkeys are just a small horse and they have the same needs" P1. Across NGOs there was a programming focus on working in partnership with veterinary universities to bring elements of equine training to the curriculum. "So we've tried to do a 101 working equid curriculum . That might be ideally , teachers that will take this on board and make them their own and listen to our recordings . And we're building slowly , sort of a database that they can have . " P31. Programmes where vet students can gain experience in the handling and treatment of equids were also run by NGOs with the hope of ensuring a better coverage of equid trained vets returning to their regions. "when the vets come and work with us , they can come with us in the field as part of the internship programme , start touching the donkeys and injecting donkeys , in our presence so it builds their confidence . " P8. Equid trained para-vets can also help to fill this gap in provision. Monitoring and evaluation challenges Many participants described an increased focus on monitoring and evaluation (M&E) of initiatives. This focus parallels changes seen far earlier in the field of international development where the need for accountability in the use of governmental and institutional funding resulted in more rigorous reporting and evaluation systems [ 50 ]. M&E was not perceived negatively by participants, all participants were in favour of tracking the progress of initiatives and felt that M&E was important to know how much of an impact their projects were making. However comprehensive and structured M&E processes were described across virtually all organisations as a relatively new introduction and rigorous M&E across the different approaches now utilised by organisations was highlighted as a challenge. "I would say this is relatively new to the programme so…it is a little bit trial and error . And it's also quite new to our teams . So there's a lot of work in that" P13. Apprehension was expressed that the increased need for formal evaluation has put extra pressure and workload onto those collecting M&E data at ground level. "So there's a lot of paperwork which the field team hates ! " P23. It also carried the risk that the time spent evaluating projects would result in less time devoted to carrying out the activities that would create desired change, a focus on upwards accountability seen in other NGO sectors [ 29 ]. "suddenly we went from actually going out there helping donkeys to spending all our time trying to prove we were helping donkeys" P4. With the implementation of new frameworks came concern that the expectations were not compatible with the practicalities of situations on the ground. "it was a very sort of tricky conversation to say , you know , we are not sure if this is realistic" P32. Participants spoke about the need to use indicators that are 'meaningful'. Historically numbers of animals reached by NGO programmes were the most common indicator of success measured [ 5 ]. However, this did not capture the impact that initiatives had on those animals reached and whether this led to any type of long-term welfare benefit. "Instead of just saying we treated , I think it used to be like 360 , 000 donkeys a year at that time , as [a colleague] quite rightly said 'that could be one donkey 360 , 000 times because you treated it really badly and haven't done anything for it ! ' What does it mean ? It doesn't mean anything" P4. It was also raised that what is deemed a meaningful indicator may differ by area. "So we also have the balance between indicators that Western developed countries think of as indicators versus what are really useful indicators here in more developing countries . " P19. With the focus shifting from numbers of animals reached to longer-term success, interviewees expressed difficulties in finding the most effective indicators to measure complex outcomes that can be influenced by multiple factors. This was especially true of multidisciplinary initiatives which seek to improve equid welfare indirectly through livelihood enhancement. "I think the biggest challenge with that is that it is successful , but it's been hard to measure how… how can you then know that that's directly improving equine welfare , if you're improving someone's livelihood" P20. It was also a concern that external factors could lead to an unrepresentative negative evaluation. "actually measuring real impact , particularly when…in the middle of your three years , there's a massive drought and people lose their livestock and body condition goes right down . You've been doing these beautiful welfare assessments and then suddenly… there's so much else that how do you then have all this data and have the thousand caveats that come with it ? " P13. However one participant suggested that the perceived difficulties surrounding comprehensive M&E can be unfounded. " one of the myths is that M&E is expensive and complicated and it certainly doesn't have to be . You can do it at a lot of different levels and I think that often the baby is thrown out with the bath water because people think you have to do so much when really you can still gain a lot of information from a smaller but dedicated and strategic approach ." P5. It was suggested that smaller changes, more reflective of the incremental nature of changes made by welfare initiatives could be a more suitable avenue than trying to immediately link initiatives to direct welfare improvement that may take far longer to be realised. "I think there is not enough emphasis placed on smaller indicators and looking at success in other ways… . So for example provision of water could be a small one , you could say 100% of people weren't providing water all the time that their animals were stabled and now 80% are , and although that might not lead all the way through to animal-based indicators of improvements in dehydration say , you've got those human behaviour indicators of what they're doing" P5. Advocacy It was apparent that there has been a relatively recent shift placing advocacy as a strong focus across virtually all organisations. This is not unique to the working equid sector; the growth of advocacy across both generalist and specialist UK based NGOs has been documented [ 18 ]. A larger focus on advocacy was perceived as having both benefits and disadvantages. Some participants felt that advocating for policy change was the best way to enhance the lives of large numbers of animals, many more than could be reached through more practical, on the ground initiatives. "I actually think targeting governments , and getting them to implement the [WOAH] chapters and getting them to enforce good welfare is probably your best way of making a huge difference , and having a huge reach . " P14. It was recognised that advocacy can help change systems in a strategic way that aims to address root causes of poor welfare, rather than addressing the symptoms of individual animals. "going for the right level would get better spread , better results , I think , in the end but it's more of a long-term plan…firefighting is never going to get you a result at the end of the day . It's not a strategic way of thinking . " P14. Recent advocacy efforts have seen equid welfare NGOs working together to form groups such as the International Coalition for Working Equids to implement the World Organisation for Animal Health (WOAH) Welfare of Working Equids Terrestrial Code Chapter 7.12 [ 19 ]. Other groups such as the Equid Power Network [ 20 ] have been involved in promoting educational events for the public and professionals across disciplines. It was acknowledged that for certain large-scale international problems (such as the trade in donkey skins [ 21 ]), correspondingly large-scale advocacy was likely to be the only effective solution. However, it was felt by some participants that the current prominence of advocacy on NGO agendas came at the expense of other approaches. "I think , particularly nowadays , every charity organisation is actually advocating advocacy… It seems as if the practical work on the ground has been neglected" P11. It was also expressed that advocacy could be essentially risky in terms of its rewards compared to the effort put in to achieve changes. Dependent upon the political climate, policy changes can be difficult to achieve and the process was described as being both lengthy and often frustrating for NGO staff. "the least successful approach…(is) the engagement with the policymakers because these policymakers they are bureaucrats and the government systems are very slow" P28. A second highlighted issue was that even if an advocacy strategy was successful in creating a policy change, this was no guarantee that the effect of the change would be felt by animals and communities. "So that [law] existed when I first started working with the project and I don't think I ever saw it enforced and I was there for 2 and a half years . So like although it was written down it didn't make any difference on the ground as such . That's just one of those things isn't it ? " P2. A gap due to lack of enforcement of welfare legislation was often discussed as an obstacle to achieving welfare change. "There is some policy on document , this policy has not been implemented , you know , these are really challenges for us . " P25. A number of reasons were suggested for failure to enforce regulations including a lack of staff capacity, lack of resources at the disposal of the enforcing body and a lack of education of enforcers. "policeman for example . Suddenly he is told that if he sees someone beating a donkey or carrying more than 300 kilos weight that man is… breaking the law . So he sees somebody carrying 200 kilos in the cart with 2 flat tyres , a bad harness and you know dogs biting the donkeys and everything else . And he's not broken the rules so he lets him go , that's the first problem . That the people that are enforcing the rules don't understand what they are enforcing…without that knowledge they can't make a judgement . " P4. These factors are often systematic within enforcement entities (most commonly police or government enforcement bodies) which presents another challenge for NGOs who can have little control over how the policy is eventually enforced. This led to discussions that some political infrastructures are less likely to successfully support effective policy changes. "lobbying and influencing future policy regarding equids can have greater and lasting impacts on welfare…mainly when the country is … resourceful enough to enforce it . " P29. Although acknowledging that it was resource intensive, some participants described an ideal balance needed of top down (advocacy based) and bottom-up (practical field based) programming. They felt that working from both directions was the most effective in terms of advocacy enabling higher up changes that supported an environment where good welfare could be created and sustained through on the ground programming. "in an ideal world , you need to be doing advocacy , potentially , on a regional and international level . And then on all the subsequent levels down , right to the community . But you also need to be doing that community level work , and all the way up and having the combination of the two . " P27. Continuing work on the ground alongside advocating enables immediate animal welfare needs to be addressed whilst informing advocacy efforts through having a thorough understanding of community level needs. "I think it's great that there's a lot of organisations , I think , now moving to this more advocacy work , but I think , unless you know , the situation on the ground , I think it's very difficult to advocate for those people and those horses , because you don't know the context . " P13. Research Organisations differed in the level of research that they carried out. Some considered research a priority and employed dedicated research teams to support their projects. "I would focus on truly understanding the situation before you try to intervene" P5. The investment of NGOs in research can be seen across sectors including animal welfare, international development and the environment, where research outputs are used to guide programming development [ 22 – 24 ]. The need for a thorough understanding of the community context, priorities and issues were underscored when planning a programme, especially if entering a new community. "I really like how we got some decent research at the beginning of our project… we had an understanding of those places and that really helped to then design a better project from the word go . " P2. Research was described as important both in informing the direction of programming by gathering contextual information about target communities and later ensuring that the approaches utilised were creating the desired changes. Although participants acknowledged that research could be a useful tool, some voiced feasibility concerns over the level of time and resources that research requires. "I think it can be , it can be quite time consuming . And it can be quite expensive , as we all know , with research . But you know that investment is really worthwhile for that long term and sort of wider impact . " P13. Size of organisation was also a factor that affected research engagement, with small national NGOs having limited ability to carry out their own research projects. "Yeah , we're not really big research people . I think it's more capacity and time" P23. Where research was being used as a tool to identify prevalent welfare problems in an area, some participants expressed frustration. "The issues for most of the working equids are pretty simple , so they are overworked , underfed and beaten . It doesn't take a lot of research to see that . " P4. This mainly centred on the time taken to conduct research on a subject that participants either felt was obvious to identify or could feasibly start to be addressed before research results were formalised. " you know from the people or from what you see on the ground from the animals , what exactly the problem is . I think you have to start while … you are prioritising , while you are actually identifying the problems" P11. Another issue identified by participants was not to do with the research process itself but the way in which research outputs were able to be used. "actually I think that what is missing in research is a link between the university or the academic setting and the farm setting , the real life" P1. Some felt that despite research being carried out, the findings were not routinely put into practice. "I think the charity organisations they do one thing very nicely but that stops there when the research towards papers has been published , it goes there on the shelf . " P11. Planning for utilisation of research was described as important and some organisations only carried out research that they felt would be of direct practical value within their projects. Many participants described gaps where they felt further research was needed and across organisations three common topics were identified. Firstly, there was a call for research that provided further evidence on the socioeconomic contribution of equids to livelihoods. This was particularly mentioned as important for advocating for equids at a policy level. "the socioeconomic value of the donkey and the donkey-owning people , what is the contribution of these people to the GDP of the country because it goes unnoticed . Of course imagine if the brick stops ? Imagine all the houses , real-estate would collapse but people don't know the bricks are made on the donkey back" P8. Within the topic of livelihoods, it was also felt that evidence directly linking improvements in human livelihoods to animal welfare was needed. "we have been promoting the livelihood of the communities , but we don't have any evidence , scientific evidence to support that whether the increase in the livelihood improves the animal welfare or vice versa" P28. Secondly, further information was required on infectious diseases of equids, a topic that has major welfare implications but was not felt to receive enough research attention. "so many nasty diseases actually really hampering the health and welfare of equids in general in most developing countries" P11. Within the topic of infectious diseases, zoonoses were included, an area likely to receive increased research attention in the wake of the COVID-19 pandemic. The final topic that was raised by multiple participants was the need for information or guidelines for the amount of weight animals should be carrying or pulling. "how much weight can a donkey carry , that's a question that comes up a lot . " P9. Veterinary clinics The provision of free (or subsidised) veterinary care was the approach that generated the largest amount of discussion. "I think it is one of those things that divides opinions the most I'm going to say running free vet clinics . " P2. Whilst historically the most commonly used approach, a shift away from the provision of free veterinary clinics has been seen in recent years across many NGOs. A spectrum of opinions was expressed with large differences seen between the philosophies of individual organisations. The main concern regarding the provision of free veterinary clinics was that of sustainability. Criticisms both relating to the amount of resources necessary to run clinics and questions over the creation of lasting welfare change were expressed, mirroring debates in the literature [ 8 , 25 ]. "[our organisation] cannot make sustainable scale level impact by veterinary service provision directly . There are too many donkeys and not enough vets and we wouldn't be able to afford all the vets even if there were enough . " P7. Creation of dependency on the service was a barrier to independent long-term welfare changes when NGO services are likely to, at some point, be withdrawn, enabling organisations to move to another location [ 7 ]. "I think that's always the difficulty when you go in , provide that free stuff , or subsidises , it's much then harder to not provide it because it becomes an expectation , and I think we have to be aware of that" P13. It was also expressed that free clinics, as an intensive approach, reached a limited number of animals in comparison to those able to be reached by other types of initiative. "for years and years we may have worked in an area where the 200 donkeys in this immediate area are greatly benefitting from our weekly clinic but the other 2 million donkeys in this country never see us so what can we do to actually reach all these animals ? " P10. Although free clinics are one of the most immediate ways to relieve animal suffering, participants highlighted that they address the symptoms rather than the root causes of welfare problems leading to lack of long-term change. "if you're not addressing the root causes , then you're not really going to be achieving , you know essentially being a band aid kind of project . " P15. This view supports previous research advocating for a move to more prevention focussed programming [ 3 ]. Having a free service available to address animal health problems also led to questions over owner motivation to prevent those problems occurring. "the feeling of some of the staff were that the same horses just kept coming back … they felt that they were very much treating symptoms rather than addressing the causes . " P5. The lack of cost associated with veterinary treatment provided by NGOs had, in some instances, led to owner delays in treatment seeking for serious welfare issues. This potential to cause harm to the animals that were meant to benefit from treatment provision was described by one participant as a significant factor in the organisation's decision to stop providing a free clinic treatment model. "with the mobile clinics , sometimes , because people would know that it would be free , people would wait a month to seek treatment because they'd be like , well , the [NGO] van will be here in a month , and then it'll be free . So they had evidence , not just that it wasn't being successful , but that potentially , there was actually harm . " P27. There were also concerns that an unintended consequence of providing free treatment was a negative impact on local veterinary professionals through loss of business, a concern echoed in other publications [ 4 ]. "I was in a brick kiln in [country] working with one of the private para-vets…and I asked him what his biggest challenges were in his work . And he said , 'it was when you guys used to turn up and give free treatments , and nobody would pay for any of my treatment . " P27. Parallels can also be drawn with the human primary healthcare sector where NGO healthcare provision has been criticised for undermining and fragmenting local health systems [ 26 ]. However, in areas where there is no established veterinary infrastructure, NGO clinics may provide the only treatment that animals are able to access. "just recognising that there are welfare needs , and how do you address them right now , if you don't have provision of free care ? " P31. Whilst concerns regarding sustainability existed, addressing poor health and injury is the quickest way to alleviate suffering and all participants acknowledged that veterinary treatment was likely to make a large positive impact on the quality of life of the animals receiving treatment. "who can deny that veterinary clinics are good for the animals that get to see them ! " P5. In areas where health problems were severe, it was felt that efforts to address other areas of welfare ahead of or instead of veterinary treatment would achieve limited success. "saying let's sit down and talk about aspects of welfare is great but you need to reach a certain level of veterinary support before that tends to happen in my opinion . " P6. Some organisations used veterinary clinics as an initial approach when entering a new community in order to establish a positive relationship and demonstrate the professionalism and trustworthiness of the organisation. "providing those services to the communities is , is a good introduction method for your ability . So I think that's a very important aspect to include in a community . " P21. This was especially pertinent in areas where equids are not considered a priority species for receiving treatment. "So free treatments is always used as a gateway to create a win and show them that actually treating the animals improves the body condition , improves the life expectancy , improves efficiency and work capacity" P11. Employing a free clinic approach to treat the worst welfare problems in these situations can be used as an avenue to engage with owners about other aspects of welfare and lead on to discussions about preventative care. "So once we treat and we've made the donkey's life better , the livelihood is saved and the community engagement gets even better . They get really engaged with us because they know the value 'these people came here all to help our livelihoods , our donkeys so I think we should listen to them about the better harness to prevent wounds , I should listen to them about better hoof care'" P8. Veterinary clinics can also support other domains such as training and research. Clinics provide an opportunity for vet students to learn skills in recognising and treating common equine health problems and positive handling methods. This was highlighted as being especially important in countries where equines are absent from the veterinary curriculum. Clinics also offer the opportunity for research to be conducted which can form an important part of disease surveillance and prevention. "the other important thing in veterinary clinics is actually the charities they do support the student research schemes to do surveillance programmes rather than doing it by themselves in most of the cases in collaboration with the government . " P11. Acknowledging the balance of positive and negative aspects surrounding the provision of veterinary care, some participants suggested that careful and strategic planning and clear communication can minimise some concerns. "that's something that you have to be really careful when working with community members and people , unfounded expectations . And you have to be very clear , and therefore you need to have a strategy saying that , okay , we're going to be doing this for this amount of time" P21. Others preferred to use clinics only under certain circumstances or in combination with other approaches. "I don't think that long-term that is purely the best way forward . In the right time and the right place . " P2. There was also a distinction made between short-term or one-off clinics run by vets flying in from overseas and clinics that build local capacity. "it would depend how it was provided… if you're building skills in somebody , in a locally trained vet , then that is slightly different . Obviously , then , hopefully , you're creating a more sustainable , long-term solution and that you've got locally skilled people who understand local context and can provide a locally appropriate service . " P32. Para-vets (community animal health workers) The role of para-vets (sometimes called community animal health workers (CAHWs) or community based welfare advisors) was felt by many participants to be important in contributing to the sustainability of local scale health and welfare provision. "about 1000 para-vets a year I think were being trained there and they would go back to their communities and they would take all of their knowledge with them" P4. Accessibility of para-vet services was highlighted as a big advantage to communities. As community-based practitioners, para-vets are available locally and at short notice. "they are… from within the communities . So , they can help the communities in need especially in the night hours when all the hospitals and medical schools are closed . " P28. Whereas particularly in rural and remote areas, accessing more formal veterinary services at clinics or health posts can mean a long journey. "And if a community is lucky , he has to travel with the sick animal at times , over two kilometres minimum , it is very difficult . " P29. It was also mentioned that nomadic communities who are reliant on animal breeding for their livelihood, but often fall outside of the reach of veterinary services, can benefit from para-vet training. The affordability of para-vet services was felt to be an important factor in the success of para-vet programmes. "if you can make some basic first aids available at the place where they are , that is something that could be affordable . " P8. The flexibility of payment that para-vets can offer was preferred by some communities as it was more practically feasible for their livelihoods. "They can pay anytime , the government service provider has to accept the money now , you know , they don't have cash economy . Farmers can get money when they sell some of their produce during market days . They may not have money essentially . But the CAHWs can treat them readily and accepts the money next month or next week . " P29. Although limited in scope in comparison to a veterinary professional, para-vets can help reduce the pressure on veterinary services by treating more minor health issues. " They took the pressure off the team because instead of having to go out , keep getting calls from you know 'my donkey's got a cut' that could be treated by the CAHW . If the donkey had something more serious they would still go to the CAHW , he would assess it and then say 'yeah I can do that' or say 'this is beyond me' and he would then call the team" P4. Para-vets can also then act as bridge between communities and veterinarians or other equine service providers to promote preventative healthcare and positive welfare. "So apart from this , we link them with the existing services that we are supporting , like vaccination" P29. In a human health parallel, there has been a rekindling of interest in the use of community health workers to help meet child survival goals, particularly in areas of the world where there is low coverage of primary healthcare professionals [ 27 ]. Despite the positive aspects that para-vets can bring, some participants mentioned concerns about regulation of para-vet services, this especially applied to areas where para-vets had greater scope to provide treatments such as administration of drugs by injection. "some states are giving them three months of training and some states are giving them two years training , three years training . So it's very , it is a huge variation among different states . " P28. Ethical concerns were also voiced that some individuals profit from a business opportunity and as a consequence the quality of care given can be questionable. "They end up in under dosing the animals , and not giving proper doses…even though it is so crucial , it has to be strictly followed that actually they properly operate in particular areas . " P11. Incorrect dosing of animals, leading to potential increases in multidrug resistance, has also been documented in drug retail outlet workers [ 28 ]. Regulation of para-vet training, monitoring of service quality and inclusion of veterinary ethics in training were suggested as avenues to ensure that service providers were having the desired positive impact on welfare. General welfare messaging Although discussed less frequently than some of the other approaches, general welfare messaging had been utilised more frequently recently due to the COVID-19 pandemic impacting NGO's ability to provide a physical presence in communities. "So we've been trying to send messages to them , educational messages…making sure that they understand that even if we're not there physically , that we're there for them as support . " P24. Amid concerns that welfare may regress during this time, participants described welfare messaging via mobile phones as a way to keep in touch with project participants. "we send out bulk SMSs . So if we need to let them know anything important , or we want to give any tips…like with COVID , we kept on sending information about how to keep themselves safe" P23. It was described that general welfare messages were a good way to reach a large amount of people at a time in a format that was easily accessible and were used in particular to cascade preventative care messages to communities. "We're able to reach about 70 to 80% of our target group , with this approach , we also understand that we can cover also communities that are out of our intervention area . " P30. However one disadvantage was that there was little avenue for following up what the impact of the information was. "If you had done a radio announcement or broadcast…you had reached like maybe 1000 people but you had no real clear idea of what effect you had on those people" P4. In the humanitarian sector issues have been raised regarding the utility and function of general messaging such as programme signage, with concerns that increasing individual NGO visibility in order to account to donors may be prioritised over messaging efficacy [ 29 ]. Monitoring and evaluation of this type of indirect method is needed to ensure that the desired impact is being created. Microcredit The formation of microcredit groups was one of the least utilised approaches discussed despite being successfully and widely used across other NGO sectors [ 30 , 31 ]. Some participants expressed their reservations at the prospect of being responsible for others' finances. "We've done a bit looking at the microcredit but I mean , it's such a minefield for me" P13. Some felt that such approaches would not be suitable for the cultural context or communities that they worked in. "we don't get involved in their financial… it's too problematic for us to try and have like saving schemes and hold people's money" P23. Success was suggested to be contingent on a cohesive community and was described as more likely to be successful in certain countries in which community savings schemes such as Iddir in Ethiopia are common practice [ 32 ]. Those who had undertaken a microcredit approach described the benefits that could be realised when a group of people worked together to invest in improving welfare. Collectively groups could afford items that they never would have been able to purchase alone. "we had lead people in each little community group we were working with , it actually came from their idea… so that was great for trust reasons to let them lead it . And when they had enough money then , they didn't wait until everyone had enough , just each month when they grouped together they could buy a bit of a harness so they decided who would get the harness that month and they all kept going until everyone got one . " P2. Despite the potential that microcredit schemes offer, one participant observed a reluctance among donors to offer credit related types of programme. "A lot of donors are really skittish about that . Because , you know , giving people money – what is that ?! We are so afraid of giving people money because we think they are going to be , that they are going to misuse them" P21. However, it was pointed out that without access to funding, (such as that provided by seed fund schemes) equine service providers were unlikely to be able to establish themselves as viable, self-sufficient businesses. "you know , people tend to think 'but we are giving them training , that should be enough for them to , you know , create a large business and work' ! No… I think , the most important aspects for creating businesses are , you know , the accessibility of funding . " P21. Equine saving and credit groups, in certain contexts, had far wider societal implications. "two things which is good . One is the equid owners , the equid family have a direct access on that front . The other thing is that it's decreased dependency on the local moneylenders" P25. It was described that before the existence of the NGO saving and credit group, borrowing of money for emergency expenses was only available from high interest rate local moneylenders. When the equid owners were unable to pay, they entered a cycle of bonded labour; the role of moneylending in the persistence of the poverty cycle has also been highlighted in other NGO sectors [ 33 ]. The avoidance of bonded labour was mentioned as the primary reason to consider starting a microcredit scheme by an NGO that worked mainly with owners in the brick kiln industry. "So the brick kiln owner exploits the people who want money for emergencies , it might be a child's medical emergency or a marriage or a death in the family… the small little group can help them in crisis . That is something we should also try I think . I have heard about and there's a possibility we should bring it into the programme , I feel strongly" P8. Education Across organisations, a common theme in successful educational approaches was interactive, practical learning. Whether it was in the context of training vet students, government officials, owners at veterinary clinics or staff at behavioural workshops, it was felt that if learners could carry out practical activities themselves, the educational message would be better remembered and overall learning would be more effective. "personally I would always go for practical training , I don't like chalk and talk in a classroom , it is all about getting out and seeing it on the ground" P3. It has been documented that practical demonstrations of novel management techniques can accomplish recognition of the technique's benefits within the owner population [ 34 ]. Hands on contact and demonstrations with animals were seen as being particularly memorable. "the one that was most successful , I think was where we were able to do a practical session and go out and actually handle working equids and show them . " P14. This also applied to training in new techniques for members of NGO staff. "there have been animals I guess that have changed the way a whole team might think . " P6. Practical participation when introducing owners to basic management practices such as wound cleaning can be especially important as individuals need to be able to repeat the techniques unsupervised outside of a clinical setting. "And when you do the veterinary care , okay , that's when I started introducing the method of , I do it for you halfway , and I tell you what to do , and I sit there and watch you , and you do it yourself . " P22. Practical participation was also mentioned as being useful as low levels of literacy in some equid owning communities mean that other commonly used methods of communication such as written welfare messaging are unsuitable. One participant also described a higher level of community engagement for practical tasks. "not just words , because they're very much you know , tired , bored by the extension messages coming from government and talks . " P29. Interactive exercises with owners were also discussed as a method of encouraging owners to be able to visualise things from the perspective of their equid. "we prepare a sack full of sand… and then make to carry the owners of the donkey and then feel how heavy . See look . This is what you're actually loading your donkey . Can you feel how heavy it is ? " P11. Empathy has been directly implicated as an influence on animal welfare [ 35 ] and studies have demonstrated relationships between higher levels of empathy towards working horses and more accurate perception of their pain [ 13 ]. It has been suggested that in decisions relating to welfare, owner empathy may play a bigger role than socio-economic status [ 36 ] therefore the fostering of empathy in owners has been recommended as a method for the improvement of equid welfare [ 13 ]. The development of empathy was also a large focus of school directed education programmes. NGOs varied in their level of school education offered; one charity had a dedicated education team specifically to develop humane education programmes delivered across their projects. Other NGOs described running after school welfare clubs where fun activities such as singing, art and drama conveyed equine welfare messages. It was felt that children who were aware of animal welfare needs would grow up to be more empathetic owners, taking better care of their animals. "One key approach that I think should be adopted is focusing on child education , you know , from the early years on animal welfare , because these are the future generation . " P26. At a young age it was felt that children were better able to take on board new perspectives than adults who may be more entrenched in their beliefs. "it is fundamental , working with children… you know , because children really , really can change their mindset . " P21. It was described that the reach of children's education programmes extended beyond just the children themselves, an effect also seen in the field of child-centred environmental education [ 37 ]. The information given was disseminated by children to their parents and friends and generated awareness of future NGO events and initiatives. "not only will they then grow up to be members of the community , but they're also going home and talking about animal welfare to their parents . And so also potentially talking about [our organisation] if [we are] coming or… other organisations that have been coming into the communities so that we get better uptake of community training initiatives that we run . " P31. Many of the children involved in schools programmes already have roles in taking care of their family's equid and so the information given can be of direct benefit. "their neighbours , their families own an animal , and then they can influence back their parents to , to at least provide the necessary needs for animals . And these kids , you know , they provide some water to the animals , they groom them . " P29. One difficulty of school education was being able to measure the impact that programmes are having. "we have quite a large school education programme… it would be really great to be able to demonstrate the benefits that aren't just anecdotal or theoretical . But yeah again , that's a really difficult thing" P31. The lack of direct evidence from monitoring and evaluation of school education initiatives was, for some, a discouraging factor for investing in child focussed programming. "I think there are some things that take absolutely ages to be of benefit like the education of kids in schools and stuff , that sounds like something that is really really useful to do but in terms of success…you've got to be watching that project for what 15 odd years before you get an idea really of how successful it is . " P2. Community participatory exercises A focus on the human drivers of welfare had seen an increase in the use of community participatory exercises such as participatory rural appraisal (PRA) by working equid NGOs. PRA uses a variety of participatory tools such as transect walks, scoring and ranking exercises, informal mapping and the creation of diagrams to enable local people to share their perspectives and knowledge [ 38 ]. PRA approaches are well established in epidemiological research, international development and environmental education and are now becoming more frequently used in animal welfare initiatives [ 39 – 42 ]. Exercises are often pictorially based in order to ensure that group members with low levels of literacy are not excluded from participating. Participants described participatory approaches as an effective way of engaging individuals from the very beginning of the initiative process. Exercises were especially useful for identifying community views on priority welfare issues. Upjohn et al. [ 34 ] highlighted that priority welfare issues for people in target communities are not necessarily the same as those identified by an NGO, and this was echoed by participants. "actually the owners are not prioritising the kinds of issues that are implicitly assumed to be priority issues within those interventions" P32. Understanding these differences in perception can avoid previously seen scenarios where a mismatch in priorities has led to the failure of an implemented initiative through lack of community engagement. "when we started training , we see the impact is very limited , because we didn't involve the communities" P16. It was described that approaches such as PRA have enabled a move away from previously used techniques where information was imparted to the target community (with purely an educational focus) but local people did not participate in the knowledge generation, share their insights or have any ownership over the learning process and outcomes generated [ 38 ]. Higher levels of engagement and investment in initiatives was one of the biggest advantages discussed regarding the use of participatory approaches. "they feel included and more invested" P21. Participatory exercises are designed to allow individuals to share their local knowledge, perspectives and opinions. "you manage to have them on board so you are not the teacher teaching them , but having them collaborating with you" P1. Utilising these opinions to come to a shared agreement regarding the direction of welfare focussed efforts enhanced investment in the subsequent initiative. "if someone comes up with an idea they always try harder to make it work because it is their idea and they take it forward more . " P2. Enabling communities to develop their own strategies was felt to empower communities and ensure that future actions were appropriate and tailored to the local context. "instead of us telling them what to do , we try to get them to facilitate each other in terms of recognising what are the issues with animal welfare and what they can do about it . So we are empowering them and the answers came from them . " P19. Another advantage of participatory exercises discussed was their ability to identify the good practices that communities already have established. Highlighting these practices not only reinforced positive welfare aspects within sessions but allowed these practices to be built on to facilitate further welfare improvement. " in that workshop , they got the communities to identify what they already did that was really good . And that they wanted to kind of keep… . And then they got the communities to identify what they then wanted to change " P27. The long-term impact of initiatives born though participatory techniques was also felt by participants to be better. "it creates greater and more sustainable behaviour change when you do it in partnership with people . " P7. This was also linked to the ability to use participatory exercises to identify and discuss the root causes of welfare problems. "I think that that's very important to find out , you know , what are the issues and not just the symptoms but actually delving down into what are the causes of something" P15. Being able to then address the root causes of welfare problems represented an avenue for sustainable welfare improvement. Skills initiatives These initiatives aim to train individuals in specialist (non-veterinary) skills such as farriery, saddlery and harnessing in order to create businesses providing those services to the community. Participants described: "education and skills initiative…this is a very crucial action in creating awareness and equipping the communities…with basic knowledge and skills" P11. Studies have reported as many as 78% of equids in a population affected by working equipment related injuries, and these skills initiatives can be especially useful in such areas [ 43 ]. "they taught the para-vets how to make basic harness , just pack saddles and very basic harness but that's had a massive effect . The knock-on effect from that was tremendous" P4. With huge variability seen in the standards of local service providers [ 14 ], creating a network of highly trained providers can ensure that both animals and owners are receiving high quality services, helping to maintain welfare within the community. "where we were based…pretty much all of the owners would use one of the farriers that we trained . I think that was so successful" P2. Most initiatives aim to train members of a community so that they can become full-time service providers within their community. "you want for them to work with you to be viable local businesses . " P21. However for this to happen, individuals need to have a financially feasible business which some participants described as difficult to establish. "People were really keen to learn how to make stuff but trying to then sell it was quite hard because it was obviously way more expensive than like old clothing or the rags people were using" P2. In order for initiatives to be successful, there needed to be a demand within the community for the service that was being offered. "just because you create a service doesn't necessarily mean that you fixed the problem" P32. It was felt that some previously run initiatives had been unsuccessful because the skills imparted by NGOs, and services subsequently offered, were not targeting priority issues for local people and resulted in a lack of investment. Differences in welfare priorities between communities and NGOs have been previously noted [ 34 ]. Due to this need for service demand, some organisations only focussed on upskilling existing service providers within communities. "we don't want to start from zero we just want to start where there are already service providers . " P26. In this scenario individuals are already established which removes the barrier of creating both service demand and local infrastructure. "there was already this setup to use that service so it wasn't like this big stepping stone it was just improving a service that was already there . " P2. Relatively small changes to the operations of these individuals can then have widespread effects within a community. Another challenge discussed was dropout of individuals enrolled in training courses. "you're taking 10 people… at the end of the year you are left with six people . So my concern is those four people why are they leaving ? " P22. As courses are intensive to teach, a substantial amount of resources and time are invested in candidates and participant dropout was a significant loss. Trained individuals subsequently leaving their communities was another area of concern discussed which was felt to hinder the success of initiatives looking to build networks of local service providers within an area. Participants highlighted the need for continued support for skills initiatives, particularly for ensuring the longevity of service provision. Previously used models for the delivery of skills initiatives saw UK based professionals teaching courses for short time periods in target countries. This was not felt to be adequate for the long-term maintenance of these skills within communities. "If you do it in a way that's shipping them in and then inject some knowledge and go away again , I have found that those are very unsuccessful although they have a lot of excitement at the time . " P5. Many NGOs have now shifted to in-country delivery of skills training programmes which allows courses to be taught by local professionals, tailored to the local context and provides increased scope for ongoing support. Barriers to welfare improvement Although participants discussed the strengths and weaknesses of the different initiatives used, similarities emerged in factors that impacted NGO programming regardless of the approach taken. These inherent challenges when trying to engage in welfare improvement initiatives cross-cut initiative categories and were widely described by participants working across countries. Equid status One such barrier, described across many institutions, was people's attitude towards equids, both at the community and policymaking level. This was especially apparent in the case of donkeys who, across countries, are frequently associated with low status. " even now in [this country] donkeys are not considered great… most of the people , not even the veterinarians want to go and treat the donkey , thinking it is a very neglected abused animal . " P8. It was felt that donkeys' association with menial or household work and comparatively lower sale price influenced opinion of their 'worth'; this has also been documented in other research studies [ 12 , 44 ]. "Donkeys are generally not taken care of well by the communities because they are cheaper and they are let loose for grazing" P28. This results in a systematic lack of basic care for donkeys, which is not reflected for other equine species. The perception within communities that donkeys are low maintenance and 'take care of themselves' was frequently mentioned. "malnutrition is one of the problems because as I said , donkeys their priority is so down some of them don't even feed them . Because donkeys normally… get whatever they get from roaming that is how they live" P11. Due to donkeys' low status, encouraging consideration of their welfare needs can be met with incredulity and sometimes derision. "apparently lots of them turned up just because they couldn't believe there was going to be a project about donkeys and they just came to find out , because that was the ultimate craziest thing [a project about] being nice to a donkey . " P5. The association of equids with low status can also extend to professions related to equids such as farriery. This perception and a lack of respect associated with the profession can prevent the recruitment and retention of individuals to farriery training programmes. "it's not a very respected profession . So there's not much pride in being a farrier . Often , the animal's feet are described as dirty , farriery is therefore described as a dirty job" P27. The cultural status of equids in some areas also affected their treatment by vets. "so getting them to engage with the idea of equids as a species you want to be a vet for is often challenging . " P32. The potential reasons for this were discussed and included: a lack of equid inclusion in veterinary curricula, wider societal views that do not prioritise equids and a lack of uptake of services as some owners are reluctant to pay for treatment for equids. Despite their importance to livelihoods and domestic tasks, the value of equids and the activities they perform are not always recognised by owners. "I would definitely hear owners saying , there's no point , you know , there's no point putting money into donkeys because they die anyway . " P32. Other livestock (those that provide a more direct benefit such as milk or meat) are often considered to be more important and as a consequence receive more care and veterinary attention in situations where financial resources are scarce [ 45 ]. This situation is also mirrored in the advocacy struggles encountered by NGO staff when trying to influence policy changes in favour of working equids. "the invisible equine within the development agenda is a problem and then animal welfare is also a bit of a blind spot . " P15. A lack of recognition of the value of equids by governments and policymakers both at local and national levels was described. "And when you actually look at the population statistics with FAO , some countries don't even register that they've actually got working equids . " P14. It was felt that research demonstrating the contribution that equids make could favourably influence their status both in the eyes of the public and policymakers. "we need to bring a change in the attitude of the general public of how good and innocent and hard-working an animal it is . How is their contribution to the economy" P8. There was understanding that other important societal issues will compete for government funds and legislative action but issues around working equids appeared to be low on policymakers' list of priorities. "In Africa you've got poverty , unemployment , HIV/AIDS . Huge , these are huge cross-cutting issues . They equally demand funding and skills" P15. The indirect benefits that working equids provide make it difficult to quantify their value to a country's economy [ 4 ]; and in countries where working equid populations are decreasing (as mechanisation becomes more widespread), there were concerns that their position is viewed as irrelevant to a countries' future growth. "the population of equids as per the government census has decreased significantly … so the government is not focusing on it…the equids are not a priority" P28. Working equids' exclusion from livestock policies was a barrier that was discussed across organisations and NGOs, and advocating for their inclusion was a focus of their programming. Successful advocacy resulted in the acknowledgement of equids as 'working livestock' contributing to food security by the UN Committee of Food Security in 2016 [ 46 ]. Recognising working equids as livestock covered within national livestock policy definitions can have a range of beneficial effects [ 47 ]. These include access to national animal health strategies and services such as vaccination campaigns, disease surveillance and emergency relief schemes [ 45 ]. One participant highlighted the need for equid inclusion in current livestock disease control strategies, particularly in the case of zoonotic diseases that pose a serious risk to human health such as anthrax. "it affects cattle , it affects sheep , goats , horse , donkey , human , and other animals too . The funny thing is the government actually do the vaccine against cattle , but they don't vaccinate against donkeys… they're actually opening a pocket , the problem is still there" P11. Leaving equids unvaccinated can create reservoirs of disease in populations that can then aid transmission to other animals or potentially humans. Access Accessibility was raised as a barrier to the success of welfare initiatives, both regarding access to communities and access to all individuals responsible for maintaining equid welfare. It was discussed that some equid owning communities live in very remote rural areas where physical access is limited by the environment, such as mountain communities. By virtue of their position, these are often the communities that rely heavily on working equids for many roles including the transport of people and goods across difficult to navigate terrain where the use of a vehicle is not possible [ 48 ]. But nevertheless, for NGOs attempting to reach these communities, access is a challenge. "I think it's more their location makes it quite difficult for even our teams that are even based near them to even be able to access them , to bring the stuff that they need to kind of do the work . And it's usually the worst welfare conditions are in the areas that are the most remote . " P17. The COVID-19 pandemic has further exacerbated issues surrounding community access, a finding that has also been documented across NGO sectors [ 49 ]. It was described that remote communities also have problems accessing veterinary services due to their distance. This was especially mentioned in the case of brick kilns which, due to pollution regulations, are situated away from urban areas. "even if they want to call veterinary services , people don't come because of the distance and the remoteness of the place . " P8. Access to the full spectrum of individuals involved in the day-to-day care of equids was also another challenge highlighted by professionals. This is especially pertinent to cultures where the gender division of labour can mean that the individual working with the animal is not necessarily the person who undertakes routine equid management or attends an animal clinic. " So for example in [country] you often have children bring the donkey to the clinic for treatment and I remember on a particular occasion a young girl had brought a donkey , it had abscesses in both front feet and the vet said 'I thought I told you , you weren't to work the donkey' , clearly by the foot pads the donkey had been worked in between the treatments and the girl's reply through translation was 'I know what you said but how do I tell my father ? '" P6. Many participants described the need for programmatic activities such as meetings, workshops and educational events to include all individuals responsible for equid care. However, this was sometimes challenging; cultural norms in some areas can prohibit mixed-gender community participation and different time constraints across genders can mean that certain individuals are not available to take part in NGO programmes. In these situations NGOs can struggle to include the views of, or impart knowledge to all of the people necessary to make improving welfare successful. A lack of equid trained vets A frequently highlighted barrier to the success of equid welfare projects was the absence of training related to equines in the curriculum of veterinarians across countries. This has resulted in local veterinary service providers who do not have the specialist knowledge or skills to treat equine disease or injury. As many communities do not have the option to seek treatment for their animals elsewhere, this lack of access to appropriate treatment further compounds the poor welfare state of working equids. "If they manage to get somebody there for the veterinary services , the person who deals with the case aren't trained or have the skills to deal with donkeys and mules . " P8. While production animals appear to be the focus of much of the veterinary curriculum, equids in some areas are not commonly encountered by vets. "5th year vet students and they've not touched a horse , they're scared of it ! " P13. In some areas, even if there is provision made for the veterinary treatment of horses, knowledge of donkey treatment is severely lacking. It is common that no differentiation is made between treatments for different equid species. "Another problem is that the majority of people think that donkeys are just a small horse and they have the same needs" P1. Across NGOs there was a programming focus on working in partnership with veterinary universities to bring elements of equine training to the curriculum. "So we've tried to do a 101 working equid curriculum . That might be ideally , teachers that will take this on board and make them their own and listen to our recordings . And we're building slowly , sort of a database that they can have . " P31. Programmes where vet students can gain experience in the handling and treatment of equids were also run by NGOs with the hope of ensuring a better coverage of equid trained vets returning to their regions. "when the vets come and work with us , they can come with us in the field as part of the internship programme , start touching the donkeys and injecting donkeys , in our presence so it builds their confidence . " P8. Equid trained para-vets can also help to fill this gap in provision. Monitoring and evaluation challenges Many participants described an increased focus on monitoring and evaluation (M&E) of initiatives. This focus parallels changes seen far earlier in the field of international development where the need for accountability in the use of governmental and institutional funding resulted in more rigorous reporting and evaluation systems [ 50 ]. M&E was not perceived negatively by participants, all participants were in favour of tracking the progress of initiatives and felt that M&E was important to know how much of an impact their projects were making. However comprehensive and structured M&E processes were described across virtually all organisations as a relatively new introduction and rigorous M&E across the different approaches now utilised by organisations was highlighted as a challenge. "I would say this is relatively new to the programme so…it is a little bit trial and error . And it's also quite new to our teams . So there's a lot of work in that" P13. Apprehension was expressed that the increased need for formal evaluation has put extra pressure and workload onto those collecting M&E data at ground level. "So there's a lot of paperwork which the field team hates ! " P23. It also carried the risk that the time spent evaluating projects would result in less time devoted to carrying out the activities that would create desired change, a focus on upwards accountability seen in other NGO sectors [ 29 ]. "suddenly we went from actually going out there helping donkeys to spending all our time trying to prove we were helping donkeys" P4. With the implementation of new frameworks came concern that the expectations were not compatible with the practicalities of situations on the ground. "it was a very sort of tricky conversation to say , you know , we are not sure if this is realistic" P32. Participants spoke about the need to use indicators that are 'meaningful'. Historically numbers of animals reached by NGO programmes were the most common indicator of success measured [ 5 ]. However, this did not capture the impact that initiatives had on those animals reached and whether this led to any type of long-term welfare benefit. "Instead of just saying we treated , I think it used to be like 360 , 000 donkeys a year at that time , as [a colleague] quite rightly said 'that could be one donkey 360 , 000 times because you treated it really badly and haven't done anything for it ! ' What does it mean ? It doesn't mean anything" P4. It was also raised that what is deemed a meaningful indicator may differ by area. "So we also have the balance between indicators that Western developed countries think of as indicators versus what are really useful indicators here in more developing countries . " P19. With the focus shifting from numbers of animals reached to longer-term success, interviewees expressed difficulties in finding the most effective indicators to measure complex outcomes that can be influenced by multiple factors. This was especially true of multidisciplinary initiatives which seek to improve equid welfare indirectly through livelihood enhancement. "I think the biggest challenge with that is that it is successful , but it's been hard to measure how… how can you then know that that's directly improving equine welfare , if you're improving someone's livelihood" P20. It was also a concern that external factors could lead to an unrepresentative negative evaluation. "actually measuring real impact , particularly when…in the middle of your three years , there's a massive drought and people lose their livestock and body condition goes right down . You've been doing these beautiful welfare assessments and then suddenly… there's so much else that how do you then have all this data and have the thousand caveats that come with it ? " P13. However one participant suggested that the perceived difficulties surrounding comprehensive M&E can be unfounded. " one of the myths is that M&E is expensive and complicated and it certainly doesn't have to be . You can do it at a lot of different levels and I think that often the baby is thrown out with the bath water because people think you have to do so much when really you can still gain a lot of information from a smaller but dedicated and strategic approach ." P5. It was suggested that smaller changes, more reflective of the incremental nature of changes made by welfare initiatives could be a more suitable avenue than trying to immediately link initiatives to direct welfare improvement that may take far longer to be realised. "I think there is not enough emphasis placed on smaller indicators and looking at success in other ways… . So for example provision of water could be a small one , you could say 100% of people weren't providing water all the time that their animals were stabled and now 80% are , and although that might not lead all the way through to animal-based indicators of improvements in dehydration say , you've got those human behaviour indicators of what they're doing" P5. Conclusion The results highlight some of the major advantages and limitations of the variety of approaches used by equid welfare NGOs. The study provides an informed insight into potential factors to consider when designing future welfare initiatives. Discussion with experienced NGO staff from multiple roles has allowed a wide range of perspectives to be documented and has revealed the division of opinions regarding some of the approaches utilised by organisations. It has also demonstrated some of the current trends in the sector and highlighted approaches that have worked particularly well in certain contexts. Striking a balance appears to be a strong theme, at many levels. Balancing top-down versus bottom-up approaches was a frequently voiced concern, as was balancing the utility of certain approaches against associated factors warranting caution. Combinations of approaches that were felt be complementary provided a balance that drew on the strengths and mitigated for the weaknesses of different approaches. The need to tailor the approaches used to individual contexts was also raised and is especially relevant to INGOs working across multiple countries, cultures and political structures as some approaches were deemed unsuitable or unlikely to work in local environments. The collaborative pooling of experience across different NGOs could aid future programme planning, help make welfare initiatives more effective and provide a framework for NGOs in other fields to learn from each other's experiences and collective knowledge. Supporting information S1 File Semi structured interview themes. Participant document setting out the themes of the interview. (DOCX) Click here for additional data file. S2 File Approaches to improving working equid welfare. Participant document setting out the most commonly used programmatic approaches. (DOCX) Click here for additional data file.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2765779/

Synthesis of Homopolymers and Copolymers Containing an Active Ester of Acrylic Acid by RAFT: Scaffolds for Controlling Polyvalent Ligand Display

We describe the synthesis of activated homopolymers and copolymers of controlled molecular weight based on the controlled radical polymerization of N -acryloyloxysuccinimide (NAS) by reversible addition fragmentation chain transfer (RAFT). We synthesized activated homopolymers in a range of molecular weights with polydispersities between 1 and 1.2. The attachment of an inhibitory peptide to the activated polymer backbone yielded a potent controlled molecular weight polyvalent inhibitor of anthrax toxin. To provide greater control over the placement of the peptides along the polymer backbone, we also used a semi-batch copolymerization method to synthesize copolymers of NAS and acrylamide (AAm). This approach enabled the synthesis of copolymers with control over the placement of peptide-reactive NAS monomers along an inert backbone; subsequent functionalization of NAS with peptide yielded well-defined polyvalent anthrax toxin inhibitors that differed in their potencies. These strategies for controlling molecular weight, ligand density, and ligand placement will be broadly applicable for designing potent polyvalent inhibitors for a variety of pathogens and toxins, and for elucidating structure-activity relationships in these systems.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6584765/

Spatial relationships between small-holder farms coupled with livestock management practices are correlated with the distribution of antibiotic resistant bacteria in northern Tanzania

We examined the spatial distribution of antibiotic-resistant coliform bacteria amongst livestock from three distinct cultural groups, where group-level differences in practices (e.g., antibiotic use) may influence the magnitude of antibiotic resistance, while livestock interactions (e.g., mixing herds, shared markets) between these locations may reduce heterogeneity in the distribution of antibiotic resistant bacteria. Data was collected as part of a larger study of antibiotic-resistance in northern Tanzania. Simple regression and generalized linear regression were used to assess livestock management and care practices in relation to the prevalence of multidrug-resistant (MDR) coliform bacteria. Simple and multivariable logistic regression were then used to identify how different management practices affected the odds of households being found within MDR "hotspots." Households that had a higher median neighbourhood value within a 3000 m radius showed a significant positive correlation with livestock MDR prevalence (β = 4.33, 95% CI: 2.41–6.32). Households were more likely to be found within hotspots if they had taken measures to avoid disease (Adjusted Odds Ratio (AOR) 1.53, CI: 1.08—2.18), and if they reported traveling less than a day to reach the market (AOR 2.66, CI: 1.18—6.01). Hotspot membership was less likely when a greater number of livestock were kept at home (AOR 0.81, CI: 0.69–0.95), if livestock were vaccinated (AOR 0.32, CI: 0.21—0.51), or if distance to nearest village was greater (AOR 0.46, CI: 0.36–0.59). The probability of MDR increases when herds are mixed, consistent with evidence for passive transmission of resistant bacteria between animals. Reduced MDR with vaccination is consistent with many studies showing reduced antibiotic use with less disease burden. The neighbourhood effect has implications for design of intervention studies. 1 Introduction Our ability to treat infectious diseases has been seriously compromised by the emergence and dissemination of microorganisms that are resistant to antimicrobial agents [ 1 ]. Resistant pathogens, particularly multidrug-resistant (MDR) strains, are found globally, irrespective of geographic, economic, or cultural differences. In a 2014 report by the World Health Organization (WHO), high prevalence rates of antimicrobial resistance have been observed in all WHO-defined regions around the world, including the African Region, the Region of the Americas, the South-East Asia Region, the European Region, the Eastern Mediterranean Region, and the Western Pacific Region [ 2 ]. Despite being a global phenomenon, antimicrobial resistance (AMR) is particularly acute in low and middle income countries (LMICs), and can be attributed to several different factors [ [3] , [4] , [5] , [6] , [7] ]. High rates of infectious diseases drive greater demand for antibiotics and in turn, results in more widespread resistance [ 7 ]. The unregulated sale of antimicrobials for both people and livestock also contributes heavily to this issue [ [5] , [6] , [7] , [8] ], and is exacerbated by low rates of compliance with best practices for antimicrobial use— a problem likely attributed to reduced access to professional veterinary and medical care in LMICs [ [6] , [7] , [8] , [9] ]. Finally, financial constraints in LMICs often prevent the widespread availability of the most effective antimicrobials needed to tackle resistant infections, making management and containment of AMR extremely difficult here [ 5 , 7 , 8 ]. Another factor that is an important contributor to the rise and dissemination of resistant microorganisms is the close relationship between people and livestock in livestock-dependent communities [ [10] , [11] , [12] ]. In sub-Saharan Africa, over 600 million people live in households that rear livestock [ 13 ], which in turn can be a source of zoonotic and AMR pathogens. For example, in 2013, investigators in Burkina Faso examined 729 fecal samples of production animals for Salmonella ; 14% were resistant to antibiotics [ 14 ]. Other investigators in Hungary and China have reported possible links between antimicrobial use in cattle and swine, respectively, and resistant infections in humans [ 12 , 15 ]. While many aspects of AMR have been studied extensively worldwide, it is rare to incorporate information about spatial relationships between prevalence of resistant organisms and household- or community-level factors (but see [ [16] , [17] , [18] , [19] , [20] ]). This is a potentially important omission when neighbourhood effects (e.g., sharing of bacteria between adjacent households) impact the variance in the distribution of resistant bacteria. This can be particularly critical when developing and testing interventions because significant neighbourhood effects have the potential to interfere with treatment effects, increasing the probability of type 1 errors and consequently the rejection of test hypotheses. Adopting a spatial perspective is also critical for identifying areas where there is clustering of high AMR because knowledge of such clusters can play a key role in the development of targeted interventions in these areas. To better partition the variance in the distribution of resistant bacteria, a spatial technique called "hotspot analysis," can be implemented to evaluate whether high or low values of a variable cluster together [ 16 , 17 , 21 ]. A study by Kiffer et al. in 2011 used hotspot analysis to detect cluster emergence of ciprofloxacin-resistant E. coli in São Paulo, Brazil, which was then used to help demonstrate a positive correlation between ciprofloxacin use and resistance from a spatial perspective [ 17 ]. This is an interesting example of how spatial analyses can be used to elucidate important risk factors for AMR and highlights the potential importance of neighbourhood effects, such as one's increased risk of resistance due to antimicrobial usage by others in the area. Using data from a recent study by Caudell et al. [ 22 ] that examined antimicrobial use and veterinary care amongst three ethnic groups (Maasai pastoralists, Arusha agropastoralists, and Chagga highland farmers) in Tanzania's northern region, we employed a spatial approach to identify household clusters of increased livestock MDR bacteria, and the risk factors related to livestock management and care practices that are significantly associated with these clusters. 2 Methods 2.1 Survey data This analysis draws on data from a study of AMR in northern Tanzania [ 22 ]. A survey instrument was used to collect data on >200 variables covering topics about livelihood, veterinary care, and practices surrounding the use of antimicrobials, livestock movement networks, and geospatial data. The survey was conducted between March 2012 and July 2015 and was administered to over 400 households across 13 villages in the region. A subset of households ( n = 289) had accompanying antibiotic resistance data that was used in the current analysis. Detailed descriptions of the original study can be found here [ 22 ]. 2.2 Study groups Surveys included three different ethnic groups in northern Tanzania—Maasai, Chagga, and Arusha— each with varying livestock practices. The Maasai are pastoralists, who primarily tend large free-range cattle and some small-stock herds. While cattle are mostly kept for dairy and as household equity, other small-stock are typically sold for cash or represent a source of meat for regular household consumption. The Chagga are highland farmers, and primarily tend small, zero-grazing herds of cattle and small-stock for subsistence and market sales. The Arusha are agro-pastoralists and have animal husbandry systems that share similarities to both the Maasai and Chagga. They also typically tend small herds of cattle and small-stock. These three groups provided a wide range of livestock practices that could be assessed in terms of their contribution to the risk of carrying MDR bacteria. More detailed descriptions about the three groups can be found here [ 22 ]. 2.3 Sampling Focal villages were selected based on ethnic composition, and in consultation with local research assistants. Within these villages, households were randomly sampled using census data supplied by local Village Executive Officers. Household heads provided information, and household latitude/longitude were recorded. 2.4 Sampling and antibiotic susceptibility testing The same sampling protocol was used for each household. Up to three fresh but distinct fecal samples were collected for each species and combined separately by species into sterile plastic bags (sheep and goats were treated as the same species for this analysis). If a household did not own more than three animals, samples were collected from all animals. Samples were then transported to a lab at the Nelson Mandela African Institute of Science and Technology (Arusha, TZ). Samples were plated on MacConkey agar for isolation of lactose-fermenting Gram-negative bacteria (mostly Escherichia coli , Klebsiella and Enterobacter ). Up to 48 isolates were selected per sample based on colony morphology consistent with E coli . Isolates ( n > 70,000) were then shipped to Washington State University where they were tested for resistance to nine antibiotics using a "breakpoint" assay. With this assay, if a bacterial isolate grows on a MacConkey agar plate in the presence of a concentration of antibiotic that normally inhibits susceptible bacteria, then the isolate is classified as resistant. In-depth methods for sample processing and antibiotic susceptibility testing were published earlier [ 23 ]. 2.5 Statistical analysis Antibiotic-resistance testing data (binary result for each antibiotic) was pooled across cattle, sheep and goats. MDR values at the household level represent the prevalence of isolates that were resistant to three or more antibiotics. This MDR value served as the dependent variable ("livestock MDR") for the subsequent analyses. Simple regression was used to first examine the household practices relating to livestock management and care, as well as spatial factors that might influence the prevalence of livestock MDR. Twelve independent variables relating to livestock management and care were selected for the regression analyses. Variables were chosen based on behaviours that have been theoretically or empirically linked to MDR [ 26 ], and through previous ethnographic work in these populations [ 22 ]. These can largely be grouped according to the following categories: Livestock management practices included whether livestock came into contact with other livestock (0 = no, 1 = yes), and the numbers of livestock a household kept at home and managed (continuous). Livestock feeding patterns included whether a household had access to common grazing land ("communal graze") (0 = no, 1 = yes), whether or not grazing patterns changed if livestock were sick (0 = no, 1 = yes), and whether wildlife and livestock shared a water source (0 = no, 1 = yes). Disease avoidance measures included the number of vaccines administered to livestock ("vaccinations") (continuous), antibiotic use ("antibiotic use") (measured on a scale from 0 (low) to 5 (high)), whether households participated in withdrawal of milk or meat for a period after treatment of livestock with antimicrobials (0 = no, 1 = yes), and the number of measures taken to avoid disease (continuous),. This last variable represents a variety of possible measures the household may partake in to avoid disease in their livestock, including but not limited to building a shed or pen to prevent contact of livestock with other domestic or wild animals, grazing sick livestock separately, feed supplementation, treatment with allopathic medicine, treatment with traditional medicine, spraying, and vaccinations. Sales yards and villages included the length of time it takes for a person in the household to reach a sales yard (market) to purchase livestock ("time to sales yard 1 day"; "time to sales yard >2 days"), and also the distance to the nearest village (continuous). Given that only a small percentage of people use cars, and public transportation is not common in these regions, almost all people walk to the nearest village on small paths. The Euclidian metric was therefore chosen to measure the distance to the nearest village as these paths are understandably the closest possible to such a beeline measurement. To investigate whether households with high MDR values are spatially clustered, a point distance analysis was performed for two different search radii (3000 m and 4000 m) using ArcGIS version 10.5.1. Search radii were pre-determined using incremental spatial autocorrelation. These radii were used to create "neighbourhood boundaries" to determine the median neighbourhood MDR value for each household. The median neighbourhood MDR value for a household was calculated using only the MDR values of those households within the set radius; the household for which the median neighbourhood value was being determined was not included in the calculation. Simple regression was performed for each to examine whether median neighbourhood MDR values are associated with an increased likelihood of households having a higher prevalence of livestock-associated MDR isolates. Using these spatial variables in addition to the livestock management and care practices for which the regression was significant at P 200 variables covering topics about livelihood, veterinary care, and practices surrounding the use of antimicrobials, livestock movement networks, and geospatial data. The survey was conducted between March 2012 and July 2015 and was administered to over 400 households across 13 villages in the region. A subset of households ( n = 289) had accompanying antibiotic resistance data that was used in the current analysis. Detailed descriptions of the original study can be found here [ 22 ]. 2.2 Study groups Surveys included three different ethnic groups in northern Tanzania—Maasai, Chagga, and Arusha— each with varying livestock practices. The Maasai are pastoralists, who primarily tend large free-range cattle and some small-stock herds. While cattle are mostly kept for dairy and as household equity, other small-stock are typically sold for cash or represent a source of meat for regular household consumption. The Chagga are highland farmers, and primarily tend small, zero-grazing herds of cattle and small-stock for subsistence and market sales. The Arusha are agro-pastoralists and have animal husbandry systems that share similarities to both the Maasai and Chagga. They also typically tend small herds of cattle and small-stock. These three groups provided a wide range of livestock practices that could be assessed in terms of their contribution to the risk of carrying MDR bacteria. More detailed descriptions about the three groups can be found here [ 22 ]. 2.3 Sampling Focal villages were selected based on ethnic composition, and in consultation with local research assistants. Within these villages, households were randomly sampled using census data supplied by local Village Executive Officers. Household heads provided information, and household latitude/longitude were recorded. 2.4 Sampling and antibiotic susceptibility testing The same sampling protocol was used for each household. Up to three fresh but distinct fecal samples were collected for each species and combined separately by species into sterile plastic bags (sheep and goats were treated as the same species for this analysis). If a household did not own more than three animals, samples were collected from all animals. Samples were then transported to a lab at the Nelson Mandela African Institute of Science and Technology (Arusha, TZ). Samples were plated on MacConkey agar for isolation of lactose-fermenting Gram-negative bacteria (mostly Escherichia coli , Klebsiella and Enterobacter ). Up to 48 isolates were selected per sample based on colony morphology consistent with E coli . Isolates ( n > 70,000) were then shipped to Washington State University where they were tested for resistance to nine antibiotics using a "breakpoint" assay. With this assay, if a bacterial isolate grows on a MacConkey agar plate in the presence of a concentration of antibiotic that normally inhibits susceptible bacteria, then the isolate is classified as resistant. In-depth methods for sample processing and antibiotic susceptibility testing were published earlier [ 23 ]. 2.5 Statistical analysis Antibiotic-resistance testing data (binary result for each antibiotic) was pooled across cattle, sheep and goats. MDR values at the household level represent the prevalence of isolates that were resistant to three or more antibiotics. This MDR value served as the dependent variable ("livestock MDR") for the subsequent analyses. Simple regression was used to first examine the household practices relating to livestock management and care, as well as spatial factors that might influence the prevalence of livestock MDR. Twelve independent variables relating to livestock management and care were selected for the regression analyses. Variables were chosen based on behaviours that have been theoretically or empirically linked to MDR [ 26 ], and through previous ethnographic work in these populations [ 22 ]. These can largely be grouped according to the following categories: Livestock management practices included whether livestock came into contact with other livestock (0 = no, 1 = yes), and the numbers of livestock a household kept at home and managed (continuous). Livestock feeding patterns included whether a household had access to common grazing land ("communal graze") (0 = no, 1 = yes), whether or not grazing patterns changed if livestock were sick (0 = no, 1 = yes), and whether wildlife and livestock shared a water source (0 = no, 1 = yes). Disease avoidance measures included the number of vaccines administered to livestock ("vaccinations") (continuous), antibiotic use ("antibiotic use") (measured on a scale from 0 (low) to 5 (high)), whether households participated in withdrawal of milk or meat for a period after treatment of livestock with antimicrobials (0 = no, 1 = yes), and the number of measures taken to avoid disease (continuous),. This last variable represents a variety of possible measures the household may partake in to avoid disease in their livestock, including but not limited to building a shed or pen to prevent contact of livestock with other domestic or wild animals, grazing sick livestock separately, feed supplementation, treatment with allopathic medicine, treatment with traditional medicine, spraying, and vaccinations. Sales yards and villages included the length of time it takes for a person in the household to reach a sales yard (market) to purchase livestock ("time to sales yard 1 day"; "time to sales yard >2 days"), and also the distance to the nearest village (continuous). Given that only a small percentage of people use cars, and public transportation is not common in these regions, almost all people walk to the nearest village on small paths. The Euclidian metric was therefore chosen to measure the distance to the nearest village as these paths are understandably the closest possible to such a beeline measurement. To investigate whether households with high MDR values are spatially clustered, a point distance analysis was performed for two different search radii (3000 m and 4000 m) using ArcGIS version 10.5.1. Search radii were pre-determined using incremental spatial autocorrelation. These radii were used to create "neighbourhood boundaries" to determine the median neighbourhood MDR value for each household. The median neighbourhood MDR value for a household was calculated using only the MDR values of those households within the set radius; the household for which the median neighbourhood value was being determined was not included in the calculation. Simple regression was performed for each to examine whether median neighbourhood MDR values are associated with an increased likelihood of households having a higher prevalence of livestock-associated MDR isolates. Using these spatial variables in addition to the livestock management and care practices for which the regression was significant at P  1 day No 258 (89.58%) Yes 30 (10.42%) Time to sales yard > 2 days No 270 (93.75%) Yes 18 (6.25%) Nearest village distance (km) (median, IQR) (3.44, 1.96–6.40) IQR = interquartile range a Number of livestock at home refers to livestock that do not leave the house to graze but are brought fodder and water. These animals are kept at the home all day, and differ from livestock that go in and out of the household for grazing and watering. b Vaccinations include anthrax, East Coast fever, foot-and-mouth disease, contagious bovine pleuropneumonia, and rabies. Five of the sixteen variables were identified from the simple regression analyses for consideration in the model ( P 1 day 0.05 (−0.39–0.45) – >2 days −0.09 (−0.69–0.42) – Nearest village distance (km) −0.06 (−0.13–0.16) −0.012 (−0.09-0.06) Median neighbourhood MDR 3000 m, 4.41 (2.54–6.35) ⁎ 4.33(2.41–6.32)** Median neighbourhood MDR 4000 m, 4.39 (2.52–6.34) ⁎ – ⁎ P 1 day 0.36 (0.12–1.05) – Time to sales yard >2 days 0.00 (0.00–0.00) – Nearest village distance (km) 0.53 (0.44–0.65) ⁎ 0.46 (0.36–0.59) ⁎ ⁎ P < 0.05 4 Discussion While many studies have examined risk factors contributing to MDR globally, spatial factors concerning the prevalence of resistant organisms are often overlooked. However, consideration of spatial context and neighbourhood effects are critical to better understand risk factors associated with MDR, and in particular for developing and testing effective interventions, as findings from our research suggest. The first model examined was the adjusted median neighbourhood analysis model and it identified only one variable as significant— the median neighbourhood livestock MDR value for a 3000 m radius. Households having a higher neighbourhood MDR value for a 3000 m radius showed significant positive correlation with having a higher prevalence of MDR isolates in their livestock population, suggesting a strong spatial component to the spread of MDR beyond the livestock management practices of a particular household . That is, even if a household adheres to recommended best practices for antimicrobial use, having neighbours who do not may be enough to put that household at risk. This finding is important not only for understanding the spread of MDR bacteria but also for understanding interventions to address MDR. Currently, many interventions focus on household-level practices (i.e. using antibiotics more prudently) [ 9 ], and not necessarily on practices linked directly to transmission. Nevertheless, if neighbourhood effects do play a role in prevalence of household MDR as these results suggest, then implementing an intervention without considering spatial effects could impact its efficacy. An improved understanding of the spatial component of MDR bacteria is therefore required if we are to develop effective interventions to manage resistance. Results of the hotspot analysis performed on the study area also indicate statistically significant spatial clusters of households with high (hotspot) or low (coldspot) livestock MDR values, further highlighting the link between proximity and MDR. These hotspots serve as key indicators where multi-drug resistance is most pressing— a critical first step in implementing interventions to reduce the emergence and dissemination of resistance. The follow-up adjusted hotspot analysis model can be considered the second critical step for interventions, as it offers insight into the potential risk factors giving rise to the distribution of MDR bacteria. Hotspot membership was more likely for households with a shorter time to the sales yard and households that take more measures to avoid disease. Hotspot membership was less likely for households that keep a greater number of livestock at home, households with a greater distance to the nearest village and for households who gave a greater number of vaccinations to livestock. For households that visited sales yards less than one day away, the increased likelihood result aligns with our expectations that households with shorter travel time might frequent sales yards more often as opposed to those further away, resulting in increased passive transmission [ 23 , 27 ]. For households that take more measures to avoid disease, this result was surprising, however it may reflect greater willingness to use antibiotics in livestock [ 1 ]. As this variable represents a variety of different disease avoidance measures, further investigation is needed to better pinpoint which measures specifically used by these households contribute to this result. Conversely, households that kept a greater number of livestock at home, households with greater distances to the nearest village, and households who demonstrated increased vaccine use all reduced likelihood of hotspot membership. By keeping livestock at home, it is likely that instances of passive transmission that can allow MDR to propagate—including contact with other herds during grazing and watering—were minimized, thereby reducing the transmission of MDR. Several lines of indirect evidence are consistent with the possibility of passive transmission. First, examination of E. coli isolates from waterholes in our research area where people, livestock, and wildlife congregate show resistant profiles similar to livestock [ 23 ]. Second, we documented the presence of resistant E. coli in samples from wildlife (consistent with other studies) [ 27 ], which could not be the result of direct selection given antibiotics are not used on these populations. The observation regarding households with greater distances to the nearest village is consistent with studies finding that proximity to urban areas increases AMR prevalence [ [28] , [29] , [30] ], and reiterates the importance of spatial consideration with regards to resistance. Finally, the observation regarding vaccine use could reflect reduced reliance on antibiotics with livestock even though some of the vaccines were for viral agents, suggesting that increased vaccine use might be an important tool to interrupt the spread of livestock-related AMR. Potential limitations of the current study include its reliance on some self-reported data that could be skewed by a social-desirability bias [ 22 ]. This might be especially true if respondents were aware of behaviours that promote MDR bacteria, like frequent, widespread antimicrobial use. In less literate populations where ethnomedical beliefs often run counter to Western biomedical science [ 31 ], this bias may be limited. The primary outcome for these analyses (household prevalence of livestock MDR bacteria) is a quantitative measure that was not self-reported, and consequently not subject to such bias. Furthermore, while fecal samples were easily distinguishable by species, it was not possible to consistently identify the breed and age of the animal producing fecal pats that were sampled. Consequently, these two variables could not be included in the model. Importantly, however, there was no significant species effect on the likelihood of detecting MDR bacteria. This study also only examines factors contributing to bacterial resistance amongst livestock, which is just one facet driving the emergence and transmission of AMR. While some studies have begun to examine the human dimension to this issue [ 32 ], more research is required if we are to develop a more substantive and robust explanatory model of AMR. Acknowledgements We are grateful for the invaluable assistance of many people including Chagga, Arusha, and Maasai study participants; village chairmen of focal villages, Misters Isaya Rumas, Godfrey Naisikye, Lemuta Naisikye, Imma Laiser, Willium Kanunga, Imani Baraka, Paul Sangre, Joseph Tarimo, Joseph Temo, Rigobert Tarimo, Deogratius Mshanga; Profs. Guy Palmer, Terry McElwain, and Jonathan Yoder (Washington State University), Profs. Joann Sharpe (University of Glasgow) and Jonathan Rushton (University of London), and Dr. Julius Keyyu and Dr. Victor Kakengi (Tanzania Wildlife Research Institute). Funding This work was supported by the National Science Foundation : Ecology and Evolution of Infectious Diseases [DEB1216040]. Ethics approval Ethical review for human subjects was approved by the Washington State University (IRB #12533) and by the Tanzania National Institute for Medical Research (Tanzania Commission for Science and Technology permit #2012–151). Research permits were obtained from the Tanzania Commission for Science and Technology, the Tanzania Wildlife Research Institute, and by the regional, district, and ward offices of the Arusha, Manyara, and Kilimanjaro areas where the research was conducted. Consent for publication Given high rates of illiteracy amongst the study population, consent for publication was obtained verbally, and the local research assistants signed on their behalf. Approval of this consent procedure was obtained from Washington State University's Institutional Review Board. Availability of data and material The data that support the findings of this study are available upon reasonable request from the author [DRC]. Declarations of interest None. Funding National Science Foundation: Ecology and Evolution of Infectious Diseases: DEB1216040. Authors' contributions Study concept and design: OA, LR; development of survey protocols and acquisition of data: DRC, MAC; analysis and interpretation of data: OA, LR; drafting of the manuscript: LR; critical revision of the manuscript for important intellectual content and for final approval: OA, DRC, MAC, NS; principal investigator: DRC. All authors have read and approved the final version.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3351443/

Suppression of Ribosomal Function Triggers Innate Immune Signaling through Activation of the NLRP3 Inflammasome

Some inflammatory stimuli trigger activation of the NLRP3 inflammasome by inducing efflux of cellular potassium. Loss of cellular potassium is known to potently suppress protein synthesis, leading us to test whether the inhibition of protein synthesis itself serves as an activating signal for the NLRP3 inflammasome. Murine bone marrow-derived macrophages, either primed by LPS or unprimed, were exposed to a panel of inhibitors of ribosomal function: ricin, cycloheximide, puromycin, pactamycin, and anisomycin. Macrophages were also exposed to nigericin, ATP, monosodium urate (MSU), and poly I:C. Synthesis of pro-IL-ß and release of IL-1ß from cells in response to these agents was detected by immunoblotting and ELISA. Release of intracellular potassium was measured by mass spectrometry. Inhibition of translation by each of the tested translation inhibitors led to processing of IL-1ß, which was released from cells. Processing and release of IL-1ß was reduced or absent from cells deficient in NLRP3, ASC, or caspase-1, demonstrating the role of the NLRP3 inflammasome. Despite the inability of these inhibitors to trigger efflux of intracellular potassium, the addition of high extracellular potassium suppressed activation of the NLRP3 inflammasome. MSU and double-stranded RNA, which are known to activate the NLRP3 inflammasome, also substantially inhibited protein translation, supporting a close association between inhibition of translation and inflammasome activation. These data demonstrate that translational inhibition itself constitutes a heretofore-unrecognized mechanism underlying IL-1ß dependent inflammatory signaling and that other physical, chemical, or pathogen-associated agents that impair translation may lead to IL-1ß-dependent inflammation through activation of the NLRP3 inflammasome. For agents that inhibit translation through decreased cellular potassium, the application of high extracellular potassium restores protein translation and suppresses activation of the NLRP inflammasome. For agents that inhibit translation through mechanisms that do not involve loss of potassium, high extracellular potassium suppresses IL-1ß processing through a mechanism that remains undefined. Introduction Aberrant interleukin-1ß (IL-1ß) signaling has been implicated in a variety of inflammatory diseases ranging from arthritis to diabetes, making the manipulation of the IL-1 pathway an attractive therapeutic option for a growing number of pathologies that stem from innate immune activation [1] , [2] . Critical to the efficacy of the innate immune system is the proper detection of invading microbes and toxic substances by macrophages that express pattern recognition receptors (PRRs) in the cytosol and at the cell surface. The Nod-like receptor (NLR) family member, NLRP3, is a cytosolic PRR that is activated by a large array of pathogen- and danger-associated molecular patterns to stimulate IL-1ß processing by a multiprotein complex termed the inflammasome [3] . The NLRP3 inflammasome consists of NLRP3, caspase-1, and the adaptor protein, ASC [3] , [4] , [5] , [6] . Bacterial pore-forming toxins, viruses, asbestos, ATP, double-stranded RNA, and uric acid crystals all stimulate IL-1ß processing via NLRP3 inflammasomes [6] , [7] , [8] , [9] . Although the importance of the inflammasome in mediating the release of IL-1ß from cells is well recognized, the mechanism(s) by which disparate activators trigger inflammasome activation are incompletely understood. In macrophages, proinflammatory signals are required to mediate the expression of mRNA from the IL-1ß gene, resulting in the accumulation of pro-IL-1ß protein. These initial, or priming, signals are mediated by Toll-like receptor ligands such as lipopolysaccharide (LPS), which direct the NF-kappaB-dependent expression of pro-IL-1ß [3] , [10] . The proteolytic processing of pro-IL-1ß by caspase-1 and the subsequent release of IL-1ß from cells requires a second signal to stimulate the assembly of inflammasome complexes. Loss of intracellular potassium has emerged as a frequent correlate of NLRP3 inflammasome activation and has been proposed to constitute one such signal. The conclusion that decreased intracellular potassium acts as a second signal to trigger activation of the NLRP3 inflammasome was based initially on the observation that loss of potassium induced by nigericin, a potassium ionophore, or by ATP results in the robust release of IL-1ß from cells in an NLRP3-dependent manner [7] , [11] . However, the mechanism by which loss of intracellular potassium is linked with activation of the NLRP3 inflammasome is unclear. The production of reactive oxygen species (ROS) as a result of mitochondrial dysfunction has also been proposed as an activator of the NLRP3 inflammasome [8] , [11] , [12] , [13] , although the validity of this conclusion has been questioned [14] , [15] , [16] . It has been shown that sufficient levels of potassium are required for elongation of the peptide chain on the ribosome in vitro [17] , [18] , [19] , [20] and that reduced concentrations of intracellular potassium fail to support protein synthesis in mammalian cells [20] , [21] . Recently, we reported that ricin toxin, a well-defined inhibitor of protein synthesis, leads to IL-1ß- and macrophage-dependent inflammation when delivered to the lungs of mice [22] . By examining ricin-mediated release of IL-1ß from primary bone marrow-derived macrophages (BMDM), we found that the release of IL-1ß from BMDM requires the known components of the NLRP3 inflammasome: NLRP3, caspase-1, and ASC [23] . This finding led us to test whether inhibition of protein translation per se could constitute a signal that is sufficient to activate the NLRP3 inflammasome. Here we employed a panel of well characterized protein synthesis inhibitors and found that each inhibitor induced the release of IL-1ß from LPS-primed BMDM in a manner that was dependent on the NLRP3 inflammasome. In BMDM treated with inhibitors of protein synthesis, intracellular potassium concentrations remained constant during the time in which IL-1ß release was observed from these cells, suggesting that activation of the NLRP3 inflammasome did not result from loss of potassium. To address whether inhibition of protein synthesis is a common feature of NLRP3 inflammasome activation, we examined two clinically relevant inflammasome triggers, monosodium urate, which causes inflammatory gout, and poly I:C, which mimics the proinflammatory actions of viral dsRNA. We report that monosodium urate and poly I:C each suppressed protein synthesis at doses and times that correspond with the release of IL-1ß from BMDM. We treated BMDM with NLRP3 stimuli in the presence of proteasome inhibitors, and found that proteasome activity was required for the ability of translational inhibitors, including MSU and poly I:C, to induce IL-1ß secretion from cells. Our results demonstrate that impaired ribosomal function may constitute a common mechanism of NLRP3 inflammasome activation through a process that may require proteasomal activity. These studies further suggest that decreased cellular potassium may trigger IL-1ß release by inhibiting protein translation. Results Activation of the NLRP3 Inflammasome by Inhibitors of Translation We tested the ability of a panel of well-characterized inhibitors of translation (cycloheximide, emetine, puromycin, pactamycin, and anisomycin) to stimulate IL-1ß release from BMDM via the NLRP3 inflammasome. When added at doses that suppress incorporation of [ 3 H]-leucine by at least 90%, each inhibitor led to release of IL-1ß from LPS-primed BMDM by 4 h ( Fig. 1A ). Conversion of pro-IL-1ß to IL-1ß induced by cycloheximide proceeded in a dose-dependent manner ( Fig. 1B ), suggesting that the degree of translational impairment correlated well with the amount of IL-1ß that was released. As expected, there was a reciprocal relationship between the amount of pro-IL-1ß present in the cell lysate ( Fig. 1A , upper panel) and the amount of IL-1ß detected in the culture medium by ELISA ( Fig. 1A , lower panel). Cells remained strongly adherent to culture dishes at the termination of the experiments, and did not show morphological signs of apoptosis or activation of caspase-3 (not shown). To gain insight into the mechanism of IL-1ß release triggered by translational inhibitors, we exposed BMDM from mice deficient in NLRP3, ASC, or caspase-1 to agents that inhibited translation by a variety of mechanisms and compared their responses to those of wild-type (WT) cells. When BMDM deficient in NLRP3, ASC, or caspase-1 were exposed to the inhibitors, the cells secreted substantially reduced amounts of IL-1ß compared with WT BMDM exposed to the same inhibitors ( Fig. 1C ), suggesting that processing and release of IL-1ß were mediated through formation of the NLRP3 inflammasome. 10.1371/journal.pone.0036044.g001 Figure 1 Inhibition of protein synthesis triggers secretion of IL-1ß from primed BMDM. A) WT BMDM were stimulated or not with 50 ng/ml LPS for 4 h followed by exposure to 10 µg/ml ricin, 25 µg/ml cycloheximide, 10 µg/ml emetine, 75 µg/ml puromycin, 0.2 µg/ml pactamycin, or 10 µg/ml anisomycin for 4 h prior to harvesting. Cell extracts were analyzed by immunoblotting for p38 MAPK (loading control) and pro-IL-1ß (upper panel). Media supernatants were analyzed by ELISA for released IL-1ß (lower panel). ELISA data are represented as mean ± SEM from triplicate dishes. B) LPS-primed WT BMDM were stimulated for 4 h with increasing doses of cycloheximide (from right to left: 0.03, 0.1, 0.3, 1.0, 3.0, 10, 30, and 100 µg/ml). Media supernatants were analyzed by ELISA for released IL-1ß (upper panel). Cell extracts were analyzed by immunoblotting for p38 MAPK (loading control) and pro-IL-1ß (lower panel). ELISA data are represented as mean ± SEM from triplicate dishes. C) Release of IL-1ß triggered by inhibitors of translation was determined by immunoblotting of cell extracts (cell) or culture medium (medium) from LPS-primed BMDM obtained from WT mice or mice deficient in NLRP3, ASC, or caspase-1, as indicated. P38 MAPK was loading control. Extracellular Potassium Suppresses Release of IL-1ß Induced by Inhibitors of Translation Previous studies have described a requirement for potassium efflux by a variety of agents reported to activate the NLRP3 inflammasome, including nigericin and ATP [11] . Here, substitution of 130 mM Na + /5 mM K + by 5 mM Na + /130 mM K + in the culture medium suppressed the appearance of mature IL-1ß by translation inhibitors to a similar extent as nigericin and ATP, as detected by ELISA ( Fig. 2A ). Immunoblotting of the high potassium culture medium revealed undetectable amounts of mature IL-1ß in the culture medium of LPS-primed cells exposed to nigericin or ATP ( Fig. 2B ), or ricin, emetine, or cycloheximide ( Fig. 2C ). 10.1371/journal.pone.0036044.g002 Figure 2 Effect of extracellular potassium on IL-1 processing and release. A) Bone marrow-derived macrophages were plated in triplicate wells in 12-well plates and primed with 50 ng/ml of LPS for 4 h. Cells were then incubated in medium containing 130 mM NaCl/5 mM KCl or 5 mM KCl/130 mM NaCl in the presence of absence of 0.01 µg/mL ricin, 25 µg/mL cycloheximide, 10 µg/mL, 10 µg/mL emetine, 75 µg/mL puromycin, 0.2 µg/mL pactamycin, 10 µg/mL anisomycin, 3.4 µM nigericin, or 5 mM ATP for 4 h. Medium was collected, and p17 IL-1 was determined by ELISA. B) Macrophages were plated, primed with LPS and incubated in medium containing 130 mM NaCl/5 mM KCl (MEM-Na) or 5 mM NaKCl/130 mM KCl (MEM-K) in the presence or absence 5 mM ATP or 3.4 µM nigericin for 4 h. Proteins were precipitated from the media with TCA and analyzed by Western blotting. C) Macrophages were plated, primed with LPS, and incubated in medium containing 130 mM NaCl/5 mM KCl (MEM-Na) or 5 mM NaKCl/130 mM KCl (MEM-K) in the presence or absence 10 ng/mL ricin, 10 µg/mL emetine, or 25 µg/mL cycloheximide for 4 h. Proteins were precipitated with TCA and analyzed by Western blotting. Lack of Correlation between Translation Inhibition and Potassium Efflux The ability of high extracellular potassium to suppress activation of the NLRP3 inflammasome by multiple proinflammatory agonists [11] has been used to support a model demonstrating that decreased cellular potassium constitutes a general trigger for inflammasome activation. Initial studies were conceived because nigericin and ATP, agents that are known to cause potassium efflux [7] , [24] , are also potent activators of the NLRP3 inflammasome. The conclusion that loss of cellular potassium is responsible for activation of the NLRP3 inflammasome was based on the ability of high extracellular potassium to suppress pro-IL-1ß processing and release of mature IL-1ß in the presence of high potassium [11] . Subsequent studies have employed high extracellular potassium to block activation of the NLRP3 by a variety of agents [13] , [25] , [26] , [27] , [28] , [29] , [30] . Although these studies demonstrated that high extracellular potassium blocks release of IL-1ß, they did not determine whether the proinflammatory agents employed actually cause a loss of cellular potassium. To determine whether inhibition of translation would trigger efflux of potassium from macrophages, we exposed LPS-primed BMDM to emetine, a potent inhibitor of translation [31] , [32] that irreversibly inhibits the elongation cycle of translation by greater than 99% within 1 minute [33] and produces release of IL-1ß within 4 h. Intracellular potassium was measured by inductively coupled plasma mass spectroscopy (ICP-MS). Nigericin and ATP, agents that have been shown to reduce intracellular potassium [7] , [24] , produced a 50% decrease in intracellular potassium concentration by 15 min and 45 min, respectively ( Fig. 3A ). BMDM exposed to nigericin and ATP detached from the culture dishes by 1 h. By contrast, cells exposed to emetine maintained normal levels of intracellular potassium for at least 4 h ( Fig. 3A ). Ricin and cycloheximide similarly failed to cause release of cellular potassium (not shown). These data suggest that in BMDM the inhibition of translation promoted the activation of the NLRP3 inflammasome but failed to induce leakage of cellular potassium in the interval of time (4 h) that preceded activation of the inflammasome and release of IL-1ß. 10.1371/journal.pone.0036044.g003 Figure 3 Inhibition of protein synthesis fails to elicit efflux of potassium from cells. A) Intracellular potassium in WT BMDM was analyzed by ICP-MS after priming with LPS for 4 h and exposure to emetine, ATP, and nigericin for indicated times. Data are represented as mean ± SEM from triplicate dishes. Measurements of potassium release by cells exposed to ATP and nigericin were terminated at 10 min and 60 min, respectively, as a result of cytotoxicity and detachment of cells from the dishes. Cells exposed to emetine showed no signs of cytotoxicity or detachment by 4 h. B) Elevated potassium blocks emetine-mediated release of IL-1ß and processing of caspase-1. LPS-primed BMDM were either exposed or not exposed to 10 µg/ml emetine for 4 h, at which time proteins were precipitated from the media with TCA and analyzed by Western blotting as shown. C) BMDM in triplicate wells were pulse-labeled in medium containing [ 3 H]-leucine for 15 min prior to harvest at the indicated times and the amount of [ 3 H]-leucine incorporation was measured. Four h after exposure of LPS-primed BMDM to emetine in medium containing low potassium, both IL-1ß and a proteolytic fragment of caspase-1 (p10) appeared in the medium ( Fig. 3B , 2nd lane). In the presence of elevated potassium, neither p10 capase-1 nor IL-1ß appeared in the medium of cells exposed to emetine ( Fig. 3B , 4th lane), suggesting that high potassium blocked caspase-1-mediated processing of pro-IL-1ß. Coupling between Potassium Efflux and Inhibition of Protein Translation In view of the differential requirement for potassium efflux in promoting NLRP3 activation, we sought to determine whether nigericin, ATP, and panel of translation inhibitors that we employed shared any mechanistic features. It has been well established that the rate of protein synthesis is directly dependent on the concentration of intracellular potassium [18] , [34] and that potassium ionophores inhibit translation at the same doses and kinetics as potassium efflux [35] . To test the ability of nigericin and ATP to suppress protein synthesis in BMDM, we subjected cells to nigericin or ATP and measured incorporation of [ 3 H]-leucine. Exposure of BMDM to either nigericin or ATP triggered rapid inhibition of protein synthesis by 30 min in normal medium (130 mM Na + /5 mM K + ; Fig. 3C ). Replacement of extracellular Na + by K + (130 mM K + /5 mM Na + ) partially reversed the blockade of protein synthesis in ATP-treated cells. Replacement of extracellular Na + by K + in nigericin-treated cells restored protein synthesis to a level even greater than that observed in untreated cells ( Fig. 3C ). These data are consistent with the known association of decreased cellular potassium with the inhibition of protein synthesis by nigericin, ATP, and other potassium ionophores [34] , [35] , [36] . Monosodium Urate Crystals Suppress Protein Translation Gout is a chronic inflammatory response that is associated with the formation of crystals of monosodium urate (MSU). Incubation of LPS-primed murine macrophages with MSU crystals leads to the conversion and release of IL-1ß by a process that requires activation of the NLRP3 inflammasome. Because increased extracellular potassium blocks the activation of the NLRP3 inflammasome by MSU, it has been concluded that MSU acts through a mechanism that involves a decreased concentration of cellular potassium [11] , [37] , [38] . To determine if MSU inhibits protein synthesis at doses that stimulate release of IL-1ß from BMDM, we exposed cells to increasing doses of MSU or monopotassium urate (MPU), a salt of uric acid that is ineffective in promoting IL-1ß release [38] . The secretion of mature IL-1ß induced by MSU crystals was associated in a dose-dependent manner with inhibition of [ 3 H]-leucine incorporation into BMDM ( Fig. 4A,B ). Monopotassium urate (MPU), which, as expected, failed to stimulate conversion of pro-IL-1ß to mature IL-1ß ( Fig. 4B ), also failed to inhibit protein synthesis as measured by incorporation of [ 3 H]-leucine ( Fig. 4A ). These results demonstrated that MSU-induced suppression of protein synthesis in macrophages correlated with the release of IL-1ß by MSU. 10.1371/journal.pone.0036044.g004 Figure 4 MSU crystals inhibit protein synthesis at concentrations that induce processing and release of IL-1ß. WT BMDM were primed with LPS for 4 h prior to exposure MSU or MPU at indicated concentrations. A) BMDM in triplicate wells were pulse-labeled in medium containing [ 3 H]-leucine for 15 min prior to harvest at the indicated times and the amount of [ 3 H]-leucine incorporation was measured. B) Cells were harvested 4 h after addition of indicated concentrations of MSU and MPU. Cell lysates (cell) and culture medium (medium) were examined by immunoblotting. P38 MAPK was loading control. Double-stranded RNA Inhibits Protein Translation and Activates the NLRP3 Inflammasome Many forms of cellular stresses, including exposure to dsRNA and accumulation of unfolded proteins [39] , activate pathways that lead to the phosphorylation of the translation initiation factor eIF-2alpha on serine 51. Phosphorylation of eIF-2alpha on serine 51 results in global, frequently transient, inhibition of protein synthesis [39] . Intracellular poly I:C, an analog of dsRNA employed experimentally to mimic the effects of viral dsRNA in cells, has been shown to activate the NLRP3 inflammasome in BMDM [6] . Indeed, exposure of BMDM to poly I:C induced the release of IL-1ß from primed BMDM as well as the phosphorylation of eIF-2alpha on serine 51 ( Fig. 5A ). The release of IL-1ß from primed BMDM induced by poly I:C was accompanied by inhibition of protein synthesis ( Fig. 5B ). 10.1371/journal.pone.0036044.g005 Figure 5 Inhibition of protein synthesis by dsRNA and inhibition of IL-1ß processing by MG-132. A) BMDM were treated with or without 4 h of LPS priming, as indicated. Cells were then rinsed in fresh medium and treated with either LipofectAMINE 2000 or LipofectAMINE 2000-poly I:C complex for 4 h, in the presence or absence of 30 µM MG-132, as indicated. Cell lysates (cell) or media (medium) samples were subjected to immunoblotting with the antibodies indicated. B) BMDM were treated with either LipofectAMINE 2000 alone or with LipofectAMINE 2000-dsRNA complex for the times indicated. Fifteen minutes before each time-point, 1 µCi of [ 3 H]-leucine was added, and leucine incorporation was terminated by trichloroacetic acid. Each treatment was conducted in triplicate wells, and values are shown as mean ± S.D. Percent incorporation of [ 3 H]-leucine at each point was calculated as the [ 3 H]-leucine incorporated into cells exposed to LipofectAMINE 2000-dsRNA complex/[ 3 H]-leucine incorporated into cells exposed to LipofectAMINE 2000 alone×100. Proteasome Inhibitors Block Activation of the NLRP3 Inflammasome Previously we showed that inhibitors of proteasome activity reduce the ricin-mediated activation of the NLRP3 inflammasome in LPS-primed BMDM [22] . Others have reported that proteasome inhibitors reduce the activity of the NLRP1 inflammasome but not the accumulation of pro-IL-1ß by anthrax lethal toxin [40] . To determine if proteasome inhibitors would block the ability of translation inhibitors to activate the NLRP3 inflammasome, we employed two proteasome inhibitors, MG-132 and bortezimib. IL-1ß release mediated by cycloheximide or ricin was strongly suppressed in LPS-primed BMDM co-treated with either bortezimib or MG-132 ( Fig. 6A ). Inclusion of MG-132 blocked the ability of each of the translation inhibitors to induce the processing of pro-IL-1ß and the release of IL-1ß from cells ( Fig. 6B ). MG-132 similarly suppressed the release of IL-1ß from cells exposed to MSU ( Fig. 6C ) and poly I:C ( Fig. 5A ), suggesting that proteasomes may participate more generally in activation of the NLRP3 inflammasome. 10.1371/journal.pone.0036044.g006 Figure 6 Proteasome inhibitors block processing and release of IL-1ß. A) LPS-primed WT BMDM were incubated in control medium or medium containing 10 ng/ml ricin or 25 µg/ml cycloheximide for 4 h. MG-132 (30 µM) or Bortezimib (0.5 µM) was included as indicated. Secreted IL-1ß was measured by ELISA in triplicate wells. B) LPS-primed WT BMDM were incubated in the presence or absence of MG-132 for 4 hours, in the presence or absence of inhibitors of protein synthesis, as indicated. Cell lysates (cell) and culture medium (medium) were examined by immunoblotting. C) LPS-primed or unprimed WT BMDM were or exposed to MSU, MG-132, or both for 4 h, as indicated. Cell lysates (cell) and culture medium (medium) were examined by immunoblotting. Activation of the NLRP3 Inflammasome by Inhibitors of Translation We tested the ability of a panel of well-characterized inhibitors of translation (cycloheximide, emetine, puromycin, pactamycin, and anisomycin) to stimulate IL-1ß release from BMDM via the NLRP3 inflammasome. When added at doses that suppress incorporation of [ 3 H]-leucine by at least 90%, each inhibitor led to release of IL-1ß from LPS-primed BMDM by 4 h ( Fig. 1A ). Conversion of pro-IL-1ß to IL-1ß induced by cycloheximide proceeded in a dose-dependent manner ( Fig. 1B ), suggesting that the degree of translational impairment correlated well with the amount of IL-1ß that was released. As expected, there was a reciprocal relationship between the amount of pro-IL-1ß present in the cell lysate ( Fig. 1A , upper panel) and the amount of IL-1ß detected in the culture medium by ELISA ( Fig. 1A , lower panel). Cells remained strongly adherent to culture dishes at the termination of the experiments, and did not show morphological signs of apoptosis or activation of caspase-3 (not shown). To gain insight into the mechanism of IL-1ß release triggered by translational inhibitors, we exposed BMDM from mice deficient in NLRP3, ASC, or caspase-1 to agents that inhibited translation by a variety of mechanisms and compared their responses to those of wild-type (WT) cells. When BMDM deficient in NLRP3, ASC, or caspase-1 were exposed to the inhibitors, the cells secreted substantially reduced amounts of IL-1ß compared with WT BMDM exposed to the same inhibitors ( Fig. 1C ), suggesting that processing and release of IL-1ß were mediated through formation of the NLRP3 inflammasome. 10.1371/journal.pone.0036044.g001 Figure 1 Inhibition of protein synthesis triggers secretion of IL-1ß from primed BMDM. A) WT BMDM were stimulated or not with 50 ng/ml LPS for 4 h followed by exposure to 10 µg/ml ricin, 25 µg/ml cycloheximide, 10 µg/ml emetine, 75 µg/ml puromycin, 0.2 µg/ml pactamycin, or 10 µg/ml anisomycin for 4 h prior to harvesting. Cell extracts were analyzed by immunoblotting for p38 MAPK (loading control) and pro-IL-1ß (upper panel). Media supernatants were analyzed by ELISA for released IL-1ß (lower panel). ELISA data are represented as mean ± SEM from triplicate dishes. B) LPS-primed WT BMDM were stimulated for 4 h with increasing doses of cycloheximide (from right to left: 0.03, 0.1, 0.3, 1.0, 3.0, 10, 30, and 100 µg/ml). Media supernatants were analyzed by ELISA for released IL-1ß (upper panel). Cell extracts were analyzed by immunoblotting for p38 MAPK (loading control) and pro-IL-1ß (lower panel). ELISA data are represented as mean ± SEM from triplicate dishes. C) Release of IL-1ß triggered by inhibitors of translation was determined by immunoblotting of cell extracts (cell) or culture medium (medium) from LPS-primed BMDM obtained from WT mice or mice deficient in NLRP3, ASC, or caspase-1, as indicated. P38 MAPK was loading control. Extracellular Potassium Suppresses Release of IL-1ß Induced by Inhibitors of Translation Previous studies have described a requirement for potassium efflux by a variety of agents reported to activate the NLRP3 inflammasome, including nigericin and ATP [11] . Here, substitution of 130 mM Na + /5 mM K + by 5 mM Na + /130 mM K + in the culture medium suppressed the appearance of mature IL-1ß by translation inhibitors to a similar extent as nigericin and ATP, as detected by ELISA ( Fig. 2A ). Immunoblotting of the high potassium culture medium revealed undetectable amounts of mature IL-1ß in the culture medium of LPS-primed cells exposed to nigericin or ATP ( Fig. 2B ), or ricin, emetine, or cycloheximide ( Fig. 2C ). 10.1371/journal.pone.0036044.g002 Figure 2 Effect of extracellular potassium on IL-1 processing and release. A) Bone marrow-derived macrophages were plated in triplicate wells in 12-well plates and primed with 50 ng/ml of LPS for 4 h. Cells were then incubated in medium containing 130 mM NaCl/5 mM KCl or 5 mM KCl/130 mM NaCl in the presence of absence of 0.01 µg/mL ricin, 25 µg/mL cycloheximide, 10 µg/mL, 10 µg/mL emetine, 75 µg/mL puromycin, 0.2 µg/mL pactamycin, 10 µg/mL anisomycin, 3.4 µM nigericin, or 5 mM ATP for 4 h. Medium was collected, and p17 IL-1 was determined by ELISA. B) Macrophages were plated, primed with LPS and incubated in medium containing 130 mM NaCl/5 mM KCl (MEM-Na) or 5 mM NaKCl/130 mM KCl (MEM-K) in the presence or absence 5 mM ATP or 3.4 µM nigericin for 4 h. Proteins were precipitated from the media with TCA and analyzed by Western blotting. C) Macrophages were plated, primed with LPS, and incubated in medium containing 130 mM NaCl/5 mM KCl (MEM-Na) or 5 mM NaKCl/130 mM KCl (MEM-K) in the presence or absence 10 ng/mL ricin, 10 µg/mL emetine, or 25 µg/mL cycloheximide for 4 h. Proteins were precipitated with TCA and analyzed by Western blotting. Lack of Correlation between Translation Inhibition and Potassium Efflux The ability of high extracellular potassium to suppress activation of the NLRP3 inflammasome by multiple proinflammatory agonists [11] has been used to support a model demonstrating that decreased cellular potassium constitutes a general trigger for inflammasome activation. Initial studies were conceived because nigericin and ATP, agents that are known to cause potassium efflux [7] , [24] , are also potent activators of the NLRP3 inflammasome. The conclusion that loss of cellular potassium is responsible for activation of the NLRP3 inflammasome was based on the ability of high extracellular potassium to suppress pro-IL-1ß processing and release of mature IL-1ß in the presence of high potassium [11] . Subsequent studies have employed high extracellular potassium to block activation of the NLRP3 by a variety of agents [13] , [25] , [26] , [27] , [28] , [29] , [30] . Although these studies demonstrated that high extracellular potassium blocks release of IL-1ß, they did not determine whether the proinflammatory agents employed actually cause a loss of cellular potassium. To determine whether inhibition of translation would trigger efflux of potassium from macrophages, we exposed LPS-primed BMDM to emetine, a potent inhibitor of translation [31] , [32] that irreversibly inhibits the elongation cycle of translation by greater than 99% within 1 minute [33] and produces release of IL-1ß within 4 h. Intracellular potassium was measured by inductively coupled plasma mass spectroscopy (ICP-MS). Nigericin and ATP, agents that have been shown to reduce intracellular potassium [7] , [24] , produced a 50% decrease in intracellular potassium concentration by 15 min and 45 min, respectively ( Fig. 3A ). BMDM exposed to nigericin and ATP detached from the culture dishes by 1 h. By contrast, cells exposed to emetine maintained normal levels of intracellular potassium for at least 4 h ( Fig. 3A ). Ricin and cycloheximide similarly failed to cause release of cellular potassium (not shown). These data suggest that in BMDM the inhibition of translation promoted the activation of the NLRP3 inflammasome but failed to induce leakage of cellular potassium in the interval of time (4 h) that preceded activation of the inflammasome and release of IL-1ß. 10.1371/journal.pone.0036044.g003 Figure 3 Inhibition of protein synthesis fails to elicit efflux of potassium from cells. A) Intracellular potassium in WT BMDM was analyzed by ICP-MS after priming with LPS for 4 h and exposure to emetine, ATP, and nigericin for indicated times. Data are represented as mean ± SEM from triplicate dishes. Measurements of potassium release by cells exposed to ATP and nigericin were terminated at 10 min and 60 min, respectively, as a result of cytotoxicity and detachment of cells from the dishes. Cells exposed to emetine showed no signs of cytotoxicity or detachment by 4 h. B) Elevated potassium blocks emetine-mediated release of IL-1ß and processing of caspase-1. LPS-primed BMDM were either exposed or not exposed to 10 µg/ml emetine for 4 h, at which time proteins were precipitated from the media with TCA and analyzed by Western blotting as shown. C) BMDM in triplicate wells were pulse-labeled in medium containing [ 3 H]-leucine for 15 min prior to harvest at the indicated times and the amount of [ 3 H]-leucine incorporation was measured. Four h after exposure of LPS-primed BMDM to emetine in medium containing low potassium, both IL-1ß and a proteolytic fragment of caspase-1 (p10) appeared in the medium ( Fig. 3B , 2nd lane). In the presence of elevated potassium, neither p10 capase-1 nor IL-1ß appeared in the medium of cells exposed to emetine ( Fig. 3B , 4th lane), suggesting that high potassium blocked caspase-1-mediated processing of pro-IL-1ß. Coupling between Potassium Efflux and Inhibition of Protein Translation In view of the differential requirement for potassium efflux in promoting NLRP3 activation, we sought to determine whether nigericin, ATP, and panel of translation inhibitors that we employed shared any mechanistic features. It has been well established that the rate of protein synthesis is directly dependent on the concentration of intracellular potassium [18] , [34] and that potassium ionophores inhibit translation at the same doses and kinetics as potassium efflux [35] . To test the ability of nigericin and ATP to suppress protein synthesis in BMDM, we subjected cells to nigericin or ATP and measured incorporation of [ 3 H]-leucine. Exposure of BMDM to either nigericin or ATP triggered rapid inhibition of protein synthesis by 30 min in normal medium (130 mM Na + /5 mM K + ; Fig. 3C ). Replacement of extracellular Na + by K + (130 mM K + /5 mM Na + ) partially reversed the blockade of protein synthesis in ATP-treated cells. Replacement of extracellular Na + by K + in nigericin-treated cells restored protein synthesis to a level even greater than that observed in untreated cells ( Fig. 3C ). These data are consistent with the known association of decreased cellular potassium with the inhibition of protein synthesis by nigericin, ATP, and other potassium ionophores [34] , [35] , [36] . Monosodium Urate Crystals Suppress Protein Translation Gout is a chronic inflammatory response that is associated with the formation of crystals of monosodium urate (MSU). Incubation of LPS-primed murine macrophages with MSU crystals leads to the conversion and release of IL-1ß by a process that requires activation of the NLRP3 inflammasome. Because increased extracellular potassium blocks the activation of the NLRP3 inflammasome by MSU, it has been concluded that MSU acts through a mechanism that involves a decreased concentration of cellular potassium [11] , [37] , [38] . To determine if MSU inhibits protein synthesis at doses that stimulate release of IL-1ß from BMDM, we exposed cells to increasing doses of MSU or monopotassium urate (MPU), a salt of uric acid that is ineffective in promoting IL-1ß release [38] . The secretion of mature IL-1ß induced by MSU crystals was associated in a dose-dependent manner with inhibition of [ 3 H]-leucine incorporation into BMDM ( Fig. 4A,B ). Monopotassium urate (MPU), which, as expected, failed to stimulate conversion of pro-IL-1ß to mature IL-1ß ( Fig. 4B ), also failed to inhibit protein synthesis as measured by incorporation of [ 3 H]-leucine ( Fig. 4A ). These results demonstrated that MSU-induced suppression of protein synthesis in macrophages correlated with the release of IL-1ß by MSU. 10.1371/journal.pone.0036044.g004 Figure 4 MSU crystals inhibit protein synthesis at concentrations that induce processing and release of IL-1ß. WT BMDM were primed with LPS for 4 h prior to exposure MSU or MPU at indicated concentrations. A) BMDM in triplicate wells were pulse-labeled in medium containing [ 3 H]-leucine for 15 min prior to harvest at the indicated times and the amount of [ 3 H]-leucine incorporation was measured. B) Cells were harvested 4 h after addition of indicated concentrations of MSU and MPU. Cell lysates (cell) and culture medium (medium) were examined by immunoblotting. P38 MAPK was loading control. Double-stranded RNA Inhibits Protein Translation and Activates the NLRP3 Inflammasome Many forms of cellular stresses, including exposure to dsRNA and accumulation of unfolded proteins [39] , activate pathways that lead to the phosphorylation of the translation initiation factor eIF-2alpha on serine 51. Phosphorylation of eIF-2alpha on serine 51 results in global, frequently transient, inhibition of protein synthesis [39] . Intracellular poly I:C, an analog of dsRNA employed experimentally to mimic the effects of viral dsRNA in cells, has been shown to activate the NLRP3 inflammasome in BMDM [6] . Indeed, exposure of BMDM to poly I:C induced the release of IL-1ß from primed BMDM as well as the phosphorylation of eIF-2alpha on serine 51 ( Fig. 5A ). The release of IL-1ß from primed BMDM induced by poly I:C was accompanied by inhibition of protein synthesis ( Fig. 5B ). 10.1371/journal.pone.0036044.g005 Figure 5 Inhibition of protein synthesis by dsRNA and inhibition of IL-1ß processing by MG-132. A) BMDM were treated with or without 4 h of LPS priming, as indicated. Cells were then rinsed in fresh medium and treated with either LipofectAMINE 2000 or LipofectAMINE 2000-poly I:C complex for 4 h, in the presence or absence of 30 µM MG-132, as indicated. Cell lysates (cell) or media (medium) samples were subjected to immunoblotting with the antibodies indicated. B) BMDM were treated with either LipofectAMINE 2000 alone or with LipofectAMINE 2000-dsRNA complex for the times indicated. Fifteen minutes before each time-point, 1 µCi of [ 3 H]-leucine was added, and leucine incorporation was terminated by trichloroacetic acid. Each treatment was conducted in triplicate wells, and values are shown as mean ± S.D. Percent incorporation of [ 3 H]-leucine at each point was calculated as the [ 3 H]-leucine incorporated into cells exposed to LipofectAMINE 2000-dsRNA complex/[ 3 H]-leucine incorporated into cells exposed to LipofectAMINE 2000 alone×100. Proteasome Inhibitors Block Activation of the NLRP3 Inflammasome Previously we showed that inhibitors of proteasome activity reduce the ricin-mediated activation of the NLRP3 inflammasome in LPS-primed BMDM [22] . Others have reported that proteasome inhibitors reduce the activity of the NLRP1 inflammasome but not the accumulation of pro-IL-1ß by anthrax lethal toxin [40] . To determine if proteasome inhibitors would block the ability of translation inhibitors to activate the NLRP3 inflammasome, we employed two proteasome inhibitors, MG-132 and bortezimib. IL-1ß release mediated by cycloheximide or ricin was strongly suppressed in LPS-primed BMDM co-treated with either bortezimib or MG-132 ( Fig. 6A ). Inclusion of MG-132 blocked the ability of each of the translation inhibitors to induce the processing of pro-IL-1ß and the release of IL-1ß from cells ( Fig. 6B ). MG-132 similarly suppressed the release of IL-1ß from cells exposed to MSU ( Fig. 6C ) and poly I:C ( Fig. 5A ), suggesting that proteasomes may participate more generally in activation of the NLRP3 inflammasome. 10.1371/journal.pone.0036044.g006 Figure 6 Proteasome inhibitors block processing and release of IL-1ß. A) LPS-primed WT BMDM were incubated in control medium or medium containing 10 ng/ml ricin or 25 µg/ml cycloheximide for 4 h. MG-132 (30 µM) or Bortezimib (0.5 µM) was included as indicated. Secreted IL-1ß was measured by ELISA in triplicate wells. B) LPS-primed WT BMDM were incubated in the presence or absence of MG-132 for 4 hours, in the presence or absence of inhibitors of protein synthesis, as indicated. Cell lysates (cell) and culture medium (medium) were examined by immunoblotting. C) LPS-primed or unprimed WT BMDM were or exposed to MSU, MG-132, or both for 4 h, as indicated. Cell lysates (cell) and culture medium (medium) were examined by immunoblotting. Discussion Ricin is a potent ribosomal toxin considered to be a potential bioterrorist agent due to the possibility of aerosol delivery to human populations [41] . Studies in animals demonstrated that delivery of ricin to the pulmonary system leads to acute lung injury and symptoms resembling acute respiratory distress syndrome [42] , [43] , [44] , [45] . Previously we reported that the inflammatory and lethal effects of inhaled ricin were prevented in mice with a genetic deficiency in IL-1ß. Lung inflammation was blocked in mice that had been co-treated with IL-1 receptor antagonist (IL1RA; anakinra) or depleted of macrophages, suggesting the primacy of macrophage-derived IL-1ß in orchestrating pulmonary inflammatory responses to ricin [22] . Exposure of LPS-primed BMDM to ricin in vitro induced the processing pro-IL-1ß to mature IL-1ß in a NLRP3 dependent manner [23] , raising the possibility that the toxic action of ricin on ribosomal function was responsible for activation of the NLRP3 inflammasome. These results prompted us to examine the role that suppression of ribosomal activity plays in facilitating initiation of inflammatory signaling by IL-1ß. Indeed, the results presented here demonstrate that a panel of translation inhibitors, acting through different mechanisms on peptide initiation or elongation, promoted the conversion and release of IL-1ß by a process that required participation of the NLRP3 inflammasome ( Fig. 1 ). These data led us to question whether inhibition of translation, which occurs as a result of a variety of cellular stresses, might serve as a relevant trigger for inflammasome activation in human inflammatory disease. It has been proposed that decreased levels of intracellular potassium act to trigger release of IL-1ß as a result of activation of the NLRP3 inflammasome [11] . It is well documented that reduced concentrations of intracellular potassium also fail to support protein synthesis in mammalian cells [20] , [21] . Potassium ionophores such as nigericin and valinomycin and compounds that poison the membrane-associated Na + /K + ATPase rapidly trigger translational inhibition in cells concomitant with efflux of cellular potassium, and application of high extracellular potassium is sufficient to rescue ribosomal activity [21] , [36] , [46] , [47] . Replacement of extracellular sodium with potassium suppresses the ability of nigericin and ATP to activate the NLRP3 inflammasome, supporting the notion that low intracellular potassium serves as a trigger for inflammasome activation [11] . Our results demonstrating the rescue of protein translation by increased extracellular potassium in BMDM exposed to nigericin or ATP ( Fig. 3C ) suggested to us that impairment of ribosomal activity may explain why loss of cellular potassium activates the NLRP3 inflammasome. However, the inability of emetine, a potent and irreversible inhibitor of protein synthesis, to promote efflux of potassium ( Fig. 3A ) demonstrated that an intracellular environment of low potassium was not required for activation of the NLRP3 inflammasome by molecules that directly interfere with ribosomal function. It has been concluded that MSU crystals activate the NLRP3 inflammasome [48] through a process that leads to decreased concentration of cellular potassium [11] , [38] . The conclusion that decreased concentration of cellular results from a leakage of potassium from cells after MSU treatment was based on the ability of high extracellular potassium (150 mM) to block processing of pro-IL-1ß to its active form [11] . Exposure of BMDM to MSU results in the engulfment of the insoluble crystals within the acidic milieu of endosomes [38] , resulting in solubilization of the crystals and the subsequent rapid increase in cellular volume caused by release of sodium ions. The increase in cellular volume is thought to be responsible for a drop in concentration of intracellular potassium by dilution [38] . Our results showing that exposure of BMDM to MSU, but not MPU, results in the dose-dependent inhibition of protein translation and release of IL-1ß ( Fig. 2 ), is consistent with the latter model and further supports the close association between inhibition of protein translation and activation of the NLRP3 inflammasome. In view of our data showing that inhibition of translation fails to mediate loss of cellular potassium ( Fig. 3 ), we were surprised that high extracellular potassium was able to block appearance of IL-1ß in the medium. Interestingly, Arlehamn et al. reported that high extracellular potassium inhibited IL-1ß release from cells after bacterial infection with P. aeruginosa and S typhimurium , which was dependent on the NLRC4 inflammasome, but that potassium leakage from cells could not be detected by flame photometry [49] . They conjectured that a minority of cells had undergone a loss of potassium due to the nature of the pathogens, which did not infect every cell, and for this reason they could not observe measurable potassium loss from the population of cells. Alternatively, high extracellular potassium may block the release of IL-1ß by a mechanism that is independent of intracellular potassium concentration. For example, high potassium (150 mM) has been shown to suppress activation and cleavage of recombinant caspase-1 in vitro [50] . Petrilli et al reported that MSU-treated primed macrophages release pro-IL-1ß and procaspase-1 into the medium, and that cells exposed to high extracellular potassium release pro-IL-ß and procaspase-1 into the high potassium medium, but that proteolytic processing of these proteins failed to occur [11] . These data and our data ( Fig. 3B ) suggest that high extracellular potassium may directly suppress the cleavage of pro-IL-1ß after externalization of the inflammasome complex by inhibiting the activation of procaspase-1. Our data suggest that cells in high potassium may suppress the activation of the NLRP3 inflammasome by two independent mechanisms: 1) by restoring intracellular potassium to normal levels in cells that have undergone leakage of potassium via pore formation (e.g. by nigericin) or stimulation of P2X7 receptors (e.g. by ATP), thereby preventing translational inhibition; and 2) by suppressing the activation of caspase-1 by an unknown mechanism. The generation of reactive oxygen species (ROS) is commonly associated with NLRP3 inflammasome activation in response to a variety of agonists, including the mitochondrial inhibitors, rotenone and antimycin A [8] , [11] , [12] , [13] . However, the conclusion that generation of ROS is responsible for inflammasome activation has been questioned [14] , [15] , [16] . Inhibition of mitochondrial Complex I or Complex III following exposure of cells to rotenone or antimycin A, respectively, leads to the generation of ROS through loss of mitochondrial membrane potential [51] , [52] , [53] . Rotenone and antimycin A have been shown to activate the NLRP3 inflammasome, presumably as a result of ROS production, since treatment of macrophages with Mito-Tempo, a scavenger of mitochondrial ROS, inhibited inflammasome activation [54] . Uncouplers of mitochondrial function such as rotenone and antimycin A are also potent inhibitors of translation, reducing the protein synthetic rate by more than 90% at concentrations employed to generate ROS in cultured cells [36] , [55] . Several reports have demonstrated that ROS can inhibit mRNA translation [56] , [57] , [58] . Although the mechanism by which mitochondrial inhibitors inhibit protein synthesis is incompletely understood, it has been shown recently that mitochondrial inhibitors suppress protein synthesis by inducing the rapid phosphorylation of both eIF-2alpha and the elongation factor eEF2, presumably by stimulating PERK [59] . Indeed, peroxide- and hypoxia-mediated ROS have been shown to inhibit translation by increasing PERK- and PKR-mediated phosphorylation of eIF-2alpha and eEF2 [60] . Our data demonstrating that inhibition of translation can activate the NLRP3 inflammasome is consistent with the notion that the generation of ROS by mitochondrial dysfunction may activate the NLRP3 inflammasome by suppressing protein synthesis through stress-activated phosphorylation of eIF-2alpha and eEF2. Inhibition of translation occurs in a variety of circumstances in nature, triggered by exposure to toxins, pathogens that co-opt host cell machinery, hypoxia, and sterile inflammatory signals released from damaged tissues. Phosphorylation of eIF-2alpha at Ser51 mediates translational inhibition in response to cellular signals [61] , [62] by preventing the formation of the eIF2/GTP/Met-tRNA complex [63] . Stress-induced inhibition of translation through phosphorylation of eIF-2alpha is induced by viral dsRNA through activation of protein kinase R (PKR); hypoxia through the PRK-like endoplasmic reticulum kinase (PERK); and by glucose deprivation through activation of both PKR and PERK [39] , [64] , [65] , [66] . An important consequence of eIF-2alpha phosphorylation is the regulation of gene expression, as mutations that interfere with eIF-2alpha phosphorylation lead to defective expression of stress-induced genes [67] . Recent evidence suggests that PKR acts as a central integrator in the inflammatory component of metabolic control by linking nutrient- and pathogen-sensing pathways in development of insulin resistance, type 2 diabetes, and other chronic metabolic pathologies [68] . Poly I:C-mediated activation of the NLRP3 inflammasome has been previously reported [6] . In LPS-primed BMDM we found that poly I:C mediates eIF-2alpha phosphorylation, inhibition of protein synthesis, and the release of IL-1ß ( Figure 5 ). Phosphorylation of eIF-2alpha is required not only for attenuation of translation, but also for transcriptional induction and survival in response to endoplasmic reticulum-mediated stress (ER stress) [67] . ER stress activates the NLRP3 inflammasome via a pathway that does not involve the unfolded protein response [27] . Repression of translation through phosphorylation of eIF-2alpha leads to activation of NF-kappaB and the subsequent transcription of NF-kappaB-directed genes by promoting the turnover of the labile inhibitor, IkappaB alpha protein [69] . Stress-induced translational inhibition by phosphorylated eIF-2alpha may contribute to inflammatory responses by simultaneously promoting the two necessary events required to produce IL-1ß: the NF-kappaB-mediated synthesis of pro-IL-1ß and the release of IL-1ß through activation of the NLRP3 inflammasome. This model could explain how stress signals that converge on eIF-2alpha could induce IL-1ß-dependent inflammatory responses. Maintenance of the intracellular level of proteins that exhibit short half-lives, such as p53 and IkappaB, is frequently regulated by the balance between their rate of synthesis and proteasome-directed degradation [70] , [71] , [72] . For example, inhibition of protein translation by stress-induced phosphorylation of eIF-2alpha leads to activation of NF-kappaB through proteasome-dependent degradation of IkappaB [69] , [73] . Our experiments determined that co-treatment of BMDM with proteasome inhibitors plus ricin, cycloheximide, puromycin, pactamycin, anisomycin, MSU, or dsRNA led to a reduction or complete suppression of IL-1ß release as measured by ELISA ( Figure 4A ) or immunoblot ( Figures 4B , 5 ). LPS-primed cells treated with poly I:C in the presence of MG-132 exhibited a reduction in IL-1ß release but did not change the phosphorylation status of eIF-2alpha ( Fig. 5 ), suggesting that involvement of the proteasome in activating the inflammasome is positioned downstream of translational inhibition in these cells. In light of our data showing a link between suppression of protein synthesis and activation of the NLPR3 inflammasome, we propose that labile protein(s) may suppress the formation of the NLRP3 inflammasome. In such a scenario, inhibition of translation, which accompanies many types of cellular stresses, would lead to a decrease in abundance of putative repressor protein(s), perhaps through proteasome-mediated degradation. In this model, blockade of proteasomal activity would extend the lifetime of the putative repressor protein(s) . The validity of this model would require identification of labile protein(s) that inhibit the processing of pro-IL-1ß. The current study demonstrates that suppression of ribosomal function by molecules acting by disparate mechanisms is sufficient to activate the NLRP3 inflammasome. These data suggest that inhibition of translation may constitute a common stimulus by which seemingly dissimilar activators promote the processing and release of IL-1ß. A decreased rate of translation may constitute a regulatory node that integrates signals from toxins, pathogens, and metabolic disturbances, thereby enhancing systemic inflammation by promoting the processing and release of IL-1ß. The suppression of IL-1ß release by proteasome inhibitors suggests that labile protein(s) may be responsible for blocking the activation of the NLRP3 inflammasome under non-stressed conditions. A decrease in translation rate may lead to reduction in cellular levels of these protein(s), thereby leading to formation of active NLRP3 inflammasomes. Further studies that focus on identification of labile inhibitors of inflammasome function are clearly necessary to test the validity of this hypothesis. A graphic depiction of the proposed mechanisms underlying NLRP3 activation by different stimuli is shown in Figure 7 . 10.1371/journal.pone.0036044.g007 Figure 7 Graphic depiction of proposed mechanisms by which different stimuli activate the NLRP3 inflammasome. Inhibition of protein synthesis leads to activation of the NLRP3 inflammasome by preventing the synthesis of short-lived inhibitory protein(s) that are degraded by proteasomes. Inhibition of protein synthesis is mediated by a variety of physiological cell stressors acting through phosphorylation of eIF-2alpha, leading to a transient decrease in protein synthesis. Toxins such as ricin, Shiga toxins, and sarcin inhibit protein synthesis by interfering with the structure of the sarcin/ricin loop of 28S rRNA. Antibiotics act through various ribosome-associated mechanisms to inhibit translational initiation and/or elongation. Potassium ionophores (and receptors such as P2X7) inhibit translation by mediating loss of cellular potassium. Monosodium urate inhibits translation by inducing cell swelling, which leads to dilution of intracellular potassium. Materials and Methods Ethics Statement This study was approved by the Institutional Animal Care & Use Committee (IACUC) and the Institutional Biosafety Committee of OHSU. The authors have conformed to the policies of the Institutional Office of Integrity of OHSU. Mice are anesthetized with ketamine/xylazine. To minimize pain and discomfort, mice are euthanized by cervical dislocation as per the recommendation of the Panel of Euthanasia of the American Veterinary Medical Association. Only personnel trained in animal experimentation are involved. All procedures that involve the use of ricin and animals have been approved by the Chemical Safety Officer at OHSU and by the Institutional Animal Care Utilization Committee (IACUC) at OHSU. Personnel are gloved and wear protective coverings over their body, including full face respirator masks. All animal studies are conducted in chemical safety hoods. Animals are housed in designated areas as approved by the IACUC under protocol A900, OHSU. Reagents and Antibodies Cycloheximide, emetine, anisomycin, and puromycin were purchased from Sigma-Aldrich (St. Louis). Ricin was purchased from Vector Laboratories (Burlingame, CA). Polyinosinic-cytidylic acid (Poly I:C) was purchased from Midland Certified Reagents (Midland, Texas). Lipofectamine 2000 was purchased from Life Technologies (Carlsbad, CA). Bortezimib was purchased from LC Laboratories (Woburn, MA). MG-132 was purchased from EMD Biosciences (Gibbstown, NJ). Pactamycin was a gift from the Upjohn Company (Kalamazoo, MI). Anti-IL-1ß was purchased from Abcam (Cambridge, MA) and anti-p38 MAPK was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-phospho-eIF-2alpha (ser51) #9721 was purchased from Cell Signaling (Danvers, MA). The mouse IL-1ß enzyme-linked immunosorbent assay (ELISA) Ready-Set-Go was purchased from eBioscience (San Diego, CA). Crystals of MSU and MPU were prepared as described [74] . Animals and Animal Procedures All animal procedures were performed according to protocols that have been approved by the Institutional Animal Care and Use Committee at Oregon Health and Science University, Portland, Oregon. C57BL/6J and caspase-1 deficient mice were purchased from the Jackson Laboratory (Bar Harbor, ME). ASC- and NLRP3-deficient mice were kindly provided by V. Dixit (Genentech, San Francisco, CA). Male mice, 8–10 weeks of age, were used throughout the experiments. Isolation of Bone Marrow- Derived Macrophages (BMDM) Male mice, 8 to 10 weeks of age, were used throughout the experiments. Cells were prepared from WT C57BL/6J, ASC-, Caspase-1-, and NLRP3-deficent mice. Marrow was flushed from femurs and tibias with PBS and cultured in alpha-Minimum Essential Medium (αMEM, Cellgro, Herndon, VA), supplied with 10% Fetal Bovine Serum (FBS, Cellgro, Herndon, VA), 50 µg/ml gentamicin, and 100 ng/ml recombinant mouse Colony Stimulating Factor 1 (CSF-1, R&D Systems, Minneapolis, MN) for 72 hrs on non-tissue culture treated 10-cm Petri dishes. BMDM were passaged and cultured for an additional 72 h. Each confluent 10-cm dish was transferred into one 6-well or one 12-well tissue culture plate and cultured for 24 hrs before initiating experimental treatment. Treating BMDM Cells were serum deprived in alpha-MEM for 30 minutes followed by treatment for 4 hours in the presence or absence of 50 ng/mL LPS. Cells were then rinsed and fresh media was added followed by exposure to the indicated doses of inhibitors (or solvent control), in the absence of LPS, for 4 hrs. In experiments employing elevated potassium, sodium was replaced by potassium at an equivalent molar concentration. Immunoblotting BMDM cells were lysed in 2×ESB lysis buffer in preparation for immunoblotting. Equal volumes of the cell lysates were separated on a 10% denaturing polyacrylamide gel in the presence of sodium dodecyl sulfate and were transferred onto polyvinylidene difluoride membranes according to standard laboratory procedures. For detection of IL-1ß released from cells, proteins from media supernatants were precipitated using TCA plus 200 µg insulin carrier protein and separated on 13% gels. Membranes were incubated with the indicated antibodies and the corresponding horseradish peroxidase-conjugated secondary antibodies. Signals were detected by using enhanced chemiluminescence. ELISA Media from BMDM were collected and analyzed in triplicate using IL-1ß ELISA (eBioscience) according to the manufacturer's protocol. ICP-MS Primed BMDM were exposed to inhibitors of protein synthesis for indicated times up to 4 h and cells were digested overnight in culture dishes in 1 mL 10% HNO 3 . Digests were analyzed by inductively coupled plasma mass spectrometry (ICP-MS) in the Metal Ion Core by Dr. Martina Ralle in the Department of Biochemistry and Molecular Biology at OHSU. ICP-MS analysis was performed using an Agilent 7700× system equipped with an ASX-250 Autosampler. The system was operated at a radio frequency power of 1,550 W, an argon flow rate of 15 L/min, carrier gas flow rate of 1.04 L/min, and helium (He) gas flow rate of 4.3 ml/min (only in He mode). Data were quantified using a seven-point (0, 1, 10, 100, 1,000, 2,000, and 5,000 ppb (ng/g)) calibration curve with an external standard for potassium. All data were acquired in He mode to remove interference from oxides, argides, and chlorides. For each sample, data were acquired in triplicate and averaged. Internal standards introduced with the sample were used to correct for plasma instabilities. A National Institute of Standards and Technology standard reference material (SRM 1577c) was used to ensure elemental recovery of >90%. Measurement of Protein Synthesis via Incorporation of [ 3 H]-leucine BMDM were cultured in 24-well tissue culture plates. Treatments were performed in leucine-free/serum-free Dulbecco modified Eagle medium, for the indicated times. Prior to harvesting, the cells were pulse-labeled with 1 µCi of [ 3 H]-leucine in 1 ml of leucine-free DMEM for times specified in the figure legends. Ten percent trichloroacetic acid was added to terminate incorporation. Wells were washed in water and 88% formic acid was added to solubilize the trichloroacetic acid-insoluble proteins. The samples were counted in a liquid scintillation counter. In each experiment, triplicate wells were used per experimental point. Statistical Analysis Individual groups were compared using unpaired t test analysis and were interpreted in a two-tailed manner. Ethics Statement This study was approved by the Institutional Animal Care & Use Committee (IACUC) and the Institutional Biosafety Committee of OHSU. The authors have conformed to the policies of the Institutional Office of Integrity of OHSU. Mice are anesthetized with ketamine/xylazine. To minimize pain and discomfort, mice are euthanized by cervical dislocation as per the recommendation of the Panel of Euthanasia of the American Veterinary Medical Association. Only personnel trained in animal experimentation are involved. All procedures that involve the use of ricin and animals have been approved by the Chemical Safety Officer at OHSU and by the Institutional Animal Care Utilization Committee (IACUC) at OHSU. Personnel are gloved and wear protective coverings over their body, including full face respirator masks. All animal studies are conducted in chemical safety hoods. Animals are housed in designated areas as approved by the IACUC under protocol A900, OHSU. Reagents and Antibodies Cycloheximide, emetine, anisomycin, and puromycin were purchased from Sigma-Aldrich (St. Louis). Ricin was purchased from Vector Laboratories (Burlingame, CA). Polyinosinic-cytidylic acid (Poly I:C) was purchased from Midland Certified Reagents (Midland, Texas). Lipofectamine 2000 was purchased from Life Technologies (Carlsbad, CA). Bortezimib was purchased from LC Laboratories (Woburn, MA). MG-132 was purchased from EMD Biosciences (Gibbstown, NJ). Pactamycin was a gift from the Upjohn Company (Kalamazoo, MI). Anti-IL-1ß was purchased from Abcam (Cambridge, MA) and anti-p38 MAPK was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-phospho-eIF-2alpha (ser51) #9721 was purchased from Cell Signaling (Danvers, MA). The mouse IL-1ß enzyme-linked immunosorbent assay (ELISA) Ready-Set-Go was purchased from eBioscience (San Diego, CA). Crystals of MSU and MPU were prepared as described [74] . Animals and Animal Procedures All animal procedures were performed according to protocols that have been approved by the Institutional Animal Care and Use Committee at Oregon Health and Science University, Portland, Oregon. C57BL/6J and caspase-1 deficient mice were purchased from the Jackson Laboratory (Bar Harbor, ME). ASC- and NLRP3-deficient mice were kindly provided by V. Dixit (Genentech, San Francisco, CA). Male mice, 8–10 weeks of age, were used throughout the experiments. Isolation of Bone Marrow- Derived Macrophages (BMDM) Male mice, 8 to 10 weeks of age, were used throughout the experiments. Cells were prepared from WT C57BL/6J, ASC-, Caspase-1-, and NLRP3-deficent mice. Marrow was flushed from femurs and tibias with PBS and cultured in alpha-Minimum Essential Medium (αMEM, Cellgro, Herndon, VA), supplied with 10% Fetal Bovine Serum (FBS, Cellgro, Herndon, VA), 50 µg/ml gentamicin, and 100 ng/ml recombinant mouse Colony Stimulating Factor 1 (CSF-1, R&D Systems, Minneapolis, MN) for 72 hrs on non-tissue culture treated 10-cm Petri dishes. BMDM were passaged and cultured for an additional 72 h. Each confluent 10-cm dish was transferred into one 6-well or one 12-well tissue culture plate and cultured for 24 hrs before initiating experimental treatment. Treating BMDM Cells were serum deprived in alpha-MEM for 30 minutes followed by treatment for 4 hours in the presence or absence of 50 ng/mL LPS. Cells were then rinsed and fresh media was added followed by exposure to the indicated doses of inhibitors (or solvent control), in the absence of LPS, for 4 hrs. In experiments employing elevated potassium, sodium was replaced by potassium at an equivalent molar concentration. Immunoblotting BMDM cells were lysed in 2×ESB lysis buffer in preparation for immunoblotting. Equal volumes of the cell lysates were separated on a 10% denaturing polyacrylamide gel in the presence of sodium dodecyl sulfate and were transferred onto polyvinylidene difluoride membranes according to standard laboratory procedures. For detection of IL-1ß released from cells, proteins from media supernatants were precipitated using TCA plus 200 µg insulin carrier protein and separated on 13% gels. Membranes were incubated with the indicated antibodies and the corresponding horseradish peroxidase-conjugated secondary antibodies. Signals were detected by using enhanced chemiluminescence. ELISA Media from BMDM were collected and analyzed in triplicate using IL-1ß ELISA (eBioscience) according to the manufacturer's protocol. ICP-MS Primed BMDM were exposed to inhibitors of protein synthesis for indicated times up to 4 h and cells were digested overnight in culture dishes in 1 mL 10% HNO 3 . Digests were analyzed by inductively coupled plasma mass spectrometry (ICP-MS) in the Metal Ion Core by Dr. Martina Ralle in the Department of Biochemistry and Molecular Biology at OHSU. ICP-MS analysis was performed using an Agilent 7700× system equipped with an ASX-250 Autosampler. The system was operated at a radio frequency power of 1,550 W, an argon flow rate of 15 L/min, carrier gas flow rate of 1.04 L/min, and helium (He) gas flow rate of 4.3 ml/min (only in He mode). Data were quantified using a seven-point (0, 1, 10, 100, 1,000, 2,000, and 5,000 ppb (ng/g)) calibration curve with an external standard for potassium. All data were acquired in He mode to remove interference from oxides, argides, and chlorides. For each sample, data were acquired in triplicate and averaged. Internal standards introduced with the sample were used to correct for plasma instabilities. A National Institute of Standards and Technology standard reference material (SRM 1577c) was used to ensure elemental recovery of >90%. Measurement of Protein Synthesis via Incorporation of [ 3 H]-leucine BMDM were cultured in 24-well tissue culture plates. Treatments were performed in leucine-free/serum-free Dulbecco modified Eagle medium, for the indicated times. Prior to harvesting, the cells were pulse-labeled with 1 µCi of [ 3 H]-leucine in 1 ml of leucine-free DMEM for times specified in the figure legends. Ten percent trichloroacetic acid was added to terminate incorporation. Wells were washed in water and 88% formic acid was added to solubilize the trichloroacetic acid-insoluble proteins. The samples were counted in a liquid scintillation counter. In each experiment, triplicate wells were used per experimental point. Statistical Analysis Individual groups were compared using unpaired t test analysis and were interpreted in a two-tailed manner.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7111298/

5-Ene-4-thiazolidinones – An efficient tool in medicinal chemistry

The presented review is an attempt to summarize a huge volume of data on 5-ene-4-thiazolidinones being a widely studied class of small molecules used in modern organic and medicinal chemistry. The manuscript covers approaches to the synthesis of 5-ene-4-thiazolidinone derivatives: modification of the C5 position of the basic core; synthesis of the target compounds in the one-pot or multistage reactions or transformation of other related heterocycles. The most prominent pharmacological profiles of 5-ene derivatives of different 4-thiazolidinone subtypes belonging to hit-, lead-compounds, drug-candidates and drugs as well as the most studied targets have been discussed. Currently target compounds (especially 5-en-rhodanines) are assigned as frequent hitters or pan-assay interference compounds (PAINS) within high-throughput screening campaigns. Nevertheless, the crucial impact of the presence/nature of C5 substituent (namely 5-ene) on the pharmacological effects of 5-ene-4-thiazolidinones was confirmed by the numerous listed findings from the original articles. The main directions for active 5-ene-4-thiazolidinones optimization have been shown: i ) complication of the fragment in the C5 position; ii ) introduction of the substituents in the N3 position (especially fragments with carboxylic group or its derivatives); iii ) annealing in complex heterocyclic systems; iv ) combination with other pharmacologically attractive fragments within hybrid pharmacophore approach. Moreover, the utilization of 5-ene-4-thiazolidinones in the synthesis of complex compounds with potent pharmacological application is described. The chemical transformations cover mainly the reactions which involve the exocyclic double bond in C5 position of the main core and correspond to the abovementioned direction of the 5-ene-4-thiazolidinone modification. 1 Introduction 4-Thiazolidinones and related heterocyclic based compounds have been extensively explored as the source of antiinflammatory, antitumor, antimicrobial, antidiabetic, antibacterial agents. The findings in the medicinal chemistry and pharmacology of 4-thiazolidinones have significantly increased since [1] , [2] , [3] 60 s being reflected in the rapid growth of the number of scientific papers [4] , numerous reviews and patents covering various 4-thiazolidinone derivatives [5] , [6] , [7] , [8] , [9] , [10] . The papers are dedicated to the selected 4-thiazolidinones subtypes, namely 2-(imino)amino-4-thiazolidinones [11] , 4-thiazolidinones with the exocyclic C=C double bond at the C(2)-position etc [12] . This proves that 4-thiazolidinones belong to the privileged scaffolds in the modern medicinal chemistry [4] , [13] , [14] . Combination of several reaction centers in the structure of 4-thiazolidinone derivates makes them an effective tool for the rational diversity oriented synthesis or privileged substructure-based diversity oriented synthesis [15] , [16] , [17] for the new lead-compounds creation. Major achievements in the 4-thiazolidinone field are related to the 2,4-thiazolidinedione, rhodanine (2-thioxo-4-thiazolidinone), 2-alkyl(aryl)-substituted, and 2-R-amino(imino)-substituted 4-thiazolidinone subtypes as sources of antimicrobial, antidiabetic, anti-inflammatory and anticancer lead-compounds and drug-candidates [7] , [10] , [11] . This is largely due to the investigation and introduction into the medical practice the antidiabetic drugs: glitazones (peroxisome proliferator-activated receptor-γ (PPAR) agonists, e.g. Rosiglitazone, Troglitazone, Pioglitazone, etc.) [18] , [19] and aldose reductase inhibitor Epalrestat [20] , and the simplicity in the synthesis of abovementioned compounds. Very less attention has been paid to the 4-thio- and 4-amino(imino)-derivatives (isorhodanine (4-thioxo-2-thiazolidinone), thiorhodanine (2,4-thiazolidinedithione), and especially 4-amino(imino)-2-thiazolidinone derivatives) [21] , [22] , [23] , [24] , [25] . Despite foregoing, among all 4-thiazolidinone subtypes 5 substituted thiazolidinones, namely 5-ene (5-ylidene) derivatives are of special interest in the context of chemical features and pharmacological profiles [6] , [8] , [26] , [27] . In this manuscript we tried to present the powerful pharmacological and chemical potential of 5-ene-4-thiazolidinones. Pursuing this goal we had not devided thiazolidinones according to the nature of their substituents (e.g. derivatives of 2,4-thiazolidinedione, rhodanine etc.) ( Fig. 1 ). Fig. 1 Structure of the target 5-ene-4-thiazolidinones. Fig. 1 The arguments in this favour were the following: i ) large number of reviews is devoted to the selected thiazolidinones subtypes; ii ) generally 5-ene-derivatives of different thiazolidinone cores are characterized by the same pharmacological profiles (see below) as well as the similar synthetic protocols. 2 Synthesis of 5-ene-4-thiazolidinones The methods used for the synthesis of 4-thiazolidinones and their derivatives depend undoubtedly on the nature of thiazolidinone subtypes and are well represented in numerous reviews and original articles (see above) [28] . Presented retro-synthetic approach ( Fig. 2 ) shows wide range of synthetic routes for thiazolidine ring formation based on various condensation reactions that successfully have been employed for the 4-thiazolidinones synthesis [1] , [2] , [3] , [6] . Fig. 2 Retro-synthetic approach to the 4-thiazolidinone scaffold formation, adapted from Ref. [29] . Fig. 2 These approaches are also speculatively attractive for the synthesis of 5-thiazolidinones as positional isomers of mentioned structures and they are rarely represented [30] . The most prominent and referred protocol for their synthesis is the cyclocondensation of reactants bearing the N=C=S fragment with α-halocarbonyl compounds. But, in most cases, mentioned protocol leads to the formation of 4-thiazolidinones only [31] . However, the protocols for synthesis of 5-ene derivatives of 4-thiazolidinones are often similar and can be divided into the next groups: i ) modification of the C5 position of the basic core; ii ) synthesis of the target compounds in the one-pot or multistage reactions (simultaneous formation of the core heterocycle and C5 exocyclic double bond); iii) transformation of thiazolidine derivatives or other related heterocycles in the recyclization reactions. 2.1 Modification of the C5 position of the thiazolidinone core Knoevenagel condensation. One of the most prominent and referred protocols for 5-ene-4-thiazolidinones synthesis is the Knoevenagel condensation of thiazolidinone core and oxo-compound ( Scheme 1 ). The methylene carbon atom at the С5-position possesses nucleophilic activity and can attack an electrophilic center affording the target ene-derivatives. The different aldehydes, ketones, and heterocycles (e.g. isatins, anhydrides of pyridine-3,4-dicarboxylic and phthalic acids etc.) [27] , [32] , [33] , [34] , [35] , [36] , [37] , [38] have been used as oxo compounds. Acetic acid or its anhydride and sodium acetate; ethanol and ammonium acetate or piperidine [39] ; toluene and ammonia acetate [33] , [40] , [41] ; isopropanol and potassium tert -butylate [42] , [43] ; toluene and l -proline [44] ; dimethylformamide and sodium acetate; ethanol and monoethanolamine as well as the solid carriers and phasetransfer catalysts etc. [45] , [46] have been widely used as the medium and catalysts in this reaction. The reaction performance was also described in the medium of the aldehyde or ketone (without solvent addition) [21] , polyethyleneglycol-300 [47] or based on the green chemistry approach. Moreover, performance of such condensation in aqueous solutions [48] ; usage of atypical agents as well as ionic liquids [49] ; the soluble polymer-supported synthesis [50] have also been investigated. Microwave assistant organic synthesis approach was also successfully employed [51] , [52] . Scheme 1 General scheme of Knoevenagel condensation of 4-thiazolidinones. Scheme 1 The utilization of the aromatic aldehydes leads to maximal yields of the target compounds, unlike aliphatic aldehydes and especially ketones [38] , [53] . The reactivity of 4-thiazolidinones differs also and depends on the main core's substituents (2,4-thiazolidinedione, rhodanine etc.) [27] , [43] , [43] : rhodanines react easier than 2-amino(imino)-4-thiazolidinone, 2,4-thiazolidinone etc; 2-R-substituted-4-thiazolidinones are characterized by the lower reactivity levels. When aldehydes are utilized the preferred (and mostly only one) product had been reported to have the Z geometry based on NMR and Х-ray study regardless of the thiazolidinone subtypes [53] , [54] , [55] . Knoevenagel condensation between 4-thiazolidinones and ketones generally leads to the mixtures of Z - and E -isomers. The use of the dicarbonyl compounds in this reaction leads to the formation of unfused heterocycles with two thiazolidinone fragments in molecule [36] , [56] , [57] , [58] . Despite the simplicity of such condensation, the formation of other products have been reported also [59] . For instance, thiopyrano[2,3- d ][1,3]thiazoles were obtained from highly reactive 4-thioxo-thiazolidin(thi)one-2 and ketones (in ethanol in the presence of monoethanolamine) [21] . The condensation of 2-imino-4-thiazolidinone and formaldehyde in water (pH∼9, triethylamine) leads to the formation of 5,5-bishydroxymetyl-2-imino-4-thiazolidinone ( 1 ) [37] , [60] . The uncatalyzed reaction in water with C−C bond-formation was reported to be thiazolidinedione-isatin conjugates' ( 2 ) synthetic protocol. The formation of the products was found to be thermodynamically controlled [59] , [61] and further heating leads to dehydration and double bound formation at C5 position of thiozolidinone core ( Scheme 2 ). Scheme 2 Variety of products of condensation reactions. Scheme 2 The reaction of pseudothiohydantoin (2-imino-4-thiazolidinone) and isatin (molar ratio 2:1) was carried out in absolute ethanol under refluxing affording not only target 5-ene-4-thiazolidinone ( 4 ) but also spiro-isatin conjugate (45%) 3 [57] . Based on the Knoevenagel condensation the efficient method for 5-ethoxy-4-thiazolidinones synthesis ( 5 ) using triethylorthoformate was described [32] , [62] , [63] , [64] , [65] , [66] , the latter ( 5 ) can be easily converted into the appropriate amines ( 6 ) ( Scheme 3 ). Scheme 3 5-Enamine-4-thiazolidinones formation. Scheme 3 Knoevenagel condensation often may be a stage in the tandem and/or domino reactions whilst the corresponding 5-ene-thiazolidinones appear as intermediates in the synthesis of fused thiazolidinone-based heterocycles (see bellow) [67] , [68] , [69] or simple molecules of other thiazolidinone subtypes. For example, an efficient approach to the 5-ene-2-amino-4-thiazolidinones obtaining via sulfur/nitrogen displacement [70] was the base for the development of multicomponent reactions involving Knoevenagel condensation [71] . The one-pot reaction of isatin derivatives, rhodanine and secondary amines (magnesium oxide nanoparticles used as heterogeneous catalyst, water medium, r.t.) is an efficient green approach to the preparation of novel isatin-thiazolidinone based conjugates ( 7 ) [72] ( Scheme 4 ). Scheme 4 One pot synthesis of 2.5-disubstituted thiazolidinones. Scheme 4 There has been developed a fast and efficient protocol for the generation of 3-substituted 5-arylidenerhodanines ( 8 ) ( Scheme 5 ) in sequential one-pot, two-step process combining the Knoevenagel condensation and alkylation reaction under microwave assisted conditions [73] . Scheme 5 One pot synthesis of 3,5-disubstituted thiazolidinones. Scheme 5 Other types of condensations. The active methylene group can be condensed with dimethylformamide-dimethylacetal (DMF-DMA) in dioxane to yield the corresponding enamines ( 9 ). Moreover, the methylene group in 4-thiazolidinone can be coupled with the aryl-diazonium salt in EtOH/NaOH to form the corresponding arylhydrazone ( 10 ) [3] , [74] ( Scheme 6 ). Similar to 5-arylidene-4-thiazolidinones, in all cases only Z -isomers were formed (confirmed by X-ray data) [75] . The isosteric 2-R-substituted-4-thioxothiazolidines were also prepared in a single step on simultaneous treatment of thiazolidin-4-one with DMF-DMA and Lawesson's reagent in toluene medium [76] . Under the action of DMF/POCl 3 on 3-substituted 2-ylidene-thiazolidinones the related enamine derivatives of pyrolo[2,1- b ]thiazol-3-ones 11 were synthesized ( Scheme 7 ) [77] . Scheme 6 Synthesis of 4-thiazolidinone based enamines and arylhydrazones. Scheme 6 Scheme 7 Synthesis of pyrolo[2,1- b ]thiazol-3-ones based on 4-thiazolidinones. Scheme 7 The active methylene group in the 4-thiazolidinones underwent nucleophilic addition reaction with the double bond of the various arylidene malononitriles via Michael type addition reaction (ethanol medium in the presence of piperidine) to give the same 5-ene-4-thiazolidinone derivatives ( 12 ) [75] ( Scheme 8 ). Scheme 8 Michael type addition reaction of 4-thiazolidinones. Scheme 8 For instance, reaction of α-cyano-3,4,5-trimethoxycinnamonitrile and/or ethyl-α-cyano-3,4,5-trimethoxycinnamate with 2-imino-4-oxo-thiazolidine provided appropriate 5-enes' formation (instead of the expected fused heterocycles) [78] , [79] . It was proposed that the reactions proceeded via nucleophilic addition of the thiazolidinyl-C-5 to the β-carbon of the activated double bond of nitriles forming the 1:1 adduct followed by the elimination of malononitrile or ethylcyanoacetate. The same step was proposed as a mechanism of dihydrothiophene derivatives formation via the four-component reaction of aldehyde, malonitrile, 2,4-thiazolidinedione and piperidine in the presence of Bu 4 NOH as a basic ionic liquid in al≿ohol-aqueous medium [49] or via the triethylamine-catalyzed domino reaction [80] , [81] . Isothiocyanate based synthesis. The reaction of 2-arylimino-3-R-4-thiazolidinones with arylisothiocyanate in DMF in the presence of KOH provided the appropriate thioamides ( 13 ) [82] , [83] , which may be treated as effective building blocks for the synthesis of polyfunctional compounds. The latter reacted with hydrazonoyl halides affording new 5-heterylidene derivatives 14 ( Scheme 9 ). Scheme 9 Isothiocynate based synthesis of 5-ene-4-thiazolidinones. Scheme 9 One more efficient method for 5-ene-4-thiazolidinone synthesis is the reaction of 4-thiazolidinones (e.g. rhodanine-3-acetic acid) with anilide-vinyl compounds [84] or with hemicyanines [85] with formation of dimethinemerocyanine and cyanine dyes ( 15 , 16 ) ( Scheme 10 ). Scheme 10 4-Thiazolidinone-based dimethinemerocyanine and cyanine dyes. Scheme 10 Merocyanine dyes ( 17 – 20 ) are other examples of polycyclic heterocycles derived from 5-ene-4-thiazolidinones. Their synthesis involves C5 active methylene group of thiazolidinone core and methylated thioxo-group (C2 position of rhodanine-3-acetic acid [86] ( Scheme 11 )) or other heterocyclic S-Me salts [84] , [87] , [88] ( Scheme 12 ). Scheme 11 Synthesis of 4-thiazolidinone based merocyanine dyes. Scheme 11 Scheme 12 Synthesis of 5-heterylidene 4-thiazolidinones. Scheme 12 Rhodanine derivatives, such as rhodanine-3-carboxylic acids, via the action of thionyl chloride were exposed to dimerisation at the C5 position and yielded appropriate acyl chlorides ( 21 ), which can be used for further modification ( 22 ) [89] ( Scheme 13 ). Scheme 13 Synthesis of bis-5-ene-4-thiazolidinones. Scheme 13 Reaction of 3-alkyl(aryl)rhodanines with bis-(trimethylsilyl)formamide also led to dimerisation and formation of the bis-(3-R-rhodaninyl-5)methinoxynes ( 23 ) [90] . The reaction of 4-thiazolidinones with halogens led to simple 5-halogen-derivatives ( 24 ) ( Scheme 14 ) which are the effective reagents for further synthetic transformation. Thus, based on ethyl ester of 5-bromo-2-thioxo-4-thiazolidone-3-acetic acid an original method for 5-ylidene derivative ( 25 ) synthesis was proposed [91] . Scheme 14 Utilization of 5-bromo-4-thiazolidinone for synthesis of 5-heterylidene derivatives. Scheme 14 Another efficient method for the 5-ene-4-thiazolidinones ( 26 , 27 ) synthesis (especially in the case of compounds containing carboxylic or carbonylic group in the C5 substituent) is a dehydrobromination reaction [92] , [93] , [94] ( Scheme 15 ). 5-(2-Aryl-2-oxoethyl)-rhodanines reacted with bromine in acetic acid and formed a mixture of E- and Z -stereoisomers of 5-aroylmethylenerhodanines. Scheme 15 Synthesis and transformation of 5-carboxymethylidene-4-thiazolidinone and related derivatives. Scheme 15 Described esters of 5-carboxymethylidene-4-thiazolidinone (2,4-dioxothiazolidin-5-ylidene-acetic acid) ( 26 ) are attractive building blocks for the various 4-thiazolidonone subtypes synthesis [94] , [95] including 5-ene derivatives. Addition of dithiocarbamates and thiocarbamates to the double bond in 5-ene-4-thiazolidinones resulted in spontaneous heterocylization to 5,5′-di-4-thiazolidinones intermediates, which were oxidized in the presence of triethylamine to 28 [96] ( Scheme 15 ). 2.2 Simultaneous formation of heterocycle and C5 exocyclic double bond Most of the protocols presented above are rather simple and most of them explore two- or three-step procedures. Considering the current trends in organic/medicinal chemistry different one-pot multicomponent methods were proposed for 5-ene-4-thiazolidinones synthesis ( 29 ) ( Scheme 16 ). The illustrative example is the one-pot method based on [2 + 3]-cyclocondensation of substituted thioureas/thiosemicarbazides with halogen-carboxylic acids followed by Knoevenagel condensation designed in our laboratory [59] , [97] , [98] and widely used [99] under wet chemistry conditions as well as under microwave irradiation or in the green reaction media [55] , [100] , [101] , [102] . Scheme 16 One-pot three component synthesis of 5-ene-4-thiazolidinones. Scheme 16 This approach can be successfully employed for the synthesis of fused heterocycles bearing 5-ene-4-thiazolidinones moieties e.g., 5-ylidene-[1,3]thiazolo[3,2- b ]1,2,4]triazol-6-ones ( 30 ) [103] ( Scheme 16 ). The common approach to the synthesis of 5-ylidene-4-thiazolidinones with carboxylic group ( 31 ) based on [2 + 3]-cyclocondensation reaction is the utilization of unsaturated acids and their derivatives ( Scheme 17 ). This approach is also efficient in the synthesis of various C2 substituted 4-thiazolidinones [104] , [105] , [106] (target compounds can be N-substituted or with unsubstituted N3 position). The condensation of thioamides with dimethylacethylenedicarboxylate in benzene (or acetonitrile or acetic acid for thioacetamide) led to appropriate thiazolidinone ( 32, 33 ) [107] , [108] ( Scheme 18 ). But there are reports about the formation of other products, such as 2-(1-iminoethylsulfanyl)-fumarate [28] , [109] . Scheme 17 General scheme of [2 + 3]-cyclocondensation of unsaturated acids. Scheme 17 Scheme 18 Condensation of thioamides with dimethylacethylenedicarboxylate. Scheme 18 Similar ylidene derivatives (mixture of E / Z isomers) 34 were formed in the reaction of α-carbamoyl(cyano)thioacetamides with dimethylacethylenedicarboxylate [28] , [110] , [111] . Thioamides of α,β-unsaturated acids in the reaction with acethylenedicarboxylic acid, propionic acid and their esters in acetone medium also yield the thiazolidinones as well as thiazanones [28] , [112] . The three-component reactions of amine with dialkyl acetylenedicarboxylate and carbon disulfide yielded related rhodanine derivatives ( 35 ) [113] , [114] ( Scheme 19 ). Isocyanate utilization instead of carbone sulfide under the same conditions led to maleimide derivatives formation. Scheme 19 Carbon disulfide based synthesis of 5-ene-rhodanines. Scheme 19 The exploration of α-chloro-β,γ-alkenoate esters in such type of multicomponent reactions led to the formation of 5-( Z )-alkylidene-2-thioxo-1,3-thiazolidin-4-ones ( 36 ) which are uncommon compounds [115] . Arylpropiolates are efficient building blocks for 5-ylidenethiazolidinone constructing ( Scheme 20 ). Reaction of the latter with bifunctional sulfur pronucleophiles is the phosphine-catalyzed tandem process which includes umpolung addition and intramolecular cyclization ( 37 ). Similarly, rhodanine derivatives ( 38 ) [116] were synthesized from dithiocarbamates in the phosphine catalyzed reaction ( Scheme 20 ). Scheme 20 Arylpropiolates based synthesis of 5-ene-4-thiazolidinones. Scheme 20 Reaction of trisubstituted thioureas with dimethylacetylenedicarboxylate (DMDA) in methanol medium can also lead to 1:1 adducts ( 39 ) [108] ( Scheme 21 ). Scheme 21 Example of the thioureas utilization in 5-ene-4-thiazolidinone derivatives synthesis. Scheme 21 Similarly, pseudothiohydantoin (2-amino-1,3-thiazol-4(5 H )-one) derivatives ( 40 ) can be easily obtained under the propenoylthioureas oxidation (e.g., by bromine action) ( Scheme 22 ) [117] . Scheme 22 Synthesis of the pseudothiohydantoin derivatives. Scheme 22 Novel 2-hydrazolyl-4-thiazolidinone-5,6-α,β-unsaturated esters ( 41 ) were synthesized in the multicomponent reaction of aldehydes, thiosemicarbazides and dimethylacetylenedicarboxylate in the ethanol medium. Interestingly, the reaction doesn't depend on the presence of electron-withdrawing or electron donating groups. It involves thiosemicarbazone formation followed by Michael addition of the sulfur atom to the triple bond and sequential cyclization ( Scheme 23 ) [118] . Scheme 23 Multicomponent reaction for 2-hydrazolyl-4-thiazolidinones synthesis. Scheme 23 Some epoxy acids ( cis - and trans -epoxysuccinic, 2,3-epoxybutyric and methyl-cis-epoxysuccinic acids) constitute another equivalent of the dielectrophilic synthon [C 2 ] 2+ in [2 + 3]-cyclocondensation reactions with thiourea when obtaining 5-substituted pseudothiohydantoines ( 42 ) and 2,4-thiazolidinediones ( 43 ) [119] ( Scheme 24 ). Scheme 24 Synthesis of pseudothiohydantoines and 2,4-thiazolidinediones from epoxy acids. Scheme 24 A fast and high yielding protocol for the generation of substituted 5-arylidenerhodanines ( 44 ) in sequential one-pot two-step process combining the Holmberg method and the Knoevenagel condensation under microwave assisted conditions has been developed [73] ( Scheme 25 ). Scheme 25 Holmberg method and Knoevenagel condensation based one-pot method. Scheme 25 An important synthetic route to 3-substituted 4-thiazolidinones is the reaction of α-mercaptocarboxylic acid derivatives with ethylcyanoacetate, cyanoacetamides, malonodinitrile etc. [2] , [3] , [6] , [120] , [121] as well as the cyclocondensation of oxonitriles or cyanoguanidines with 2-mercaptosuccinic acid derivatives [121] ( Scheme 26 ). Obtained 2-methyl-4-thiazolidinones 45 were exposed to regioselective bromination and dehydrogenation to form 5-ene-4-thiazolidinone derivatives 46 [122] , [123] . Scheme 26 Synthesis of 2,5-diene-4-thiazolidinone derivatives. Scheme 26 The most convenient method for 2-substituted-4-thiazolidinones synthesis is the one-pot three-component reaction of a primary amine, an oxo-compound, and a thiolic agent using various reaction conditions, such as extended heating with a dehydrating agent, using an acylation agent or microwave-assisted organic synthesis [124] , [125] , [126] , [127] , [128] , [129] . Based on the retrosynthetic approach the synthesis of 5-ene-2,3-disubstituted-4-thiazolidinones ( 47 ) was proposed [43] ( Scheme 27 ). 3-Substituted-2-mercaptoacrylic acids obtained from 5-arylidenerhodanine were used as thiolic agents. Scheme 27 Synthesis of 5-ene-2-R-4-thiazolidinones. Scheme 27 The 5-ene-4-thiazolidinones ( 48 ) were prepared in such one-pot three-component reaction where nitriles were used instead of corresponding amines [130] , [131] . The similar three-component reaction was proposed for the fused heterocycles synthesis, namely thiazolo[3,2- a ]pyridine derivatives ( 49 ) [132] , [133] ( Scheme 28 ). Scheme 28 Three-component reaction in the synthesis of fused heterocycles with 5-ene-4-thiazolidinone fragments. Scheme 28 This type of three-component domino reaction of readily available thioglycolic acid/ethylthioglycolate, aromatic aldehydes and malononitrile/ethylcyanoacetate was described in the aqueous potassium carbonate at r.t [134] . The one-pot reaction based on the condensation of ethylthiocyanoacetate, arylidenehydrazine and hydrazine hydrate in glacial acetic acid also led to the formation of 5-arylidene-2-arylidenehydrazone-4-thiazolidinone ( 50 ) [135] ( Scheme 29 ). Scheme 29 One-pot method of 5-ene-2-hydrazone-4-thiazolidinone synthesis. Scheme 29 Similarly to the 5-unsubstituted 4-thiazolidinones synthesis, the sequential three-component reaction involving primary amines, isothiocyanates, and 1,2-diaza-1,3-dienes was proposed as an efficient method for the 2-iminothiazolidin-4-ones-synthesis ( 51 ) [136] ( Scheme 30 ). Scheme 30 1,2-Diaza-1,3-dienes in 5-ene-4-thiazolidinone synthesis. Scheme 30 Formation of the 2-iminothiazolidin-4-one was explained by the initial regioselective S-Michael addition of the thiourea intermediate, which resulted from the coupling of amine and isothiocyanate, to the electrophilic center of 1,2-diaza-1,3-diene. The next step was the intramolecular attack of the -NH of the obtained isothiourea derivative on the ester group in C4 of the hydrazone chain with a loss of an alcohol molecule [136] . The hydrolytic cleavage of the hydrazide moiety of 51 afforded new 5-hydroxyethylidene thiazolidinones 52 [136] . 2.3 Transformation of related heterocycles The transformation of related heterocycles aimed to form 5-ene-4-thiazolidinone is not very often explored in the 4-thiazolidinone synthesis due to the simplicity of the methods described above. However, in the reaction of monosubstituted ureas with 4-chloro-5-formylthiazolin-2-one ( 53 ) the rearrangement occurs yielding 2,4-dioxothiazolidin-5-ylidenemethyl-urea ( 54 ). The starting 4-chloro-5-formylthiazolin-2-one ( 53 ) was obtained in the reaction of 2,4-thiazolidinedione formylation [137] ( Scheme 31 ). Scheme 31 Recyclization in the synthesis of 2,4-dioxothiazolidin-5-ylidenemethyl-urea. Scheme 31 2.1 Modification of the C5 position of the thiazolidinone core Knoevenagel condensation. One of the most prominent and referred protocols for 5-ene-4-thiazolidinones synthesis is the Knoevenagel condensation of thiazolidinone core and oxo-compound ( Scheme 1 ). The methylene carbon atom at the С5-position possesses nucleophilic activity and can attack an electrophilic center affording the target ene-derivatives. The different aldehydes, ketones, and heterocycles (e.g. isatins, anhydrides of pyridine-3,4-dicarboxylic and phthalic acids etc.) [27] , [32] , [33] , [34] , [35] , [36] , [37] , [38] have been used as oxo compounds. Acetic acid or its anhydride and sodium acetate; ethanol and ammonium acetate or piperidine [39] ; toluene and ammonia acetate [33] , [40] , [41] ; isopropanol and potassium tert -butylate [42] , [43] ; toluene and l -proline [44] ; dimethylformamide and sodium acetate; ethanol and monoethanolamine as well as the solid carriers and phasetransfer catalysts etc. [45] , [46] have been widely used as the medium and catalysts in this reaction. The reaction performance was also described in the medium of the aldehyde or ketone (without solvent addition) [21] , polyethyleneglycol-300 [47] or based on the green chemistry approach. Moreover, performance of such condensation in aqueous solutions [48] ; usage of atypical agents as well as ionic liquids [49] ; the soluble polymer-supported synthesis [50] have also been investigated. Microwave assistant organic synthesis approach was also successfully employed [51] , [52] . Scheme 1 General scheme of Knoevenagel condensation of 4-thiazolidinones. Scheme 1 The utilization of the aromatic aldehydes leads to maximal yields of the target compounds, unlike aliphatic aldehydes and especially ketones [38] , [53] . The reactivity of 4-thiazolidinones differs also and depends on the main core's substituents (2,4-thiazolidinedione, rhodanine etc.) [27] , [43] , [43] : rhodanines react easier than 2-amino(imino)-4-thiazolidinone, 2,4-thiazolidinone etc; 2-R-substituted-4-thiazolidinones are characterized by the lower reactivity levels. When aldehydes are utilized the preferred (and mostly only one) product had been reported to have the Z geometry based on NMR and Х-ray study regardless of the thiazolidinone subtypes [53] , [54] , [55] . Knoevenagel condensation between 4-thiazolidinones and ketones generally leads to the mixtures of Z - and E -isomers. The use of the dicarbonyl compounds in this reaction leads to the formation of unfused heterocycles with two thiazolidinone fragments in molecule [36] , [56] , [57] , [58] . Despite the simplicity of such condensation, the formation of other products have been reported also [59] . For instance, thiopyrano[2,3- d ][1,3]thiazoles were obtained from highly reactive 4-thioxo-thiazolidin(thi)one-2 and ketones (in ethanol in the presence of monoethanolamine) [21] . The condensation of 2-imino-4-thiazolidinone and formaldehyde in water (pH∼9, triethylamine) leads to the formation of 5,5-bishydroxymetyl-2-imino-4-thiazolidinone ( 1 ) [37] , [60] . The uncatalyzed reaction in water with C−C bond-formation was reported to be thiazolidinedione-isatin conjugates' ( 2 ) synthetic protocol. The formation of the products was found to be thermodynamically controlled [59] , [61] and further heating leads to dehydration and double bound formation at C5 position of thiozolidinone core ( Scheme 2 ). Scheme 2 Variety of products of condensation reactions. Scheme 2 The reaction of pseudothiohydantoin (2-imino-4-thiazolidinone) and isatin (molar ratio 2:1) was carried out in absolute ethanol under refluxing affording not only target 5-ene-4-thiazolidinone ( 4 ) but also spiro-isatin conjugate (45%) 3 [57] . Based on the Knoevenagel condensation the efficient method for 5-ethoxy-4-thiazolidinones synthesis ( 5 ) using triethylorthoformate was described [32] , [62] , [63] , [64] , [65] , [66] , the latter ( 5 ) can be easily converted into the appropriate amines ( 6 ) ( Scheme 3 ). Scheme 3 5-Enamine-4-thiazolidinones formation. Scheme 3 Knoevenagel condensation often may be a stage in the tandem and/or domino reactions whilst the corresponding 5-ene-thiazolidinones appear as intermediates in the synthesis of fused thiazolidinone-based heterocycles (see bellow) [67] , [68] , [69] or simple molecules of other thiazolidinone subtypes. For example, an efficient approach to the 5-ene-2-amino-4-thiazolidinones obtaining via sulfur/nitrogen displacement [70] was the base for the development of multicomponent reactions involving Knoevenagel condensation [71] . The one-pot reaction of isatin derivatives, rhodanine and secondary amines (magnesium oxide nanoparticles used as heterogeneous catalyst, water medium, r.t.) is an efficient green approach to the preparation of novel isatin-thiazolidinone based conjugates ( 7 ) [72] ( Scheme 4 ). Scheme 4 One pot synthesis of 2.5-disubstituted thiazolidinones. Scheme 4 There has been developed a fast and efficient protocol for the generation of 3-substituted 5-arylidenerhodanines ( 8 ) ( Scheme 5 ) in sequential one-pot, two-step process combining the Knoevenagel condensation and alkylation reaction under microwave assisted conditions [73] . Scheme 5 One pot synthesis of 3,5-disubstituted thiazolidinones. Scheme 5 Other types of condensations. The active methylene group can be condensed with dimethylformamide-dimethylacetal (DMF-DMA) in dioxane to yield the corresponding enamines ( 9 ). Moreover, the methylene group in 4-thiazolidinone can be coupled with the aryl-diazonium salt in EtOH/NaOH to form the corresponding arylhydrazone ( 10 ) [3] , [74] ( Scheme 6 ). Similar to 5-arylidene-4-thiazolidinones, in all cases only Z -isomers were formed (confirmed by X-ray data) [75] . The isosteric 2-R-substituted-4-thioxothiazolidines were also prepared in a single step on simultaneous treatment of thiazolidin-4-one with DMF-DMA and Lawesson's reagent in toluene medium [76] . Under the action of DMF/POCl 3 on 3-substituted 2-ylidene-thiazolidinones the related enamine derivatives of pyrolo[2,1- b ]thiazol-3-ones 11 were synthesized ( Scheme 7 ) [77] . Scheme 6 Synthesis of 4-thiazolidinone based enamines and arylhydrazones. Scheme 6 Scheme 7 Synthesis of pyrolo[2,1- b ]thiazol-3-ones based on 4-thiazolidinones. Scheme 7 The active methylene group in the 4-thiazolidinones underwent nucleophilic addition reaction with the double bond of the various arylidene malononitriles via Michael type addition reaction (ethanol medium in the presence of piperidine) to give the same 5-ene-4-thiazolidinone derivatives ( 12 ) [75] ( Scheme 8 ). Scheme 8 Michael type addition reaction of 4-thiazolidinones. Scheme 8 For instance, reaction of α-cyano-3,4,5-trimethoxycinnamonitrile and/or ethyl-α-cyano-3,4,5-trimethoxycinnamate with 2-imino-4-oxo-thiazolidine provided appropriate 5-enes' formation (instead of the expected fused heterocycles) [78] , [79] . It was proposed that the reactions proceeded via nucleophilic addition of the thiazolidinyl-C-5 to the β-carbon of the activated double bond of nitriles forming the 1:1 adduct followed by the elimination of malononitrile or ethylcyanoacetate. The same step was proposed as a mechanism of dihydrothiophene derivatives formation via the four-component reaction of aldehyde, malonitrile, 2,4-thiazolidinedione and piperidine in the presence of Bu 4 NOH as a basic ionic liquid in al≿ohol-aqueous medium [49] or via the triethylamine-catalyzed domino reaction [80] , [81] . Isothiocyanate based synthesis. The reaction of 2-arylimino-3-R-4-thiazolidinones with arylisothiocyanate in DMF in the presence of KOH provided the appropriate thioamides ( 13 ) [82] , [83] , which may be treated as effective building blocks for the synthesis of polyfunctional compounds. The latter reacted with hydrazonoyl halides affording new 5-heterylidene derivatives 14 ( Scheme 9 ). Scheme 9 Isothiocynate based synthesis of 5-ene-4-thiazolidinones. Scheme 9 One more efficient method for 5-ene-4-thiazolidinone synthesis is the reaction of 4-thiazolidinones (e.g. rhodanine-3-acetic acid) with anilide-vinyl compounds [84] or with hemicyanines [85] with formation of dimethinemerocyanine and cyanine dyes ( 15 , 16 ) ( Scheme 10 ). Scheme 10 4-Thiazolidinone-based dimethinemerocyanine and cyanine dyes. Scheme 10 Merocyanine dyes ( 17 – 20 ) are other examples of polycyclic heterocycles derived from 5-ene-4-thiazolidinones. Their synthesis involves C5 active methylene group of thiazolidinone core and methylated thioxo-group (C2 position of rhodanine-3-acetic acid [86] ( Scheme 11 )) or other heterocyclic S-Me salts [84] , [87] , [88] ( Scheme 12 ). Scheme 11 Synthesis of 4-thiazolidinone based merocyanine dyes. Scheme 11 Scheme 12 Synthesis of 5-heterylidene 4-thiazolidinones. Scheme 12 Rhodanine derivatives, such as rhodanine-3-carboxylic acids, via the action of thionyl chloride were exposed to dimerisation at the C5 position and yielded appropriate acyl chlorides ( 21 ), which can be used for further modification ( 22 ) [89] ( Scheme 13 ). Scheme 13 Synthesis of bis-5-ene-4-thiazolidinones. Scheme 13 Reaction of 3-alkyl(aryl)rhodanines with bis-(trimethylsilyl)formamide also led to dimerisation and formation of the bis-(3-R-rhodaninyl-5)methinoxynes ( 23 ) [90] . The reaction of 4-thiazolidinones with halogens led to simple 5-halogen-derivatives ( 24 ) ( Scheme 14 ) which are the effective reagents for further synthetic transformation. Thus, based on ethyl ester of 5-bromo-2-thioxo-4-thiazolidone-3-acetic acid an original method for 5-ylidene derivative ( 25 ) synthesis was proposed [91] . Scheme 14 Utilization of 5-bromo-4-thiazolidinone for synthesis of 5-heterylidene derivatives. Scheme 14 Another efficient method for the 5-ene-4-thiazolidinones ( 26 , 27 ) synthesis (especially in the case of compounds containing carboxylic or carbonylic group in the C5 substituent) is a dehydrobromination reaction [92] , [93] , [94] ( Scheme 15 ). 5-(2-Aryl-2-oxoethyl)-rhodanines reacted with bromine in acetic acid and formed a mixture of E- and Z -stereoisomers of 5-aroylmethylenerhodanines. Scheme 15 Synthesis and transformation of 5-carboxymethylidene-4-thiazolidinone and related derivatives. Scheme 15 Described esters of 5-carboxymethylidene-4-thiazolidinone (2,4-dioxothiazolidin-5-ylidene-acetic acid) ( 26 ) are attractive building blocks for the various 4-thiazolidonone subtypes synthesis [94] , [95] including 5-ene derivatives. Addition of dithiocarbamates and thiocarbamates to the double bond in 5-ene-4-thiazolidinones resulted in spontaneous heterocylization to 5,5′-di-4-thiazolidinones intermediates, which were oxidized in the presence of triethylamine to 28 [96] ( Scheme 15 ). 2.2 Simultaneous formation of heterocycle and C5 exocyclic double bond Most of the protocols presented above are rather simple and most of them explore two- or three-step procedures. Considering the current trends in organic/medicinal chemistry different one-pot multicomponent methods were proposed for 5-ene-4-thiazolidinones synthesis ( 29 ) ( Scheme 16 ). The illustrative example is the one-pot method based on [2 + 3]-cyclocondensation of substituted thioureas/thiosemicarbazides with halogen-carboxylic acids followed by Knoevenagel condensation designed in our laboratory [59] , [97] , [98] and widely used [99] under wet chemistry conditions as well as under microwave irradiation or in the green reaction media [55] , [100] , [101] , [102] . Scheme 16 One-pot three component synthesis of 5-ene-4-thiazolidinones. Scheme 16 This approach can be successfully employed for the synthesis of fused heterocycles bearing 5-ene-4-thiazolidinones moieties e.g., 5-ylidene-[1,3]thiazolo[3,2- b ]1,2,4]triazol-6-ones ( 30 ) [103] ( Scheme 16 ). The common approach to the synthesis of 5-ylidene-4-thiazolidinones with carboxylic group ( 31 ) based on [2 + 3]-cyclocondensation reaction is the utilization of unsaturated acids and their derivatives ( Scheme 17 ). This approach is also efficient in the synthesis of various C2 substituted 4-thiazolidinones [104] , [105] , [106] (target compounds can be N-substituted or with unsubstituted N3 position). The condensation of thioamides with dimethylacethylenedicarboxylate in benzene (or acetonitrile or acetic acid for thioacetamide) led to appropriate thiazolidinone ( 32, 33 ) [107] , [108] ( Scheme 18 ). But there are reports about the formation of other products, such as 2-(1-iminoethylsulfanyl)-fumarate [28] , [109] . Scheme 17 General scheme of [2 + 3]-cyclocondensation of unsaturated acids. Scheme 17 Scheme 18 Condensation of thioamides with dimethylacethylenedicarboxylate. Scheme 18 Similar ylidene derivatives (mixture of E / Z isomers) 34 were formed in the reaction of α-carbamoyl(cyano)thioacetamides with dimethylacethylenedicarboxylate [28] , [110] , [111] . Thioamides of α,β-unsaturated acids in the reaction with acethylenedicarboxylic acid, propionic acid and their esters in acetone medium also yield the thiazolidinones as well as thiazanones [28] , [112] . The three-component reactions of amine with dialkyl acetylenedicarboxylate and carbon disulfide yielded related rhodanine derivatives ( 35 ) [113] , [114] ( Scheme 19 ). Isocyanate utilization instead of carbone sulfide under the same conditions led to maleimide derivatives formation. Scheme 19 Carbon disulfide based synthesis of 5-ene-rhodanines. Scheme 19 The exploration of α-chloro-β,γ-alkenoate esters in such type of multicomponent reactions led to the formation of 5-( Z )-alkylidene-2-thioxo-1,3-thiazolidin-4-ones ( 36 ) which are uncommon compounds [115] . Arylpropiolates are efficient building blocks for 5-ylidenethiazolidinone constructing ( Scheme 20 ). Reaction of the latter with bifunctional sulfur pronucleophiles is the phosphine-catalyzed tandem process which includes umpolung addition and intramolecular cyclization ( 37 ). Similarly, rhodanine derivatives ( 38 ) [116] were synthesized from dithiocarbamates in the phosphine catalyzed reaction ( Scheme 20 ). Scheme 20 Arylpropiolates based synthesis of 5-ene-4-thiazolidinones. Scheme 20 Reaction of trisubstituted thioureas with dimethylacetylenedicarboxylate (DMDA) in methanol medium can also lead to 1:1 adducts ( 39 ) [108] ( Scheme 21 ). Scheme 21 Example of the thioureas utilization in 5-ene-4-thiazolidinone derivatives synthesis. Scheme 21 Similarly, pseudothiohydantoin (2-amino-1,3-thiazol-4(5 H )-one) derivatives ( 40 ) can be easily obtained under the propenoylthioureas oxidation (e.g., by bromine action) ( Scheme 22 ) [117] . Scheme 22 Synthesis of the pseudothiohydantoin derivatives. Scheme 22 Novel 2-hydrazolyl-4-thiazolidinone-5,6-α,β-unsaturated esters ( 41 ) were synthesized in the multicomponent reaction of aldehydes, thiosemicarbazides and dimethylacetylenedicarboxylate in the ethanol medium. Interestingly, the reaction doesn't depend on the presence of electron-withdrawing or electron donating groups. It involves thiosemicarbazone formation followed by Michael addition of the sulfur atom to the triple bond and sequential cyclization ( Scheme 23 ) [118] . Scheme 23 Multicomponent reaction for 2-hydrazolyl-4-thiazolidinones synthesis. Scheme 23 Some epoxy acids ( cis - and trans -epoxysuccinic, 2,3-epoxybutyric and methyl-cis-epoxysuccinic acids) constitute another equivalent of the dielectrophilic synthon [C 2 ] 2+ in [2 + 3]-cyclocondensation reactions with thiourea when obtaining 5-substituted pseudothiohydantoines ( 42 ) and 2,4-thiazolidinediones ( 43 ) [119] ( Scheme 24 ). Scheme 24 Synthesis of pseudothiohydantoines and 2,4-thiazolidinediones from epoxy acids. Scheme 24 A fast and high yielding protocol for the generation of substituted 5-arylidenerhodanines ( 44 ) in sequential one-pot two-step process combining the Holmberg method and the Knoevenagel condensation under microwave assisted conditions has been developed [73] ( Scheme 25 ). Scheme 25 Holmberg method and Knoevenagel condensation based one-pot method. Scheme 25 An important synthetic route to 3-substituted 4-thiazolidinones is the reaction of α-mercaptocarboxylic acid derivatives with ethylcyanoacetate, cyanoacetamides, malonodinitrile etc. [2] , [3] , [6] , [120] , [121] as well as the cyclocondensation of oxonitriles or cyanoguanidines with 2-mercaptosuccinic acid derivatives [121] ( Scheme 26 ). Obtained 2-methyl-4-thiazolidinones 45 were exposed to regioselective bromination and dehydrogenation to form 5-ene-4-thiazolidinone derivatives 46 [122] , [123] . Scheme 26 Synthesis of 2,5-diene-4-thiazolidinone derivatives. Scheme 26 The most convenient method for 2-substituted-4-thiazolidinones synthesis is the one-pot three-component reaction of a primary amine, an oxo-compound, and a thiolic agent using various reaction conditions, such as extended heating with a dehydrating agent, using an acylation agent or microwave-assisted organic synthesis [124] , [125] , [126] , [127] , [128] , [129] . Based on the retrosynthetic approach the synthesis of 5-ene-2,3-disubstituted-4-thiazolidinones ( 47 ) was proposed [43] ( Scheme 27 ). 3-Substituted-2-mercaptoacrylic acids obtained from 5-arylidenerhodanine were used as thiolic agents. Scheme 27 Synthesis of 5-ene-2-R-4-thiazolidinones. Scheme 27 The 5-ene-4-thiazolidinones ( 48 ) were prepared in such one-pot three-component reaction where nitriles were used instead of corresponding amines [130] , [131] . The similar three-component reaction was proposed for the fused heterocycles synthesis, namely thiazolo[3,2- a ]pyridine derivatives ( 49 ) [132] , [133] ( Scheme 28 ). Scheme 28 Three-component reaction in the synthesis of fused heterocycles with 5-ene-4-thiazolidinone fragments. Scheme 28 This type of three-component domino reaction of readily available thioglycolic acid/ethylthioglycolate, aromatic aldehydes and malononitrile/ethylcyanoacetate was described in the aqueous potassium carbonate at r.t [134] . The one-pot reaction based on the condensation of ethylthiocyanoacetate, arylidenehydrazine and hydrazine hydrate in glacial acetic acid also led to the formation of 5-arylidene-2-arylidenehydrazone-4-thiazolidinone ( 50 ) [135] ( Scheme 29 ). Scheme 29 One-pot method of 5-ene-2-hydrazone-4-thiazolidinone synthesis. Scheme 29 Similarly to the 5-unsubstituted 4-thiazolidinones synthesis, the sequential three-component reaction involving primary amines, isothiocyanates, and 1,2-diaza-1,3-dienes was proposed as an efficient method for the 2-iminothiazolidin-4-ones-synthesis ( 51 ) [136] ( Scheme 30 ). Scheme 30 1,2-Diaza-1,3-dienes in 5-ene-4-thiazolidinone synthesis. Scheme 30 Formation of the 2-iminothiazolidin-4-one was explained by the initial regioselective S-Michael addition of the thiourea intermediate, which resulted from the coupling of amine and isothiocyanate, to the electrophilic center of 1,2-diaza-1,3-diene. The next step was the intramolecular attack of the -NH of the obtained isothiourea derivative on the ester group in C4 of the hydrazone chain with a loss of an alcohol molecule [136] . The hydrolytic cleavage of the hydrazide moiety of 51 afforded new 5-hydroxyethylidene thiazolidinones 52 [136] . 2.3 Transformation of related heterocycles The transformation of related heterocycles aimed to form 5-ene-4-thiazolidinone is not very often explored in the 4-thiazolidinone synthesis due to the simplicity of the methods described above. However, in the reaction of monosubstituted ureas with 4-chloro-5-formylthiazolin-2-one ( 53 ) the rearrangement occurs yielding 2,4-dioxothiazolidin-5-ylidenemethyl-urea ( 54 ). The starting 4-chloro-5-formylthiazolin-2-one ( 53 ) was obtained in the reaction of 2,4-thiazolidinedione formylation [137] ( Scheme 31 ). Scheme 31 Recyclization in the synthesis of 2,4-dioxothiazolidin-5-ylidenemethyl-urea. Scheme 31 3 Pharmacological profiles of 5-ene-4-thiazolidinones Most of the pharmacologically attractive agents among 4-thiazolidinone-based compounds are exactly 5-ene-4-thiazolidinones with the exocyclic double bond. This is reflected in crucial impact of the presence/nature (namely 5-ene) of C5 substituents in thiazolidone ring on pharmacological effects [27] , [34] , [35] , [54] , [94] , [138] , [139] , [140] , [141] , [142] , [143] . At the same time, (ar)ylidene fragment conjugation with the carbonyl group at C4 of the thiazolidinone core makes such compounds electrophilic and potentially reactive due to possible Michael addition of the exocyclic double bond to the nucleophilic protein residues [144] , [145] . Moreover, 5-ene-thiazolidinones can react with glutathione and other free thiols within a cell [145] , [146] , [147] ( Sheme 32 ). Scheme 32 General scheme of thia-Michael addition of 5-ene-4-thiazolidinones. Scheme 32 This property characterizes 5-arylidene-4-thiazolidinones (mainly rhodanine derivatives) as frequent hitters or pan assay interference compounds (PAINS) that are thought to be useless in the modern drug discovery process because of their possible/predicted insufficient selectivity [4] , [14] , [148] , [149] , [150] . This, along with other structural features, offers high probability of polar interactions and hydrogen bonds formation. Causing a promiscuous behavior at concentrations within the "screening range" this statement should not be regarded as a general knock out criterion that excludes such screening hits from further development. It is suggested that special criteria for target affinity and selectivity must be applied to these classes of compounds so that their exceptional and potentially valuable biomolecular binding properties are consequently exploited in a useful way [14] . Interpretation of 5-arylidene-4-thiazolidinones as frequent hitters is often not confirmed in experimental studies [24] , [151] , [152] and a large number of lead-compounds belong to 5-ene-4-thiazolidinones. The positive perspective may also be associated with the polypharmacological approach in drug discovery, where the affinity toward various targets is regarded as an advantage [14] . Moreover, mentioned compounds as the examples of privileged scaffolds can be treated as drug-like molecules that provide baseline affinity for a whole protein family [14] , [153] . Additionally, better results in drug design can be expected when responses are evoked by multipoint interventions in more than one mechanism and towards different targets [154] following the concept of multi-target drugs, also known as dual or symbiotic drugs [155] . The exploitation of 5-ene diversity allows achieving such desired combinations of molecular fragments. There have been many reports describing the presence of heterocyclic moieties, such as thiazole, pyrazole, flavones, chromone, furan etc. at C5 position of thiazolidinone core [8] , [27] , [34] , [35] , [65] , [66] , [138] . It is even considered that the more such fragments are in the molecule, the more effective it is when comparing with the molecules including simple aryl groups (see below). In this spirit, lack of strict specificity in ligand binding and possibility of the influence of both the compound and its metabolites on the other metaboli≿ pathways are being actively discussed. This concept is one of the arguments in favour of the search for "nonspecific" biologically active compounds. Furthermore, many compounds possess different types of activity that are apparently due to their action towards differents metabolic pathways (targets) or involving the same biological targets in metabolic pathways of various pathologies. Thus, many authors consider that "thiazolidinones and related scaffolds should not be regarded as problematic or promiscuous binders per se" [14] . The intermolecular interaction profile of these scaffolds makes them prone to bind to a large number of targets with weak or moderate affinity. It may be that the observed moderate affinities, e.g. in screening campaigns, have been overinterpreted in the past and that these compounds have too easily been put forward as lead compounds for further development. In this context, it is worth mentioning that exactly the same Michael acceptors are among the most effective activators of Nrf2 through the Keap1 modification that opens new perspectives in the treatment of inflammation, cancer, etc. [144] , [156] . Moreover, Michael acceptors including 5-ene-thiazolidinones belong to a new class of highly specific and potent inhibitors of the mitochondrial pyruvate carrier [157] . In addition, it should be emphasized that the possibility of being Michael acceptors (calculated or predicted based on the molecular structure and used in the in silico studies) is often not confirmed experimentally under conditions close to physiological [156] . Besides, as was mentioned earlier, 5-ene-4-thiazolidinones are widely used in the design of complex heterocyclic systems within the hybride pharmacophore approach [158] , [159] . The significant biological potency of compounds bearing 5-ene-4-thiazolidinone fragment in combination, for example, with heterocycles [27] , [34] , [65] , [138] , [160] or other scaffolds [161] , [162] was discovered. Satisfactory toxicological parameters of such compounds may be an additional argument in favour of their development and study [163] . The next chapter represents the most prominent pharmacological profiles of 5-ene from different 4-thiazolidinone subtypes which belong to hit-, lead-compounds, drug-candidates and drugs. Two main directions of the study of biological activity of 5-ene derivatives are presented: the identification of a number of compounds possessing some types of biological activity in screening procedure (without molecular targets identification); design of high-affinity ligands to established molecular targets (for instance, a lot of thiazolidinones are proved to be ligands to the biological targets presented in protein data bank ( www.rscb.org )). Moreover, several types of activity are often described for the same compounds that clearly follows from the above said. A large number of publications are dedicated to the in silico study of 4-thiazolidinones and particularly 5-ene derivatives. In this paper we did not focus on such data as it is widely outlined in the review papers [164] , [165] , [166] , [167] , [168] , [169] , [170] . A wide range of established pharmacological activity of 5-ene-4-thiazolidinones does not allow to describe all its types in the given manuscript. Thereby, we focused on the most studied and described types of activity such as antimicrobial, antitumor etc. Besides, the compounds bearing fused 5-ene fragment (e.g. thiazolothiopyranes, thiazolopyridines etc.) are beyond the scope of this paper. 3.1 Antimicrobial agents One of the earliest directions of the 5-ene-4-thiazolidinones biological assays lies in the field of antibacterial and antifungal activities search. Apparently, the structural similarity of 4-azolidinones with penicillin antibiotics was the stimulus to the study of such type of activity [1] , [2] , [3] , [6] , [167] , [168] . Though, recently it was discovered that manifestation of antibacterial effect is not always related to the penicillins' mode of action. Usually, there is a tendency to move from the detection of antimicrobial activity [169] , [170] , [171] , [172] , [173] , [174] , [175] , [176] within screening programs to identification and design of high affinity ligands to the validated molecular targets. Regardless of the nature of the basic core (2,4-thiazolidinedione, rhodanine, etc.), modification and complication of C5 ylidene fragment are the benefits for realization of such type of activity (the role of the halogen-arylidene moieties should be stressed) [141] , [177] . The 3-substituted 5-ylidene-4-thiazolidinones, especially 3-carboxylic acids 55 – 59 [171] , [175] , [176] , [177] , [178] , [179] , [180] , [181] represent group of compounds with significant antimicrobial activity against gram-positive and gram-negative microorganisms. Their molecular mode of action is associated with the inhibition of: i) the last two stages of peptidoglycan cell wall biosynthesis (the most attractive targets for new antimicrobials design - mainly penicillin-binding proteins family , and transferases – UDP-Mur-NAc-peptidases ); іі) the activity of dehydrogenases (compound 58 and related structures can cross the bacterial cell wall and lead to the identification of catechol-rhodanine core as privileged scaffold for design of the dehydrogenases inhibitors) [153] ; iii) peptide deformylase [182] (for lead-compounds 60 , 61 IC 50 = 0.89 μM and 1.66 μM correspondly) etc ( Scheme 33 ). Scheme 33 5-Ene-4-thiazolidinone-3-carboxylic acids with antimicrobial activity. Scheme 33 A series of 5-ene-rhodanine-3-acetic acid derivatives 62 ( Scheme 34 ) were described as inhibitors of fungal protein mannosyl transferase 1 (PMT1) at micromolar levels (design of compounds was also based on the complications of C5 moiety) [183] . High-throughput screening of 5-arylidene-4-thiazolidinone-3-alkanecarboxylic acids 63 , 64 proved them to be anthrax lethal factor (as one of the major virulence factors) inhibitors (IC 50 ∼ 10 μМ) [45] , [184] . Scheme 34 5-Ene-rhodanines as antimicrobial agents. Scheme 34 Antimicrobial activity of such compounds are determined by inhibition of penicillin-binding proteins family , such as a group of transferases (including UDP-Mur-NAc-peptidases ), most of their subtypes are characterized by distinct transglycosylase activity [185] . ATP-dependent amino acid ligases (MurC, MurD, MurE, and MurF, MurG) are an illustrative examples of the 5-ene-4-thiazolidinones antibacterial agents design [41] , [186] , [187] , [188] . For instance, 65 is an efficient inhibitor of MurC ( UDP-N-acetylmuramate/ l -alanine ligase ) [139] and 66 – of MurG [189] . It should be emphasized that the 4-thiazolidinone and rhodanine cores are treated as novel phosphate mimics [190] ( Scheme 34 ). Structurally similar compounds are the inhibitors of dTDP-rhamnose synthesis . This is especially important for antimycobacterial agents search. The mycobacterial cell wall is unique because it contains an amycolylarabinogalactan layer bound to the peptidoglycan layer via a rhamnose–Glc-NAc sugar linker, where the dTDP-rhamnose can be vetreated as a precursor [191] , [192] . 5-Heterylidene-2,4-thiazolidinedione derivatives 67 are competitive inhibitors of recombinant bacterial arylamine-N-acetyltransferases (NATs) [193] , which were designed based on the structure modification of N-(2-naphthyl)-methyl substituted 1,1-dioxo-2,3-dihydrobenzo[1,2]-thiazine-4-ylidenethiazolidine-2,4-dione (weak inhibitor of NAT) [194] ( Scheme 35 ). Scheme 35 5-Heterylidene-4-thiazolidinedione with antimicrobial activity. Scheme 35 One of the pathogenetic processes of pro- and eukaryotes is the mechanism associated with galactofuranose-based conjugation which involves UDP-galactopyranose mutase (UGM). The inhibition of such process leads to inhibition of microbial growth and reduction of virulence (especially for Mycobacterium tuberculosis ) and improve the antimicrobial/antimycobacterial activity of UGM inhibitors. High inhibition activity of this enzyme was also discovered for 4-thiazolidinone-3-alkanecarboxylic acids 68 , 69 [180] and 2-imino derivatives 70 [145] , [195] ( Scheme 35 ). 2-Amino(imino)-4-thiazolidinones have been widely investigated [141] , [177] , [196] , [197] as antimicrobial agents, e.g. compound 71 was identified as an inhibitor of the type III secretion system of Gram-negative bacteria [196] , [198] . Pseudomonas aeruginosa had shown reduction of T2S-dependent elastase secretion in the presence of 71 (Type II secretion (T2S) systems is well conserved among Gram-negative bacteria and a key virulence factor of P. aeruginos a) [199] . The structure optimization aiming the increasing of solubility led to compound 74 which possessed superior activity (10 μM) in the S. typhimurium T3S secretion assay [196] . The N-3 position was established to be the most permissive optimization direction ( Scheme 36 ). Scheme 36 5-Ene-4-thiazolidinediones as inhibitors of the type II & III secretion systems. Scheme 36 The same findings are suitable for 2,4-thiazolidinedione derivatives – N3 modification provides to 5–40 folds increasing of the antimicrobial activity ( 72 ) [200] , [201] . Related 2-amino-4-thiazolidinone ( 73 ) was detected as high active CysK1 inhibitor . CysK1-pyridoxal phosphate-dependent O-acetylsulfhydrylase which catalyzes the formation of l -cysteine from O-acetylserine and hydrogensulfide. This cysteine biosynthetic pathway is one of the essential pathways in microbial pathogens, providing potential targets for the development of novel antibacterial compounds [202] ( Scheme 36 ). The row of 2-thiazolylimino-5-arylidene-4-thiazolidinone bearing the nitro group and small heterocyclic fragments (thiazole, benzthiazole, isatine, pyrazole, furane, thiophene) ( 75 – 77 ) ( Scheme 37 ) in positions 2 and 5 showed sufficient antimicrobial activity levels too [203] ; the nitro derivatives were characterized by high renal excretion that can be used in the design of potential diuretics [204] . The results of QSAR study revealed that the compounds with less number of atoms or less number of substituents are more likely to be active than their counterparts with higher molecular weight [205] . Scheme 37 2,5-Disubstituted-4-thiazolidinone hybrids with antimicrobial activity. Scheme 37 5-Heterylidenerhodanines and simple arylidene analogs are proposed as novel class of b-lactamase inhibitors that possesed selectivity for class C b-lactamases ( 78 is a non-b-lactam with the IC 50 ∼ 2.6 mM against the class C enzyme P99) [206] . Simple 5-alkylidene-2,4-thiazolidinediones 79 were discovered as effective inhibitors of autoinducer-2 quorum sensing (mechanism through which bacteria regulate gene expression in response to population density including regulation of the production of virulence factors etc.) Virulence targeting is representing an emerging concept in antibacterial therapy for which there are examples of compounds that inhibit virulence functions [207] . N-3-Dipeptide-thiazolidinone hybrids 80 may provide a critical step toward the validation of this strategy and the development of novel therapeutics [196] ( Scheme 38 ). Compounds having bulky aromatic substituents at position 5 and a tryptophan residue at position N-3 of the rhodanine ring were the most active against InhA ( trans -2-enoyl-acyl carrier protein reductase) with micromolar IC 50 [208] . Scheme 38 Structure of antimicrobial 5-ene-4-thiazolidinones. Scheme 38 3.2 Antiparasitic agents The in vitro antiamoebic activity of 3-substituted 2-amino-4-thiazolidinones 81 was evaluated against HM1:IMSS strain of Entamoeba histolytica and it exhibited promising activity (IC 50 - 0.11–0.172 mM being lower than that of metronidazole IC 50 - 1.64 mM) and low toxicity level [209] ( Scheme 39 ). Scheme 39 3-Substituted 2-amino-4-thiazolidinones with antiamoebic activity. Scheme 39 Several aza-fused rhodacyanines [210] were found as perspective agents when testing various heterocyclic rings on the antimalarial activity. Among tested compounds 82 showed excellent inhibitory activity with an IC 50 of 4.4 nM (K1strain). Another compound 83 with quinoline ring and two rhodanine moieties showed 78% suppression of parasitemia (25 mg/kg/day) [211] . 4-Thiazolidinone-3-carboxylic acids belonging to rhodacyanine dyes 84 are characterized by distinct antimalarial activity [87] , [212] . Significant anticancer activity of the mentioned class of compounds should be emphasized in this context (see below) ( Scheme 40 ). Scheme 40 Rhodanine-based dyes with antimalarial activity. Scheme 40 An in vitro structure–activity relationship investigation showed that the rhodacyanine MKH-57 possesses high antimalarial activity (EC 50 = 12 nM) and significant selective toxicity [87] as well as a series of its analogs [212] . Screening of more than 13000 compounds for antimalarial activity using the agglomerative structural clustering technique allowed identifying 47 starting points for lead optimization including 4-amino(imino)thiazolidinone derivatives 85 [213] . The positional isomer of mentioned 4-amino derivatives ( 85 ) – compound 86 was also described in the compound library of potential antimalarial leads [213] ( Scheme 41 ). Scheme 41 5-Ene-4(2)-thiazolidinones as antiparasitic agents. Scheme 41 In the area of antiprotozoal agents search the design of antitrypanosonal agents based on thiazolidinone scaffold is of special interest [214] . Compounds 87 showed the highest antiproliferative activity in comparison with 5-saturated analog 88 when screened on Trypanosoma cruzi epimastigotes but were inactive towards cruzipain [118] ( Scheme 41 ). While 88, obtained from a virtual screening of 500000 chemical structures (ZINC5 database) against cruzipain [215] inhibited this enzyme at micromolar concentration. 5-Benzylidenerhodanine-3-acetic acid derivatives 89 were reported to possess inhibitory activity against Trypanosoma brucei dolicholphosphate mannose synthase and glycosylphosphatidylinositol anchor synthesis as well as in vitro trypanocidal activity against the blood stream form and were non-cytotoxic against HeLa cells. Dolicholphosphatemannose synthase is a mannosyl transferase critically involved in glycoconjugate biosynthesis in T. brucei . VSG dimmers covering the parasite cell-surface are linked to the trypanosomal plasma membrane via glycosyl phosphatidylinositol anchors, which biosynthesis is essential for viability of the blood stream form of T. brucei . The 3-benzyloxy-substituted analog and the 2-hydroxyderivative 89 showed the best trypanocidal activity (ED 50 ∼ 100 μM) [215] ( Scheme 42 ). Scheme 42 5-Benzylidenerhodanine-3-acetic acid derivatives as antitrypanosomal agents. Scheme 42 Simple 5-(hydroxyphenylmethylidene)thiazolidine-2,4-diones were described as novel inhibitors of Leishmania pteridine reductase 1 where the thiazolidinone ring was treated as a bioisosteric replacement for pteridine/purine ring [216] . One of the discussed modes of action of thiazolidinone based compound is the DNA-binding process. In this study 2-imino derivatives 90 were designed and the most potent molecule bound at the DNA minor groove involving Van der Waals, H-bonding and hydrophobic interactions [217] ( Scheme 43 ). Scheme 43 2-(3-Imidazol-1-yl-propylimino)-5-(3-nitro-benzylidene)-thiazolidin-4-one as potential DNA-binding compound. Scheme 43 3.3 Antidiabetic agents Search for new antidiabetic agents among 4-thiazolidinones is one of the most studied fields and had resulted in the introduction of new class of the antidiabetic drugs – glitazones. Their mechanism of action is associated with activation of PPARγ (peroxisome proliferation activated receptors) [19] , [218] , [219] , [220] ( Scheme 44 ). Scheme 44 Structures of known glitazones. Scheme 44 However, 5-ene derivatives including C5-unsaturated analogs of glitazones are to lesser extent activators of PPARs [221] , [222] . The introduction of carboxylic acid residue in the N3 position leads to decrease of hepatotoxicity of glitazones as well as unsaturated structural analogs (5-arylidene-thiazolidine-2,4-dione-3-acetic acids 91 ). Compound 91 and related derivatives are reported to have low or no activity on PPAR, but have high antidiabetic activity in vivo (confirmed in sucrose-loaded model) [223] . This indicates other mode of antihyperglycemic action of 5-ylidene-4-thiazolidinones which can be treated as glitasone bioisosters. 5-Arylidene-3-benzyl-2,4-thiazolidinedione derivatives ( 92, 93) possess antihyperglycemic activity opposed to troglitazone under the experimental conditions (evaluation of in vitro insulin releasing activity, INS-1 cells); the latter has not decreased the level of hyperglycemia [224] ( Scheme 45 ). Structurally related 5-ene-2,4-thiazolidinediones 94 with bulky C5 fragments are treated as potent euglycemic and hypolipidemic agents too [225] as well as their simple analog 95 (the alloxan-induced hyperglycemia, in vivo mice model) [226] ( Scheme 46 ). Scheme 45 5-Ene-2,4-thiazolidinediones with antihyperglycemic activity. Scheme 45 Scheme 46 5-Ene-2,4-thiazolidinones based hypolipidemic agents. Scheme 46 However, studies of PPAR-related actions of 5-ylidene-4-thiazolidinones are continuing. Virtual screening tools including SQUIRREL (Sophisticated QUantification of InteRaction RELationships) procedure allowed detecting a set of potential PPARs ligands [227] . 14 Compounds including 4-thiazolidinones 96 – 99 showed PPARα and PPARγ agonistic activity (lead-compound 99 possessed the nano-level of activity (EC 50 PPARα = 0.044 μM)) ( Scheme 47 ). Scheme 47 5-Ene-4-thiazolidinones with PPARα/γ agonistic activity. Scheme 47 A new series of PPARγ agonists known as (β-carboxyethyl)-rhodanine derivatives were identified based on ligand-centric and receptor-centric approaches. An in vitro assay had confirmed the nanomolar binding affinity of 100 . In a cell-based transactivation assay similar PPARγ agonistic activity to that of the known PPARγ drug, pioglitazone, was shown. Based on CoMFA model and docking data it was discovered that the electrostatic interaction of the carboxyethyl group with the rhodanine core is important for its binding in the pocket of PPARγ and the rhodanine heterocycle played a different role than the thiazolidine group of rosiglitazon [228] . Currently the investigation of PPARγ antagonists is also of the great interest in the treatment of diabetes and obesity [229] , [230] including study of the 5-ene-4-thiazolidinones. The most investigated mechanism of antidiabetic mode of action of 5-ene-4-thiazolidinones is the inhibition of aldose reductase ( AR ). Aldose reductase is the limiting enzyme of polyol/sorbitol pathway of glucose oxidation, excessive activation of which leads to the accumulation of glucitol and the development of diabetic complications [231] . 5-Ene-rhodanine-3-alkanecarboxylic acids are high affinity inhibitors of aldose reductase [232] . The illustrative example of the mentioned compounds' row is epalrestat – ( Z,E )-5-(2-methyl-3-phenyl-2-propenylidene)-2-thioxo-4-thiazolidinone-3-acetic acid. Optimization of these compounds structures is mainly associated with C5 fragment modification [53] , [233] , [234] , [235] . Study of epalrestat analogs revealed that Z -isomers possessed higher activity level. There is a great interest in 2,4-thiazolidinedione derivatives as AR inhibitors since they can be viewed as hydantoin and rhodanine bioisosteres potentially free of the hypersensitive reactions which are linked to the presence of the hydantoin system [55] , [236] , [237] . In fact, to date, several thiazolidine-2,4-diones have been patented with dual activity as anti-hyperglycaemic and AR2 inhibitory agents [238] , [239] . The findings suggest that the activity of these compounds might correlate with their AR2 inhibitory ability by preventing the stimulation of PKC or MAPK and the subsequent activation of NF-kB [240] . AR2 plays a pivotal role in mediating oxidative stress-induced inflammation and is implicated in the development of various inflammatory pathologies [232] (see below). This enzyme catalyzes the reduction of lipid peroxidation-derived aldehydes, thus producing metabolites which transduce inflammatory signaling by means of the activation of protein kinases such as PKC and MAPK. This in turn activates NF-kB that is responsible for the transcription of many proinflammatory genes [241] , [242] . Antidiabetic activity of target compounds is often related to the inhibition of PTP 1B (protein tyrosine phosphatase 1B), which is an intracellular PTP and a key negative regulator of the insulin signaling pathway as well as to LMW-PTP (low molecular weight protein tyrosine phosphatase). This possible mode of action may be regarded as an attractive approach to the design of new therapeutic agents for the treatment of type 2 diabetes mellitus, obesity and, therefore, of the states associated with complex metabolic disorders known as metabolic syndrome. 4-(5-Arylidene-2,4-dioxothiazolidin-3-yl)methylbenzoic acids 101 were indicated as inhibitors of both PTP 1B and LMW-PTP ( Scheme 48 ). One of the directions in this study is the example of utilization of phosphotyrosine-mimetics to identify effective low molecular weight nonphosphorus inhibitors of PTPs ( p -methylbenzoic acid residue at N-3 position of the 5-arylidene-2,4-thiazolidinedione scaffold can act as a monoanionic pTyr-mimetic group replicating the interactions of pTyr with the catalytic site of the enzyme) [243] . In this study the authors also noted the importance of the 5-arylidene fragment. Alkylated/acylated phenolic groups and methoxy groups are desirable for the inhibitory effect. Scheme 48 5-Ene-4-thiazolidinones with antidiabetic activity. Scheme 48 5-Substituted 2-cyanimino-4-thiazolidinone ( FPFS-410 ) – a compound related to pioglitazone (PPAR-agonist) also possesses antidiabetic activity. In vivo data revealed the ability to reduce blood glucose and triglycerides levels and reduce the obesity [244] . Screening study allowed identifying and confirming the activity of 5-arylidene derivatives as inhibitors of glycogen synthase kinase-3 ( GSK-3 ) which, in turn, has been emerging as a key therapeutic target not only for type 2 diabetes, but also for Alzheimer'sdisease, cancer and chronic inflammation [245] . Simple 5-benzylidene-1,3-thiazolidine-2,4-dione derivatives 102 were presented as a new class of α-glucosidase inhibitors [246] ( Scheme 49 ). α-Glucosidases (α-D-glucosideglucohydrolases) are membrane bound exo-acting enzymes responsible for catalyzing the final step in the digestive process of carbohydrate metabolism. α-Glucosidases are the enzymes that hydrolyze O-and S-glycosyl residues and are involved in the biosynthesis and processing of oligosaccharide chains of N-linked glycoproteins in the endoplasmic reticulum [247] . Scheme 49 Structures of α-glucosidase and glycogen phosphorylase inhibitors. Scheme 49 5-Benzylidenethiazolidine-2,4-diones carrying 2,3-dihydrobenzo[1,4]dioxine fragment 103 were proved to be glycogen phosphorylase inhibitors ( Scheme 49 ) ( glycogen phosphorylase is a key enzyme in the regulation of blood sugar level and it catalyzes the formation of glucose-1-phosphate from glycogen) [248] that may be relevant to the control of blood glucose concentrations in type 2 diabetes [249] . 3.4 Anticancer agents Search for new anticancer agents is the most dynamically developing area of medicinal chemistry. A lot of papers and patents are devoted to the search for efficient anticancer agents among 4-thiazolidinones, including 5-ene derivatives (selected row of patents see) [5(ESI),250]. Research in this area could be divided into the next groups: i) in vitro screening of highly active/selective hit-compounds for further optimization, mainly with unknown mode of action (including the international programs, mainly Developmental Therapeutics Program, NCI, NIH - https://dtp.cancer.gov ); ii) design of the high-affinity ligands to the "validated" anticancer biotargets; iii) creation of the hybrid molecules that combine several pharmacologically attractive scaffolds [158] , [159] , [251] , [252] ; iv) search for antitumor agents among the compounds with known biological activity (anti-inflammatory, antidiabetic, anti- or prooxidant etc.). 4-Thiazolidinones are known to possess good activity against different types of cancer including relatively simple 5-ene derivatives as well as complex or hybrids/conjugates bearing non-fused 5-ene-4-thiazolidinone fragment. Mentioned compounds don't belong to any "classic" anticancer agent types [8] , [94] , [253] , [26] , [27] , [254] , [255] , [256] . Despite the diversity of 5-ene-4-thiazolidinones, the search for antileukemic agents is one of the most promising directions. It was found the tendency of the maximum sensitivity of leukemia cell lines to various subtypes of thiazolidinones [8] , [25] , [26] , [27] , [34] , [35] , [66] , [94] , [138] , [142] , [143] , [162] , [100] , [257] , [258] , [259] , [260] . The crucial role of the C5-(ylid)ene fragment is also confirmed and complication of ene-fragment is considered as a benefit. But anti-leukemic effects of such compounds have been less documented in comparision with all other cancer types [261] . Among 4-thiazolidinone-3-carboxylic acids derivatives, amides 104 – 107 ( Scheme 50 ) with anticancer activity were identified including samples possessing strong antileukemic activity ( 107 ). The SAR study revealed that anticancer activity has significantly decreased or disappeared after transformation into related isosteric compounds (replacement of C5 arylidene moiety by the C5-carboxymethylidene or 5-carboxymethyl fragments) or in case of compounds based on the related heterocyclic cores (e.g. 2,4-imidazolidinedione) [26] , [94] , [253] Moreover, novel 4-thiazolidinone-3-carboxylic acid amides 107 bearing furan moiety exhibited significant cytotoxicity and induction of apoptosis in human leukemia cells [259] . Scheme 50 Structures of 5-ene-4-thiazolodinones with anticancer activity. Scheme 50 Among 5-arylidene-2-arylamino(imino)-4-thiazolidinones 108 a set of active compounds with micromolar IC 50 levels has been detected (log GI 50 ∼ −5.77) [143] . Based on the obtained results the compounds with bulky C5-ene fragment and benzothiazole core were designed: compound 109 (pGI 50 = 4.97) selectively inhibited growth of the HOP-92 cell lines (CNS cancer, pGI 50 = 6.34) [138] . Study of the anticancer activity of 4-thiazolidinones with pyrazoline moiety in the C2 position of the main core has revealed high anticancer potential of the mentioned compounds 110 ( Scheme 50 ). SAR study confirmed the dependence of anticancer activity from the structure of C5 fragment [142] . The isomeric 4-amino thiazolidinones possessed much less activity levels [25] . Thiazolidinone-isatin conjugates are good examples of polycyclic compounds with high anticancer potential and their design illustrates the molecular hybridization approach when two "pharmacophores" are combined into a single molecule [262] . For instance, 111 possesed micromolar activity level and also selectively inhibited the leukemia cell lines [27] . Further structure optimization led to increasing of the anticancer activity which was reflected in the design of thiazolidinone-isatin-pyrazole hybrids 112 , 113 [34] , [35] ( Scheme 51 ). Similar isatin-pyrazoline conjugates (without thiazolidinone core) didn't show the anticancer activity. Scheme 51 Thiazolidinone-isatin conjugates with anticancer activity. Scheme 51 Another example of anticancer isatin-thiazolidinone hybrides are compounds 114 [263] and 115 [264] ( Scheme 51 , Scheme 52 ). The latter are the result of the structure optimization of sunitinib (SU11248, Sutent TM; Pfizer Inc) which is currently used in the clinics as a multi-targeting tyrosine kinase inhibitor with antiangiogenic activity [265] , [266] . The same approach of molecular hybridization was used in the design of thiazolidine-2,4-diones 116 bearing α-bromoacryloylamido moiety [267] , [268] . Scheme 52 4-Thiazolidinones hybrids as anticancer agents. Scheme 52 Compound MKT077 is known anticancer agent [269] ( Scheme 53 ) with apoptosis related mechanism of action and is an illustrative example of above mentioned compounds, though, its study has been proceeding till now. Thus, it is shown that it acts through differential interaction with the Hsp70 ( Heat shock protein 70 kDa ) allosteric states and reactivation of p53 function. MKT-077 therefore can be treated as an "allosteric drug" [270] , [271] , [272] , [273] . Scheme 53 Structure of MKT 077. Scheme 53 One more example of the hybrid approach in anticancer agents design is the combination of the natural compound, e.g. oleanane scaffold, with thiazolidinone core ( Scheme 54 ). Among oleanane-thiazolidinone hybrids 3-(2,4-thiazolidinedione-5-ylidene)-carboxyimino]olean-12-en-28-oic acid methyl ester ( 117 ) was identified as the most active substance (pGI 50 = 5.57, pTGI = 5.13 and pLC 50 = 4.64 NCI NIH protocol) with low toxicity and moderate activity level in in vivo Hollow Fiber Assay [162] . Another example of this approach is 5-arylidene-2,4-thiazolidinedione bearing cholesterol fragment 118 . Moreover, the level of activity and selectivity of antimitotic effects depends on the substituents in the C5 position of thiazolidinone core [161] . The maximum effect such compounds exhibited towards HeLa cancer cells line. Scheme 54 Combination of natural compounds and 5-ene-4-thiazolidinones fragments. Scheme 54 Promising active molecules based on the combination of thiazolidinone and chromene cores 119 were also found [160] ( Scheme 55 ). Scheme 55 Combination of thiazolidinone and chromene cores. Scheme 55 Once again it should be emphasized that for majority of mentioned compounds the tendency towards maximum sensitivity of leukemia and lung cancer cell lines is shown. This, most probably, may be a platform for the effective anti-leukemic agents design. For example, 5-isopropylidene-3-ethylrhodanine and 5-benzylidene-3-ethylrhodanine demonstrated cytotoxicity towards leukemic cell line CEM by inducing apoptosis [38] , [274] . Besides this, the modification of known drugs, for example assorafenib, with 5-ene-4-thiazolidinone fragments 120 is reported [275] , [276] ( Scheme 56 ). Scheme 56 Modification of known drug with 5-ene-4-thiazolidinone fragment. Scheme 56 Whereas, a lot of potential biotargets for new anticancer agents design are known at present, there have been almost no attempts to systematize the 4-thiazolidinones as anticancer agents in the review papers. For example, three aspects of the PPARs -independent antitumor activities of thiazolidinones had been outlined [10] : i) inhibition of Bcl-2/Bcl-x function, ii) proteasomal degradation of target proteins, iii) transcriptional repression of AR through Sp1 degradation [277] . Though, the latter definitely does not represent the whole spectrum of experimental data and is not enough for the design of novel hit- and lead-compounds. Taking into account the great number of papers, we tried to outline the most referred biotargets and their ligands among 5-ene-4-thiazolidinones. It should be noted that the majority of high-affinity ligands belong to 4-thiazolidinone-3-carboxylic acids, 5-benzylidenerhodanines, 2-amino(imino)derivatives and 2,3-disubstituted-4-thiazolidinones. Based on the established role of PPARs in the cancer and inflammation progress [219] , [278] , [279] , [280] considerable part of investigations is dedicated to the study of glitazones and related derivatives as possible anticancer agents. However, 5-ene analogs as synthetic precursors of glitazones that do not contain exocyclic double bond in the C5 position possess less expressed affinity to PPARs as their agonists. PPARγ antagonists in turn represent a new drug class that holds promise as a broadly applicable therapeutic approach for cancer treatment [229] , [281] . Among them 5-ene derivatives occupy their deserved place. For instance, the novel 3-thiazolidine-modified benzoic acid derivative HL005 ( Scheme 57 ) being a potent PPARγ-specific antagonist inhibits the proliferation of the MCF-7 cell line at the concentration-dependent manner, induces cell cycle arrest at the G2/M phase and interferes with cell adhesion [282] . Scheme 57 Structure of HL005 – PPARγ-specific antagonist. Scheme 57 PPARs indepented anticancer effects of troglitazone are mediated mainly via the repression of cyclin D1 (MCF-7 breast cancer cells) by facilitating proteasome-facilitated proteolysis [283] and partial depletion of intracellular Ca 2+ stores that leads to inhibition of translation initiation [284] . The troglitazone structure modification led to the STG28 identification – the first small-molecule agent mediating the proteasomal degradation of cyclin D1 with high specificity (exposure to STG28 did not cause any appreciable change in the expression levels of a series of other cyclins and CDK-dependent kinases ) [283] ( Scheme 58 ). Scheme 58 Structure of 5-ene-thiazolidinones with anticancer activity derived from troglitazone structure modification. Scheme 58 Following the structure optimization and screening data it was shown that the 5-arylidenerhodanines 121 are equipotent to troglitazone in Ca 2+ releasing activity, induction of eIF2a phosphorylation and more potent in inhibiting cancer cells proliferation [285] . Further modification allowed to state that a series of 5-benzylidene-2,4-thiazolidinediones and -thiones inhibited cell growth at low micromolar concentrations via the inhibition of translation initiation which involves partial depletion of intracellular Ca 2+ stores and strong phosphorylation of eIF2a . Structurally similar benzylidenerhodanines 122 showed high inhibition of protein tyrosine phosphatase PRL-3 (IC 50 = 0.9 μM), which is one of the probable prognostic markers of metastatic cells [286] . Among 5-ene-rhodanines 123 was identified as a promising and selective inhibitor of enzymes of dual-specificity phosphatases group – phosphatases JSP-1 ( JNK-stimulating phosphatase-1 ). Mentioned compounds are as well perspective agents that can be explored in the treatment of the inflammatory disorders [140] ( Scheme 59 ). Scheme 59 3-[5-(4-Fluorobenzylidene)-rhodanin-3-yl]-benzoic acid - JSP-1 inhibitor. Scheme 59 Necrostatin-7 ( Nec-7 ) ( Scheme 60 ) and related heterocycles belong to new class of "small molecules" – inhibitors of necroptosis that is regulated caspas-independent pathway of the cell death, morphological features of which are close to necrosis. This may be used in FADD-changed variant of the treatment of JurkatT cancer cells under the use of TNFα [287] , [288] . Scheme 60 5-[3-(4-Fluorophenyl)-1 H -pyrazol-4-ylmethylene]-2-imino-3-(thiazolyl-2)-4-thiazolidinone – necroptosis inhibitor. Scheme 60 On the other hand, derivatives of hydantoin-5-acetic acid are non-hydroxamate inhibitors of TNFα converting enzyme [289] that allows to treat them as promising anti-inflammatory agents and gives the prospects to establish anticancer potential of the given class of compounds. Moreover, the ability of hydantoincarboxylic acids, especially with the thiazole fragments, to inhibit activity of Ras farnesyl transferase (Ftase) was proved [290] . Ras proteins play an essential role in the processes of cell growth and differentiation and need post translational modification including farnesylation catalyzed by Ras farnesyl transferase . That is the reason why Ras-Ftase inhibitors are considered as potential anticancer agents [291] , [292] , [293] . Structural analogs of the given substances are the ligands for neuro-immunophilin FK506-binding protein (FKBP) [294] . 5-Substituted rhodanine-3-carboxylic acids 124 ( Scheme 61 ) are the inhibitors of protein-protein interaction of antiapoptotic proteins of the Bcl-2 and Bax family and their binding to the appropriate receptor domains [257] , [295] , [296] , [297] , [298] , [299] . Scheme 61 5-Arylidenerhodanine-3-carboxylic acids - Bcl-2 inhibitor. Scheme 61 Apart from the directions shown above, 5-aryl(hetyryl)idenethiazolidine-2,4-diones 125, 126 ( Scheme 62 ) were identified as potent and selective insulin-like growth factor-1 receptor (IGF-1R) inhibitors [230] . Scheme 62 Structure of insulin-like growth factor-1 receptor inhibitors. Scheme 62 IGF-1R is a growth factor receptor of tyrosine kinase family, acting as a critical mediator of cell proliferation and survival. Although, being highly related to insulin receptor, it plays a different role in organism development, being responsible for normal growth and development as opposed to glucose homeostasis. Epidemiological studies indicate that the IGF-1R is overexpressed in human cancer and is primarily responsible for tumor genesis. Signaling through IGF-1R includes the activation of PI3K and Raf pathways [300] . Inhibition of both these pathways makes IGF-1R kinase a promising target for cancer therapy. Moreover, series of 3,5-disubstituted thiazolidine-2,4-dione analogs [301] and 3-(2-aminoethyl)-5-(3-phenylpropylidene)-thiazolidine-2,4-dione [302] were shown to be potential anticancer agents via the inhibition of the Raf/MEK/ERK and PI3K/Akt signaling cascades. The inhibitors of extracellular signal-regulated kinases-1 and 2 (ERK1/2) – (e.g. ( Z )-3-(2-aminoethyl)-5-(4-ethoxybenzylidene)thiazolidine-2,4-dione 127 ) are the examples of compounds with greater selectivity for inhibiting the proliferation of melanoma cells [303] ( Scheme 63 ). Scheme 63 Structure of signal-regulated kinases inhibitor. Scheme 63 Polo-like kinase 1 (Plk1) is a key regulator of mitotic progression and cell division. In eukaryotes it acts in concert with cyclin-dependent kinase 1 – cyclin B1 and Aurora kinases to conduct a wide range of critical cell cycle events. Because Plk1 has been preclinically validated as a cancer target, small-molecule inhibitors of Plk1 have become attractive candidates for anticancer drugs development [304] . Thiazolidinone 128 ( Scheme 64 ) selectively inhibits human Plk1 (IC 50 19 nM) and various human and mouse tumor cell lines (IC 50 0.2–1.3 μM) and cause a prometaphase-like mitotic arrest [305] . Scheme 64 Structure of Polo-like kinase 1 inhibitor. Scheme 64 Simple 5-ene-rhodanines ( 129 , 130 ) ( Scheme 65 ) may also be considered as "Myc-Max compounds" that inhibit or reverse the association between c-Myc and its obligated HLH-LZ heterodimerization partner Max [306] , [307] , [308] . Scheme 65 Simple 5-ene-rhodanines as "Myc-Max compounds". Scheme 65 Among 5-ene derivatives of 2-iminothiazolidine, 2,4-thiazolidinedione and rhodanine the estrogen-related receptor-α (ERR-α) modulators were identified. They can be useful for the prophylaxis or treatment of ERR-α associated diseases such as breast cancer [309] , [310] . Following the modification of 5-(3-trifluoromethylbenzylidene)thiazolidine-2,4-dione (high selectivity Pim-1 inhibitor) a series of substituted thiazolidine-2,4-dione derivatives were identified as highly active and selective Pim-1 and Pim-2 inhibitors (nanomolar values of IC 50 for Pim-1 and ∼2.0 μ M for Pim-2) [54] . Pim-1 and Pim-2 are serine/threonine protein kinases frequently overexpressed in prostate cancer and certain forms of leukemia and lymphoma [311] ; Pim can phosphorylate the proapoptotic protein BAD (the Bcl-2-associated death promoter) leading to sequestration of 14-3-3 proteins and inhibits the apoptosis [312] , [313] . As it was already mentioned, 5-ylidene-2-amino(imino)-4-thiazolidinones 131 ( Scheme 66 ) comprise one of the most investigated thiazolidinone subtype as compounds with anticancer activity [138] , [142] , [143] , [314] , [315] . Among them the derivative MMPT was identified as a hit-compound being able to effectively inhibit the growth of some lung cancer cell lines (H460 and H460/TaxR) and at the same time do not influence normal fibroblasts in a dose response manner [151] , [316] , [317] . Study of the structure-activity relationship in a group of 372 compounds revealed some structural peculiarities: i) the nitrogen atom of thiazolidine core has to be unsubstituted; ii) dimethylaminobenzylidene fragment is the best in the C5 position; iii) 2-phenylamino moiety may contain substituents in different positions of aromatic ring [151] . Scheme 66 Structure of simple 5-ylidene-2-amino(imino)-4-thiazolidinones with significant anticancer activity. Scheme 66 5-[(4-Methylphenyl)methylene]-2-(phenylamino)-4(5 H )-thiazolone ( MMPT ) and 5-(2,4-dihydroxybenzylidene)-2-(phenylimino)-1,3-thiazolidine ( DBPT ) are effective in the treatment of multidrug-resistant (MDR) cancer. Multidrug-resistance is a phenotype of cross-resistance to multiple drugs with diverse chemical structures. One of the well-documented MDR mechanisms is the overexpression of the MDR-1gene that encodes the transmembrane, ATP-dependent drug efflux transporter P-glycoprotein (P-gp) in response to chemotherapy [318] , [319] , [320] . P-gp prevents the intracellular accumulation of many cancer drugs by increasing thei refflux out of cancer cells as well as through hepatic, renal, or intestinal clearance pathways [319] . Attempts to coadminister P-gp modulators or inhibitors to increase cellular availability by blocking the actions of P-gp have met with limited success [321] , [322] , [323] . Therefore, a more promising approach lies in the design and discovery of novel compounds that are not substrates of P-gp and are effective against drug-resistant cancer while at the same time exhibit minimal toxicity to normal cellular functions. MMPT inhibited the growth of human non-small-cell lung cancer and colon cancer cells (drug-sensitive (H460) and drug-resistant (H460/TaxR cell lines) independently of P-glycoprotein and p53 status [151] , [316] . Besides this, it selectively killed drug resistant cancer cells and induced apoptosis [317] . Among 5-ylidene-2-arylamino-4-thiazolidinone derivatives effective growth inhibitors of HT29 cell line (with a high COX-2 expression) were identified [324] . Among mentioned 4-thiazolidinone derivatives efficient CDK1 132 [325] , CDK1/cyclinB inhibitors [326] , [327] , [328] and SHP-2 inhibitors 133 and 134 ( Scheme 67 ) ( SHP-2 is a non receptor protein tyrosine phosphatase that mediates cell signaling by growth factors and cytokines acting via the RAS/MAP kinase pathway) [329] , [330] were discovered. Benzo[ d ]thiazole derivatives turned out to be more potent than their thiazole analogs. Scheme 67 Structures of CDK1 and SHP-2 inhibitors. Scheme 67 2-Substituted thiazolidinone and oxazolidinone derivatives were established as the inhibitors of phosphatases and anticancer agents [331] . Modifying the imidazolidine derivatives the new cell division cycle 7 kinase inhibitors were designed 135 and ( Z )-2-(benzylamino)-5-(1 H -pyrolo2,3- b ]pyridin-3-ylmethylene)-1,3-thiazol-4(5 H )-one] was selected as a lead compound [332] ( Scheme 68 ). Scheme 68 Structure of cycle 7 kinase inhibitors. Scheme 68 2-N,N′-Disubstituted diamines bearing 5-arylidene-4-thiazolidinone moiety 136 ( Scheme 69 ) had shown nanomolar inhibition potency (IC 50 40 nM) towards tyrosine phosphorylation-regulated kinases 1A [333] . This result prompted to explore the symmetric 1,2-diamino-linker grafted on N-3 position of two different 5-arylidenerhodanine platforms in order to modulate potential biological activity and led to the synthesis of unsymmetrical linked bis-5-arylidenerhodanine derivatives with anticancer effects [334] . Scheme 69 Structure of tyrosine phosphorylation-regulated kinases 1A inhibitors. Scheme 69 One of the established modes of action of 4-thiazolidinones anticancer activity is their antagonistic activity towards α v β 3 receptors . The value of inhibition of the α v β 3 activity is proved by the establishing of the correlation between progression of factor's expression and cancer growth. Search for effective inhibitors was held based on the pharmacophore model and virtual screening of 600000 compounds following by further confirming of the activity in vitro . This allowed to select promising compounds including 4-azolidinones. 2-Substituted 4-thiazolidinone derivatives among selected ones appeared to be efficient integrin α v β 3 antagonists 137 [335] ( Scheme 70 ). Scheme 70 General structure of integrin α v β 3 antagonists. Scheme 70 One of the aspects of anticancer effects realization of 4-azolidinone-3-alkanecarboxylic acids is the ability to influence the oxidative homeostasis of the cancer cells and the impact on the reactive oxygen species production. The modulating effect of the 5-ene-rhodanines on the proliferation and apoptosis of the cancer cells was identified as well as the influence on metabolism of NO alone with antioxidant and immunomodulating actions [336] . In this context interesting is the combining anticancer activity with anti-inflammatory effects; with antioxidant and/or other types of activity in some of the compounds [337] that is important in the development of "classic triad": "oxidative stress – inflammation – cancer". Additionally, the pharmacological profile of 5-arylidene-4-thiazolidinones derivatives is related to free radicals scavenging activity [106] , [338] , [339] . Besides this, notes about prooxidant effect of thiazolidinones with anticancer activity are frequently found in the papers. It has been shown that the derivatives of 5-ylidene-2-imino-4-thiazolidinone increased levels of reactive oxygen species (ROS) (HT29 cells), moreover, the ROS increasing was blocked by the ascorbic acid addition [340] . 5-Benzilidene-3-ethylrhodanine treatment led to increased level of ROS production and DNA strand breaks suggesting activation of apoptosis for induction of the cell death (leukemic cell line, CEM) plus by inducing a block at S phase [274] . Cell exposure to the selected compound was associated with the production of ROS, and the induction of autophagy which could lead to cell death [341] . Recent observation has been made in favour of the alternative chemotherapeutic strategy against cancer that consists in increasing the production of reactive oxygen species [340] , [342] . Moreover, generation of ROS could activate the NF-2E related factor 2/Kelch like ECH-associated protein 1 ( Nrf2/Keap1 ) pathway involved in protection of the cells against oxidative stress [343] . Activation of the Nrf2/Keap1 pathway leads to the stabilization of Nrf2 which translocates in nucleic db in ds to Maf protein. The complex, after binding to specific DNA sequence, defined as antioxidant response element localized in the promoter of Nfr2/Maf target genes, increases the transcription of genes involved in reactive species-mediated response such as HemeOxygen-1 (HO-1) , NADH quinoneoxidase-1 (NQO1) or glutamylcysteyl ligase. It could also indicate the indirect impact of the given class of compounds on the pro/antioxidant balance [156] . ROS play a key role in mitochondria mediated apoptosis. Mitochondria are the prime source of ROS which are byproducts of aerobic respiration [344] , [345] , [346] . High levels of ROS in mitochondria can result in free radical attack of membrane phospholipids and cause mitochondrial membrane depolarization. This is an irreversible step associated with the release of mitochondrial factors triggering caspase cascades [347] , [348] . On the other hand, immunoblotting analysis showed that 5-(4-hydroxy-3-dimethoxybenzylidene)-2-thioxo-4-thiazolidinone (RD-1) significantly elevated the Parkin and Miro 2 expression levels in acute MPTP treated mice and improved mitochondrial membrane potential and ATP synthesis in MPP + -treated Neuro-2a cells. Moreover, RD-1 attenuated impaired mitochondrial transport and vesicle release dysfunction has been evoked by MPP + cytotoxicity in cultured primary mesencephalic neurons. Taken together, the results indicate that improving the mitochondrial dysfunction may be a good choice to delay the neurodegenerative progression commonly associated with pakinson diseases as well as cancer [349] . Among the proposed mechanisms for the antitumor effect of target compounds apoptosis induction, cell cycle arrest and differentiation have been extensively reported [38] , [66] , [252] , [274] , [277] . Anticancer effect of such compounds can be also achieved by blocking the cell cycle progression at the G2/M phases border (not affect the G1 to S phase transition) in reversible manner and by the induction of apoptosis [252] , [325] . FACS (fluorescence-activated cell sorting) analysis of 5-ylidene-rhodanine-3-carboxylic acids, which possessed strong antiproliferative activity against human leukemia cells, showed a remarkable accumulation of subploid cells, the sub G1 phase (G0/G1) followed by the decline of both G1 and G2/M phases. Morerover, the presented findings suggest that the observed growth inhibition could be due to apoptosis [259] . The related 5-ylidene thiazolidinones with heterocyclic fragments induced cell cycle arrest of leukemic cells at G2/M phase and induced cell death that resulted in increased level of SubG1 phase population, thereby affecting its proliferation and causing depolarization of mitochondrial membrane potential [260] . The similar results were found for 2-heterylamino 4-thiazolidinones that induced cell accumulation at G2/M and in sub G0/G1 phases of the cell cycle. Furthermore, dissipation of mitochondria membrane potential was observed as well as red/ox changes in the treated cells (HT29 cell lines) [340] . Treatment of cancer cells with mentioned MMPT and DBPT led to a time-dependent accumulation of cells arrested in G2/M phase [316] . 5-Ene-4-thiazolidinones caused a prometaphase-like mitotic (G 2 –M) arrest (as the ATP-competitive inhibitors of Plk1) [305] , [350] . Despite the diversity of presented biotargets, the notes about thiazolidine derivatives being inductors of apoptosis in various cancer cells have become more frequent [351] , [352] , [353] , [354] , [355] , [356] , [357] . Above all, apoptosis-related mechanism is associated with the inhibition of Bcl-2/Bcl-x function. However, there has been described another feasibility of target compounds to affect diverse apoptotic signaling pathways, namely, induction of peroxisome proliferator-activated receptor-associated caspase-dependent [250] , [317] , [358] mitochondria-mediated apoptosis. 4-Thiazolidinones can decrease mitochondrial membrane potential which is one of the important mechanisms for the mitochondrial mediated apoptotic cell death [359] (an example of such compound is MKT 077 ) [269] , [271] . 3.5 Anti-inflammatory agents Progress in the anti-inflammatory agents search is reached via both high affinity ligand design and pharmacological screening of the compounds with unknown molecular mode of action. Anti-inflammatory activity of 4-thiazolidinones is associated, primarily, with their ability to inhibit isoforms of cyclooxygenase ( COX ) and lipoxygenase ( LOX ). The most successful representative of anti-inflammatory agents on the base of 5-ene-4-thiazolidinones is darbufelone – (5-(3,5-ditertbutyl-4-hydroxybenzylidene)-thiazol-4-one [33] , [360] , [361] ) ( Scheme 71 ) that has attracted broad attention as dual inhibitor of cellular prostaglandin and leukotriene production via cyclooxygenase activity inhibition including inhibition of prostaglandin endoperoxide synthase-1 (PGHS-1) and PGHS-2 [362] . Scheme 71 Darbufelone structure and directions of the structure modification. Scheme 71 Chemical structure of darbufelon has been considered as lead-compound for novel anti-inflammatory agents design till now. Optimization of the latter and application of the bioisostere concept allowed the elaboration of new bioactive compounds 138 (GS26) and 139 (GS28) ( Scheme 71 ) with significant anti-inflammatory activity as well as antiedematogenic and ulcerogenic activity in vivo [363] . Further studies of related compounds allowed identifying selective inhibitors of COX-2 as promising anti-inflammatory agents with fewer side effects. Obtained data and SAR analysis resulted in identifying the most active compound 140 ( Scheme 71 ) from the group of thiazolones and oxazolones with di- tert -butylphenol fragment. Compound 140 inhibited human recombinant COX-2 (IC 50 1.7 μM) and inhibited the COX-2 activity in cell line J774A.1 (IC 50 0.17 μM). This compound turned out to be inactive towards COX-1 at the concentration of 100 μM and did not inhibit the COX-1 in thrombocytes at 20 μM [360] . Among 5-ene-4-thiazolidinones CT-8 ( Scheme 72 ) was found to be potent inhibitor of 15-hydroxyprostaglandin dehydrogenase ( 15-PGDH ) [364] . 15-PGDH type I is NAD + -dependent (while Type II is NADP + preferred) cytosolic enzyme which catalyzes oxidation of prostaglandins to 15-ketoprostaglandins. Inhibition of this enzyme is related to prostaglandin E2 action and can lead not only to anti-inflammatory effects but, for instance, to reduce hair loss [365] . Structure-activity analysis indicated that the N-methylation of CT-8 abolished the inhibitory activity. It was also established that the nature of the moiety linked to benzylidenethiazolidine-2,4-dione through another linkage plays an important role in its inhibitory activity [364] . Further modification led to the new related compound 141 identification (IC 50 = 0.031 μM) [366] ( Scheme 72 ). Scheme 72 Structures of 15-hydroxyprostaglandin dehydrogenase inhibitors. Scheme 72 Another mechanism of anti-inflammatory activity has been the inhibition of phosphodiesterase (PDE4 and PDE2). Phosphodiesterase is responsible for the hydrolysis of secondary messanger cAMP, the level of which increases under inflammation processes. Besides this, phosphodiesterase inhibitors may be effective agents in astma and obstructive lung diseases treatment. It has been established the ability of 5-arylidenerhodanines, including 3-substituted carboxylic acids [367] to inhibit above mentioned enzyme. Moreover, unlike other rhodanine derivatives, carboxylic derivatives selectively inhibit only phosphodiesterase-4 that may be an argument for their study as probable agents for Parkinson's and Alzheimer's diseases. The antigenerative effect of 5-benzylidene-4-thiazolidinones with benzothiazole 142 and isothiazole 143 moieties ( Scheme 73 ) on human chondrocytes culture was estimated when studying potent inhibitors of metalloproteinase (MMP). Given experimental model reproduces the mechanisms involved in osteoarthritis [240] , [368] . The compound 143 (Ar = 4-MeO-C 6 H 4 ) showed MMP-13 inhibition activity at nanomolar concentration (IC 50 0,036 μМ). Scheme 73 Structures of 15-hydroxyprostaglandin dehydrogenase inhibitors. Scheme 73 Moreover, among anti-inflammatory 5-ene-4-thiazolidinones a series of furan-2-ylmethylene-2,4-thiazolidinediones have been investigated as the ATP-competitive PI3Ks inhibitors 144 ( Scheme 74 ) ( IC 50 0.2–0.9 μM). Class I phosphoinositide 3-kinases ( PI3Ks ), in particular PI3Kγ , have become attractive drug targets in inflammatory and autoimmune diseases treatment [369] . An acidic NH group in the thiazolidinedione core and a hydroxyl group in the furan-2-yl-phenyl part of the molecule play crucial roles in binding to PI3K and contribute to class IBPI3K selectivity. Besides the compound 145 inhibited PI3Kγ (IC 50 33 nM) (but not PI3Kα , PI3Kβ ) dependent pathways inside cells and in a murine peritonitis model, it produced a similar decrease in leukocyte infiltration. Scheme 74 Structure of ATP-competitive PI3Ks inhibitors. Scheme 74 The anti-inflammatory properties of 5-arylidene-2-phenylimino-4-thiazolidinones 146 ( Scheme 75 ) are related to their ability to block the production or action of the degenerative factors induced by IL-1b [370] and possessing antidegenerative activity on human chondrocyte cultures. Scheme 75 5-Arylidene-2-amino-4-thiazolidinones with anti-inflammatory activity. Scheme 75 5-Arylidine-2,4-thiazolidinedione derivatives without substituent in the N3 position [371] as well as N3-substituted analogs 147 [372] ( Scheme 76 ) possessed anti-inflammatory activity and related type of activities, such as analgesic activity and in vitro antioxidant activity or even antimicrobial [167] . The structural affinity to PPARγ agonists ( in silico correlation studies and PPAR-competition binding assay) indicates the possibility of PPAR mediated mode of anti-inflanmatory action [372] . Scheme 76 5-Ene-4-thiazolidinones as active anti-inflammatory agents. Scheme 76 Similarly, 5-arylidene-2-imino-4-thiazolidinones 148 ( Scheme 76 ) exhibited anti-inflammatory effect (carrageenan-induced paw edema model) causing the reduction of PGE2 level together with insignificant COX-2 inhibition when compared with celecoxib. It should be noted that 5-unsaturated thiazolidinones possess much less expressed effect or are inactive [373] . Applying procedures of virtual and highthroughput screening for 2-(thiazole-2-ylamino)-thiazol-4-ones study led to identification of compounds 149 ( Scheme 76 ) with anti-inflammatory activity. In the in vivo cyclooxygenase and lipoxygenase inhibition assays as well as anti-inflammatory assay the compounds turned out to be promising agents for further studies in this field [374] . The row of hybrids 138 [363] ( Scheme 71 ), 150 [375] , 151 [376] ( Scheme 77 ) based on indometacine, roziglitazone and darbufelone molecules was designed. The molecules possesed in vivo anti-inflammatory activity in air pouch and peritonitis models or carrageenan induced paw edema model and inhibited the cyclooxygenase-1 and 2 . Scheme 77 5-Ene-4-thiazolidinones designed based on indometacine modification. Scheme 77 A number of related 3-substituted-5-arylidenethiazolidine-2,4-diones ( 152, 153 ) ( Scheme 78 ) was screened for anti-inflammatory effects. Compounds exhibited significant levels of anti-inflammatory activity in the assay of induced edema in mouse paws [377] . Scheme 78 5-Ene-4-thiazolidinones being active in induced edema paws model. Scheme 78 Biphenyl-4-carboxylic acid 5-(arylidene)-2-(aryl)-4-oxothiazolidin-3-yl amides 154 ( Scheme 78 ) had also shown significant activity in carrageenan test [378] . Application of polypharmacological approach and the advances in the pathophysiological role of various biotargets had resulted in the great interest in search for new anti-inflanmatory agents. For example, it was established that aldose reductase (AR) is critically involved in inflammatory processes under both normoglycemic and diabetic conditions. This enzyme, which is overexpressed under various oxidative conditions, intervenes in multiple signaling pathways leading to inflammation and tissue degeneration. Accordingly, it was ascertained that AR inhibition prevents multiple inflammatory pathways [379] , [380] , [381] and, therefore, new more effective and safer AR inhibitors were designed not only as antidiabetic (see above) but also antinflammatory agents [232] . 2-(Benzylamino)-5-((thiophen-2-yl)methylene)-thiazol-4-(5 H )-one showed the highest anti-inflammatory response on PBMCs (peripheral blood mononuclear cells) exerted through the NF-kB inhibition. This and related compounds also had antioxidant activity and xanthine oxidase inhibitory potency [152] . 3.6 Antiviral agents Investigation of antiviral activity of 4-thiazolidinone derivatives has been carried out mainly in two directions: search for anti HIV agents and search for agents used in hepatitis C treatment. Tested compounds are small molecule inhibitors of validated targets such as HCV NS3 and 5 proteases [39] , [382] , HIV RT [383] etc. Recently, rhodanine derivatives, namely 2-aryl-5-(4-oxo-3-phenethyl-2-thioxothiazolidinylidenemethyl)-furans 155 ( Scheme 79 ) were reported to exhibit anti-HIV-1 activity [384] . Scheme 79 5-Ene-4-thiazolidinone derivatives with antiviral activity. Scheme 79 Among 5-arylidenerhodanine derivatives the first small molecule able to inhibit HIV replication by targeting a cellular enzyme – the RNA helicases DDX3 had been identified [385] . The precise combination of functional groups on the rhodanine scaffold was shown to be responsible for the DDX3 inhibitory activity and selectivity of the hit compound 156 (FE15) ( Scheme 79 ). The 2-thioxo-4-thiazolidinones and salicylic acid containing compounds were the most potent HIV-1 integrase inhibitors ( HIV-1 integrase catalyzes the integration of proviral DNA in to the host genome, an essential step for viral replication) among several compounds retrieved from a database of small-molecules. Compounds 157 ( Scheme 79 ) inhibited strand transfer activities of HIV-1 integrase with similar IC 50 values (10–20 μM). The replacement of either the rhodanine or salicylic acid fragments is found to reduce HIV-1 integrase inhibitory potency [386] . Rhodanine-3-alkanecarboxylic acids, their amides and bicyclic analogs 158, 159 ( Scheme 80 ) represent a new class of inhibitors of HIV-1 Integrase as antiviral agents [387] . They were identified in in silico studies when modeling HIV-1 integrase inhibitors as well as in in vitro investigations. Scheme 80 Structures of inhibitors of HIV-1 integrase. Scheme 80 Employing HCV proteins as targets, directly acting antiviral agents were identified and collectively described as specifically targeted antiviral therapy for HCV (STAT-C) [388] , [389] , [390] . Most approaches to small molecule inhibitors search for Hepatitis C virus treatment have mainly focused upon inhibition of essential viral targets, such as the NS3-4A protease (analogous to HIV protease ), NS5B polymerase , NS3 helicase and NS5A [391] , [392] . A literature survey on HCV NS5B polymerase inhibitors clearly indicates that 4-thiazolidinones could inhibit this enzyme and might be promising candidates for the development of novel antiviral agents against HCV. Screening of in-house library allowed identifying derivative 160 ( Scheme 81 ) as hit-compound (IC 50 2.0 μM) acting towards NS5B-polymerase of HCV. Based on the results of rational design and virtual screening (GOLD docking) novel 2-imino-4-thiazolidinone derivatives were synthesized and compound 161 ( Scheme 81 ) with the IC 50 (3.0 μM) level close to that of 160 was identified [393] , [394] , [395] . Scheme 81 Structures of HCV NS5B polymerase inhibitors. Scheme 81 Further investigation led to N-substituted (aryl)alkylidene-rhodanines synthesis which inhibit HCV NS3 protease and also are good inhibitors of other serine proteases ( chymotrypsin and plasmin ). But, the selectivity of some compounds ( 162, 163 ) ( Scheme 82 ) with bulkier C5 fragments bearing hydrophobic functional groups as well as simple analogs ( 164 – 166 ) ( Scheme 82 ) was increased by ten fold towards HCV NS3 protease respectively [39] . Scheme 82 Structures of HCV NS3 protease inhibitors. Scheme 82 The 4-thiazolidinones from the groups of 2,3-diaryl-4-thiazolidinones [382] , [396] ; 2-amino-5-arylidene-4-thiazolinones [397] , [398] ; and 5-arylidene-3-substituted rhodanines [399] , [400] (167 – 171) ( Scheme 83 ) were the initial structures for new inhibitors of HCV NS5B polymerase design, mainly 2-heteroarylimino-5-arylidene-4-thiazolidinones. In all cases the 5-ylidenethiazolidininones inhibited NS5B at lower IC 50 values ranging between 19.8 and 64.9 mM. Moreover, the authors argued about the contribution of these (ar)ylidene groups in stabilizing the binding mode to NS5B active site. Scheme 83 5-Ene-4-thiazolidinones as inhibitors of HCV NS5B. Scheme 83 Further investigation led to new lead-compound identification 172 [401] . Thus the derivatives of 5-ene-rhodanine-3-acetic acids 173, 174 were active towards HCV inhibiting virus protease [402] . Additionally, it was established the ability of the derivatives of this compounds' class to inhibit human serine proteases . SAR analysis showed that substitution of arylidene moiety by the carboxy methylidene fragment led to insignificant loss of the activity ( Scheme 84 ). Scheme 84 Structures of HCV protease inhibitors. Scheme 84 Based on the previous data [403] a series of 2(4)-pyrazolyl-4(2)-thiazolidinones 175, 176 ( Scheme 85 ) were synthesized and their antiviral activity against Influenza viruses (Type A and B), Corona virus SARS, Tacaribe virus, Dengue virus, Rift Valley Fever virus, Respiratory Syncytial virus, Vaccinia virus and Venezuelan Equine Encephalitis virus was tested in vitro . Scheme 85 Structures of hit-compounds with antiviral activity. Scheme 85 Molecular hybridization method allowed obtaining thiazolidinone–peptide hybrids 177 ( Scheme 85 ) that inhibited Dengue virus protease . Moreover, it was determined that thiazolidinone core (2,4-thiazolidinedione and rhodanine) and the peptides should be accomplished by relatively rigid arylidene moieties using para-substitution [404] . 3.7 Anti- and prooxidant agents Violation of the antioxidant defence system and the balance of the pro/antioxidants are described under majority of pathologies including above mentioned. Therefore, a lot of papers are dedicated to the study of the 4-thiazolidones antioxidant activity [339] . Most of the articles describe investigation of the antiradical properties of 5-ylidenethiazolidinones in the model systems and evaluation of their efficiency as free radical scavengers (e.g. scavenging activity to DPPH (1,1-diphenyl-2-picrylhydrazyde) [106] ) combined with the study of other types of activity, in particular, antitumor and anti-inflammatory (see above) [405] . Such data concern all the classes of 4-thiazolidinones bearing different types of ylidene fragment. 4-Thiazolidinone derivative 178 ( Scheme 86 ) possesses anti-radical activity including the ability to inhibit superoxide anion formation [406] and it is known that antioxidant activity increases when additional carboxylic group is introduced in the substituent at the position C5 of the heterocyclic core in both N3 substituted and N3 unsubstituted derivatives [407] . Scheme 86 Structures of 4-thiazolidinones with anti-radical activity. Scheme 86 For the simple 5-benzylidene rhodanines antioxidant activity in various LDL oxidation models, such as TBARS assay, conjugated diene formation, REM of ox-LDL, fragmentation of apo β-100 by SDS-PAGE, radical DPPH scavenging activity, and macrophage-mediated LDL oxidation was showed [405] . It should be noted that the branched aliphatic fragments prevalence (spatially-screened phenols) largely contributes to the imitation of the known agents, e.g. BHT or probucol [336] . Antioxidant properties of a series of 2,4-dichlorothiazolylthiazolidine-2,4-dione and 4-chloro-2-benzylsulfanylthiazolyl-thiazolidine-2,4-dione derivatives 179 ( Scheme 86 ) were reported in two different in vitro assays: superoxide anion radical formation and DPPH radical scavenging activity [408] . The presence of the benzylsulfanyl group at the second position of the thiazole ring plays a significant role in increasing the superoxide anion scavenging activity. No correlation was observed in results on superoxide radical and DPPH radical scavenger capacity. Simple thiazolidinedione derivative ( Z )-5-(2,4-dihydroxybenzylidene)-thiazolidine-2,4-dione 180 ( Scheme 87 ) plays a crucial role in UV-induced melanogenesis, which is known to be related to the induction of tyrosinase enzyme . Compound inhibited nitroprusside-induced NO generation dose-dependently and suppressed tyrosinase activity and melanin synthesis (B16F10 melanoma cells) [409] . Scheme 87 Structures of 4-thiazolidinones with dual activity. Scheme 87 Series of 5-arylidene-2,4-thiazolidinediones 181 , 182 ( Scheme 86 ) and their 2-phenylimino analogs inhibiting AR [236] , [410] , [411] , [412] possess dual activity as antidiabetic and antioxidant agents for the treatment of diabetic complications [413] . The results indicate that latter possess excellent antioxidant properties inhibiting the production of TBARS in the test of compounds effect on hydroxyl radical-dependent lipoperoxidation induced in rat brain homogenate by the oxidant system Fe 3+ /ascorbic acid. It was observed that the presence of electron releasing substituents on the distal phenyl ring differentiates the antioxidant activity of the compounds. Detailed mechanisms of antioxidant effects of compounds of this class is still unclear and can be explained not only by the ability to inhibit superoxide anion (or other ROS) production, but also by influencing other parts of the pro/antioxidants system. Because of the changes in the classic approaches to the evaluation of pro/antioxidants [414] , emerging of new approaches to the interpretation of antioxidant effects [415] and study of other (not related to free radical scavenging) mechanisms of the antioxidants [156] , more attention is paid to the study of the molecular mechanisms of stimulation of antioxidant defence system [416] , [417] including the possibility of these substances pro-oxidant impact with subsequent stimulation of antioxidant defence system (see above – anticancer agents). 3.8 Other types of activity A large number of bio-targets of 5-ene-4-thiazolidinones are involved in different pathologies and there are a lot of data on the efficacy of 5-ene-thiazolidinones in different experimental models [5] , [6] , [7] , [100] , [418] . For instance, molecular docking studies indicated that the 5-ene-thiazolidinedione moiety was a likely candidate for its selectivity to monoamine oxidase-B . This hypothesis was confirmed by the identification of two lead-compounds 183 ( Scheme 88 ) with IC 50 13–20 μM [419] . Scheme 88 Examples of 5-ene-4-thiazolidinones with different types of activity. Scheme 88 Some 5-ene-rhodanine-3-carboxylic acid derivatibes, namely 4-(5 Z )-5-{5-(4-bromophenyl)-2-furyl]methylene}-4-oxo-2-thioxo-1,3-thiazolidin-3-yl-butanoic acid and 6-((5 Z )-5-{5-(4-bromophenyl)-2-furylmethylene}-4-oxo-2-thioxo-1,3-thiazolidin-3-yl)hexanoic acid are efficient inhibitors of apoptosis signal-regulating kinase 1 [420] which is involved in the variety of cellular processes [421] and can be considered as a promising target for the treatment of the various cardiovascular diseases, neurodegenerative diseases, several liver and kidney diseases etc. At the same time several other kinases, for example a cyclin-dependent kinase family are attractive targets for the design of high affinity inhibitors based on 5-alkylideenrhodanines [52] . 5-Ylidenes 184, 185 ( Scheme 88 ) were identified in the virtual screening as representatives of novel glycogen synthase kinase-3 ( GSK-3b ) inhibitors and their activity was confirmed experimentally [245] (ІС 50 1.56–5.56 μМ). GSK-3b has been emerging as a key therapeutic target for type-2 diabetes melitus, Alzheimer's disease, cancer and chronic inflammation. There is experimental evidence that GSK-3b inhibitors activate negative regulation of NF-kB activity, p53-dependent apoptosis, and enhance the TRAIL-induced cell death [422] , [423] . 3-(3-Trifluoromethyl)phenyl]-5-arylidene-2-thioxo-4-thiazolidinone was identified as a potent and selective blocker of the cystic fibrosis transmembrane conductance regulator ( CFTR ) Cl-channel . Moreover, it acted differently from other known blockers of the CFTR channel. Rather than blocking the channel pore, it affects the mechanism of channel gating, possibly by binding to nucleotide binding domain (NBD-1) [424] , [425] , [426] . CFTR , a member of the ATP-binding cassette transporter superfamily, is an epithelial chloride channel that plays a critical role in fluid absorption and secretion. Defective CFTR function causes cystic fibrosis, the most common lethal genetic disease in caucasians that produces severe lung disease, pancreatic insufficiency, neonatal intestinal obstruction and infertility where the hyperfunction of CFTR chloride channels , usually resulted from bacterial enterotoxins, constitutes the basic cause for secretory diarrhea. Thiazolidinone CFTR inhibitors may be useful in developing large-animal models of cystic fibrosis and in reducing intestinal fluid loss in cholera and other secretory diarrheas. Benzylidenerhodanine derivatives showed good inhibitory activity against recombinant human PRL-3 (phosphatase of regenerating liver, the representant of protein tyrosine phosphatases ). Compounds 186 and 187 ( Scheme 88 ) were the most active in vitro in this series and showed the ability to reduce the invasiveness of tumor ( 186 ) [286] . Following virtual screening of potent serotonin N-acetyltransferase (member of the GCN5N-acetyltransferase superfamily catalyzing the penultimate step in the biosynthesis of melatonin) inhibitors the class of 5-ene-rhodanine-3-carboxylic acids 188 ( Scheme 89 ) was identified that exhibited low micromolar competitive inhibition against acetyl-CoA and proved to be effective in blocking melatonin production in pineal cell [427] . Scheme 89 Structures of 5-ene-4-thiazolidinones with action on different targets. Scheme 89 5-Arylidene-4-thiazolidinone derivatives 189 ( Scheme 89 ) exhibited antidegenerative activity and could block multiple cartilage destruction during the osteoarthritic process by reducing NO release and restoring normal levels of glycosaminoglycans in chondrocytes treated with beta L-1, and possessed MMP-13 inhibition activity ( MMPs are a large family of calcium dependent, zinc containing endopeptidases responsible for the tissue remodeling and an extracellular matrix degradation) [240] . These compounds were found to be endowed with interesting activity levels in models of acute inflammation, such as carrageenan-induced paw and pleurisy edema in rats and to be inhibitors of COX isoforms [373] . 5-Benzylidene 2,4-thiazolidinedione derivatives are proposed for the treatment of Non-Alcoholic Fatty Liver Disease [428] . Compound 190 ( Scheme 89 ) was found to upregulate the adiponectin protein expression and down regulate the secretion of leptin protein in 3T3-L1 adipocytes at a respective concentration of 10 mM. 4-{( Z )-(2 Z )-2-(2-Fluorobenzylidene)-4-oxo-1,3-thiazolidin-5-ylidene]methyl}benzoic acid 191 ( Scheme 89 ) was described as the most potent full agonist to human GPR35 [429] . A class of orphan receptors GPR35 has been described as a potentially novel drug target [430] for the design of compounds with therapeutic application in a number of diseases including inflammation, metabolic disorders, nociception and cardiovascular diseases. The olefinic thiazolidinedione 192 ( Scheme 90 ) was found to be potent and selective b3-adrenergic receptor ( b3-AR ) (b3-Ar EC 50 0.006 mM, IA = 1.03) agonists (96-fold vs b2 and 492-fold vs b1 ) being more active than primary hits (5-saturated analogs) [431] . b3-AR Agonists are potential drugs for the treatment of obesity, type II diabetes, frequent urination and related diseases [432] . Related thiazolidinones with a cyanamide ( 193 ), hydroxylamine ( 194 ) ( Scheme 90 ) or 4-amino-1-benzylpiperidine substituents were generally very potent towards the b3 receptor, however, they were not very selective against both the b2 and b1 receptors. Related benzimidazolones ( 195 ) ( Scheme 90 ) displayed a good b3 agonist selectivity profile. Scheme 90 Structures of 5-ene-4-thiazolodinone b3-adrenergic receptor agonists. Scheme 90 5-Benzylidene-2-thioxo-4-thiazolidinones 196, 197 were described as inhibitors of Aggrecanase-2 that are members of the ADAMTS ( adisinte grin and metalloprotease with thrombospondin motifs ) family of zinc metalloproteases. The inhibition of Aggrecanase-2 , therefore, protect cartilage from damage and provide the first potential therapy to halt and/or reverse the progression of osteoarthritis and the compounds were at least 3-fold more active than 2,4-thiazolidinedione analogs [433] . In the result of virtual screening 198 ( Scheme 91 ) was found as potent canabinoid receptor type 1 antagonist with promising binding affinity (IC 50 125 nM) and also displayed good pharmacokinetic profile in rats as well as in human plasma [434] , [435] . Scheme 91 Structures of 5-ene-4-thiazolodinone b3-adrenergic receptor agonists and canabinoid receptor type 1 antagonist. Scheme 91 Novel class of sphingosine-1-phosphate ( S1P1 ) receptor agonists based on the 2-iminothiazolidin-4-one scaffold 199 and ACT-128800 ( Scheme 92 ) was found as the most active compound in in vitro and in vivo assays. Selective agonists of S1P1 receptor are of therapeutic interest for their ability to halt the exit of lymphocytes from lymph nodes. This interruption of lymphocyte migration promises a new immunomodulatory therapeutic principle for a variety of autoimmune diseases [436] . Scheme 92 Novel class of sphingosine-1-phosphate receptor agonists. Scheme 92 New class of benzisothiazolylimino-5-benzylidene-4-thiazolidinones was identified as new metalloproteinase-3 i nhibitors and chondroprotective agents. The mode of action is also related to the decreasing of NF-kB level [437] . Optovin ((5 E )-5-[(2,5-Dimethyl-1-pyridin-3-ylpyrrol-3-yl)methylidene]-2-sulfanylidene-1,3-thiazolidin-4-one) ( Scheme 93 ) was detected as a small molecule that enables repeatable photo activation of motor behaviors in wild type animals. Optovin acts as a light-sensitive ligand for anion channel involved in the detection of painful sensory stimuli. It activates human TRPA1 via structure-dependent photochemical reactions with redox-sensitive cysteine residues. Optovin treatment enables control of motor activity in the paralyzed extremities by localized illumination. These studies identify a light-based strategy for controlling endogenous TRPA1 [146] , [147] . It should be mentioned that optovin is a rare example of the E -isomers of 5-heterylidenerhodanines. Scheme 93 Structure of optovin. Scheme 93 Among rhodanine-3-carboxylic (sulfonic) acids the new inhibitors of histone acetyltransferases ( 200 – 202 ) ( Scheme 94 ) were identified. Histone acetyltransferases ( HATs ) are promising epigenetic drug targets involved in the pathogenesis of a wide range of diseases [427] , [438] . Scheme 94 Structures of inhibitors of histone acetyltransferases. Scheme 94 Summarizing all the above, 5-ene-4-thiazolidinones deservedly can be considered as privileged heterocycles with the wide spectrum of biological activity. Assigning 5-ene-thiazolidinones as PAINS due to possible Michael acceptor functionality is not so unambiguous and requires further study; therefore, most compounds can not be rejected per se . There are two main directions of the presented literature data: i) screening of the compounds activity without confirmed mode of action and ii) search for highly active and selective ligands to so-called "validated" bio-targets, which definitely don't negate each other. Search for antitumor, hypoglycemic, antimicrobial, antiviral and anti-inflammatory agents are the main fields of the 5-ene-4-thiazolidinones study. Moreover, most of the works are dedicated to the investigation of the antitumor potential of these compounds. Despite a series of established bio-targets, a detailed mode of anti-cancer effect undoubtedly depends on the structure of a particular derivative and needs further study. However, for the majority of 5-ene-thiazolidinones apoptosis-dependent as well as the ROS mediated modes of action are established. The main directions of the active 5-ene-thiazolidinones optimization can be devided as follows: i ) complications of the C5 fragment (bulky-substituents), while N3 unsubstituted position is preferred; ii ) introduction of substituents in N3 position (especially fragments bearing carboxylic group or its derivatives); iii ) annulation of 5-ene-thiazolidinones in complex fused heterocyclic systems; iv ) combination with other pharmacologically attractive fragments within hybride pharmocophore approach. 3.1 Antimicrobial agents One of the earliest directions of the 5-ene-4-thiazolidinones biological assays lies in the field of antibacterial and antifungal activities search. Apparently, the structural similarity of 4-azolidinones with penicillin antibiotics was the stimulus to the study of such type of activity [1] , [2] , [3] , [6] , [167] , [168] . Though, recently it was discovered that manifestation of antibacterial effect is not always related to the penicillins' mode of action. Usually, there is a tendency to move from the detection of antimicrobial activity [169] , [170] , [171] , [172] , [173] , [174] , [175] , [176] within screening programs to identification and design of high affinity ligands to the validated molecular targets. Regardless of the nature of the basic core (2,4-thiazolidinedione, rhodanine, etc.), modification and complication of C5 ylidene fragment are the benefits for realization of such type of activity (the role of the halogen-arylidene moieties should be stressed) [141] , [177] . The 3-substituted 5-ylidene-4-thiazolidinones, especially 3-carboxylic acids 55 – 59 [171] , [175] , [176] , [177] , [178] , [179] , [180] , [181] represent group of compounds with significant antimicrobial activity against gram-positive and gram-negative microorganisms. Their molecular mode of action is associated with the inhibition of: i) the last two stages of peptidoglycan cell wall biosynthesis (the most attractive targets for new antimicrobials design - mainly penicillin-binding proteins family , and transferases – UDP-Mur-NAc-peptidases ); іі) the activity of dehydrogenases (compound 58 and related structures can cross the bacterial cell wall and lead to the identification of catechol-rhodanine core as privileged scaffold for design of the dehydrogenases inhibitors) [153] ; iii) peptide deformylase [182] (for lead-compounds 60 , 61 IC 50 = 0.89 μM and 1.66 μM correspondly) etc ( Scheme 33 ). Scheme 33 5-Ene-4-thiazolidinone-3-carboxylic acids with antimicrobial activity. Scheme 33 A series of 5-ene-rhodanine-3-acetic acid derivatives 62 ( Scheme 34 ) were described as inhibitors of fungal protein mannosyl transferase 1 (PMT1) at micromolar levels (design of compounds was also based on the complications of C5 moiety) [183] . High-throughput screening of 5-arylidene-4-thiazolidinone-3-alkanecarboxylic acids 63 , 64 proved them to be anthrax lethal factor (as one of the major virulence factors) inhibitors (IC 50 ∼ 10 μМ) [45] , [184] . Scheme 34 5-Ene-rhodanines as antimicrobial agents. Scheme 34 Antimicrobial activity of such compounds are determined by inhibition of penicillin-binding proteins family , such as a group of transferases (including UDP-Mur-NAc-peptidases ), most of their subtypes are characterized by distinct transglycosylase activity [185] . ATP-dependent amino acid ligases (MurC, MurD, MurE, and MurF, MurG) are an illustrative examples of the 5-ene-4-thiazolidinones antibacterial agents design [41] , [186] , [187] , [188] . For instance, 65 is an efficient inhibitor of MurC ( UDP-N-acetylmuramate/ l -alanine ligase ) [139] and 66 – of MurG [189] . It should be emphasized that the 4-thiazolidinone and rhodanine cores are treated as novel phosphate mimics [190] ( Scheme 34 ). Structurally similar compounds are the inhibitors of dTDP-rhamnose synthesis . This is especially important for antimycobacterial agents search. The mycobacterial cell wall is unique because it contains an amycolylarabinogalactan layer bound to the peptidoglycan layer via a rhamnose–Glc-NAc sugar linker, where the dTDP-rhamnose can be vetreated as a precursor [191] , [192] . 5-Heterylidene-2,4-thiazolidinedione derivatives 67 are competitive inhibitors of recombinant bacterial arylamine-N-acetyltransferases (NATs) [193] , which were designed based on the structure modification of N-(2-naphthyl)-methyl substituted 1,1-dioxo-2,3-dihydrobenzo[1,2]-thiazine-4-ylidenethiazolidine-2,4-dione (weak inhibitor of NAT) [194] ( Scheme 35 ). Scheme 35 5-Heterylidene-4-thiazolidinedione with antimicrobial activity. Scheme 35 One of the pathogenetic processes of pro- and eukaryotes is the mechanism associated with galactofuranose-based conjugation which involves UDP-galactopyranose mutase (UGM). The inhibition of such process leads to inhibition of microbial growth and reduction of virulence (especially for Mycobacterium tuberculosis ) and improve the antimicrobial/antimycobacterial activity of UGM inhibitors. High inhibition activity of this enzyme was also discovered for 4-thiazolidinone-3-alkanecarboxylic acids 68 , 69 [180] and 2-imino derivatives 70 [145] , [195] ( Scheme 35 ). 2-Amino(imino)-4-thiazolidinones have been widely investigated [141] , [177] , [196] , [197] as antimicrobial agents, e.g. compound 71 was identified as an inhibitor of the type III secretion system of Gram-negative bacteria [196] , [198] . Pseudomonas aeruginosa had shown reduction of T2S-dependent elastase secretion in the presence of 71 (Type II secretion (T2S) systems is well conserved among Gram-negative bacteria and a key virulence factor of P. aeruginos a) [199] . The structure optimization aiming the increasing of solubility led to compound 74 which possessed superior activity (10 μM) in the S. typhimurium T3S secretion assay [196] . The N-3 position was established to be the most permissive optimization direction ( Scheme 36 ). Scheme 36 5-Ene-4-thiazolidinediones as inhibitors of the type II & III secretion systems. Scheme 36 The same findings are suitable for 2,4-thiazolidinedione derivatives – N3 modification provides to 5–40 folds increasing of the antimicrobial activity ( 72 ) [200] , [201] . Related 2-amino-4-thiazolidinone ( 73 ) was detected as high active CysK1 inhibitor . CysK1-pyridoxal phosphate-dependent O-acetylsulfhydrylase which catalyzes the formation of l -cysteine from O-acetylserine and hydrogensulfide. This cysteine biosynthetic pathway is one of the essential pathways in microbial pathogens, providing potential targets for the development of novel antibacterial compounds [202] ( Scheme 36 ). The row of 2-thiazolylimino-5-arylidene-4-thiazolidinone bearing the nitro group and small heterocyclic fragments (thiazole, benzthiazole, isatine, pyrazole, furane, thiophene) ( 75 – 77 ) ( Scheme 37 ) in positions 2 and 5 showed sufficient antimicrobial activity levels too [203] ; the nitro derivatives were characterized by high renal excretion that can be used in the design of potential diuretics [204] . The results of QSAR study revealed that the compounds with less number of atoms or less number of substituents are more likely to be active than their counterparts with higher molecular weight [205] . Scheme 37 2,5-Disubstituted-4-thiazolidinone hybrids with antimicrobial activity. Scheme 37 5-Heterylidenerhodanines and simple arylidene analogs are proposed as novel class of b-lactamase inhibitors that possesed selectivity for class C b-lactamases ( 78 is a non-b-lactam with the IC 50 ∼ 2.6 mM against the class C enzyme P99) [206] . Simple 5-alkylidene-2,4-thiazolidinediones 79 were discovered as effective inhibitors of autoinducer-2 quorum sensing (mechanism through which bacteria regulate gene expression in response to population density including regulation of the production of virulence factors etc.) Virulence targeting is representing an emerging concept in antibacterial therapy for which there are examples of compounds that inhibit virulence functions [207] . N-3-Dipeptide-thiazolidinone hybrids 80 may provide a critical step toward the validation of this strategy and the development of novel therapeutics [196] ( Scheme 38 ). Compounds having bulky aromatic substituents at position 5 and a tryptophan residue at position N-3 of the rhodanine ring were the most active against InhA ( trans -2-enoyl-acyl carrier protein reductase) with micromolar IC 50 [208] . Scheme 38 Structure of antimicrobial 5-ene-4-thiazolidinones. Scheme 38 3.2 Antiparasitic agents The in vitro antiamoebic activity of 3-substituted 2-amino-4-thiazolidinones 81 was evaluated against HM1:IMSS strain of Entamoeba histolytica and it exhibited promising activity (IC 50 - 0.11–0.172 mM being lower than that of metronidazole IC 50 - 1.64 mM) and low toxicity level [209] ( Scheme 39 ). Scheme 39 3-Substituted 2-amino-4-thiazolidinones with antiamoebic activity. Scheme 39 Several aza-fused rhodacyanines [210] were found as perspective agents when testing various heterocyclic rings on the antimalarial activity. Among tested compounds 82 showed excellent inhibitory activity with an IC 50 of 4.4 nM (K1strain). Another compound 83 with quinoline ring and two rhodanine moieties showed 78% suppression of parasitemia (25 mg/kg/day) [211] . 4-Thiazolidinone-3-carboxylic acids belonging to rhodacyanine dyes 84 are characterized by distinct antimalarial activity [87] , [212] . Significant anticancer activity of the mentioned class of compounds should be emphasized in this context (see below) ( Scheme 40 ). Scheme 40 Rhodanine-based dyes with antimalarial activity. Scheme 40 An in vitro structure–activity relationship investigation showed that the rhodacyanine MKH-57 possesses high antimalarial activity (EC 50 = 12 nM) and significant selective toxicity [87] as well as a series of its analogs [212] . Screening of more than 13000 compounds for antimalarial activity using the agglomerative structural clustering technique allowed identifying 47 starting points for lead optimization including 4-amino(imino)thiazolidinone derivatives 85 [213] . The positional isomer of mentioned 4-amino derivatives ( 85 ) – compound 86 was also described in the compound library of potential antimalarial leads [213] ( Scheme 41 ). Scheme 41 5-Ene-4(2)-thiazolidinones as antiparasitic agents. Scheme 41 In the area of antiprotozoal agents search the design of antitrypanosonal agents based on thiazolidinone scaffold is of special interest [214] . Compounds 87 showed the highest antiproliferative activity in comparison with 5-saturated analog 88 when screened on Trypanosoma cruzi epimastigotes but were inactive towards cruzipain [118] ( Scheme 41 ). While 88, obtained from a virtual screening of 500000 chemical structures (ZINC5 database) against cruzipain [215] inhibited this enzyme at micromolar concentration. 5-Benzylidenerhodanine-3-acetic acid derivatives 89 were reported to possess inhibitory activity against Trypanosoma brucei dolicholphosphate mannose synthase and glycosylphosphatidylinositol anchor synthesis as well as in vitro trypanocidal activity against the blood stream form and were non-cytotoxic against HeLa cells. Dolicholphosphatemannose synthase is a mannosyl transferase critically involved in glycoconjugate biosynthesis in T. brucei . VSG dimmers covering the parasite cell-surface are linked to the trypanosomal plasma membrane via glycosyl phosphatidylinositol anchors, which biosynthesis is essential for viability of the blood stream form of T. brucei . The 3-benzyloxy-substituted analog and the 2-hydroxyderivative 89 showed the best trypanocidal activity (ED 50 ∼ 100 μM) [215] ( Scheme 42 ). Scheme 42 5-Benzylidenerhodanine-3-acetic acid derivatives as antitrypanosomal agents. Scheme 42 Simple 5-(hydroxyphenylmethylidene)thiazolidine-2,4-diones were described as novel inhibitors of Leishmania pteridine reductase 1 where the thiazolidinone ring was treated as a bioisosteric replacement for pteridine/purine ring [216] . One of the discussed modes of action of thiazolidinone based compound is the DNA-binding process. In this study 2-imino derivatives 90 were designed and the most potent molecule bound at the DNA minor groove involving Van der Waals, H-bonding and hydrophobic interactions [217] ( Scheme 43 ). Scheme 43 2-(3-Imidazol-1-yl-propylimino)-5-(3-nitro-benzylidene)-thiazolidin-4-one as potential DNA-binding compound. Scheme 43 3.3 Antidiabetic agents Search for new antidiabetic agents among 4-thiazolidinones is one of the most studied fields and had resulted in the introduction of new class of the antidiabetic drugs – glitazones. Their mechanism of action is associated with activation of PPARγ (peroxisome proliferation activated receptors) [19] , [218] , [219] , [220] ( Scheme 44 ). Scheme 44 Structures of known glitazones. Scheme 44 However, 5-ene derivatives including C5-unsaturated analogs of glitazones are to lesser extent activators of PPARs [221] , [222] . The introduction of carboxylic acid residue in the N3 position leads to decrease of hepatotoxicity of glitazones as well as unsaturated structural analogs (5-arylidene-thiazolidine-2,4-dione-3-acetic acids 91 ). Compound 91 and related derivatives are reported to have low or no activity on PPAR, but have high antidiabetic activity in vivo (confirmed in sucrose-loaded model) [223] . This indicates other mode of antihyperglycemic action of 5-ylidene-4-thiazolidinones which can be treated as glitasone bioisosters. 5-Arylidene-3-benzyl-2,4-thiazolidinedione derivatives ( 92, 93) possess antihyperglycemic activity opposed to troglitazone under the experimental conditions (evaluation of in vitro insulin releasing activity, INS-1 cells); the latter has not decreased the level of hyperglycemia [224] ( Scheme 45 ). Structurally related 5-ene-2,4-thiazolidinediones 94 with bulky C5 fragments are treated as potent euglycemic and hypolipidemic agents too [225] as well as their simple analog 95 (the alloxan-induced hyperglycemia, in vivo mice model) [226] ( Scheme 46 ). Scheme 45 5-Ene-2,4-thiazolidinediones with antihyperglycemic activity. Scheme 45 Scheme 46 5-Ene-2,4-thiazolidinones based hypolipidemic agents. Scheme 46 However, studies of PPAR-related actions of 5-ylidene-4-thiazolidinones are continuing. Virtual screening tools including SQUIRREL (Sophisticated QUantification of InteRaction RELationships) procedure allowed detecting a set of potential PPARs ligands [227] . 14 Compounds including 4-thiazolidinones 96 – 99 showed PPARα and PPARγ agonistic activity (lead-compound 99 possessed the nano-level of activity (EC 50 PPARα = 0.044 μM)) ( Scheme 47 ). Scheme 47 5-Ene-4-thiazolidinones with PPARα/γ agonistic activity. Scheme 47 A new series of PPARγ agonists known as (β-carboxyethyl)-rhodanine derivatives were identified based on ligand-centric and receptor-centric approaches. An in vitro assay had confirmed the nanomolar binding affinity of 100 . In a cell-based transactivation assay similar PPARγ agonistic activity to that of the known PPARγ drug, pioglitazone, was shown. Based on CoMFA model and docking data it was discovered that the electrostatic interaction of the carboxyethyl group with the rhodanine core is important for its binding in the pocket of PPARγ and the rhodanine heterocycle played a different role than the thiazolidine group of rosiglitazon [228] . Currently the investigation of PPARγ antagonists is also of the great interest in the treatment of diabetes and obesity [229] , [230] including study of the 5-ene-4-thiazolidinones. The most investigated mechanism of antidiabetic mode of action of 5-ene-4-thiazolidinones is the inhibition of aldose reductase ( AR ). Aldose reductase is the limiting enzyme of polyol/sorbitol pathway of glucose oxidation, excessive activation of which leads to the accumulation of glucitol and the development of diabetic complications [231] . 5-Ene-rhodanine-3-alkanecarboxylic acids are high affinity inhibitors of aldose reductase [232] . The illustrative example of the mentioned compounds' row is epalrestat – ( Z,E )-5-(2-methyl-3-phenyl-2-propenylidene)-2-thioxo-4-thiazolidinone-3-acetic acid. Optimization of these compounds structures is mainly associated with C5 fragment modification [53] , [233] , [234] , [235] . Study of epalrestat analogs revealed that Z -isomers possessed higher activity level. There is a great interest in 2,4-thiazolidinedione derivatives as AR inhibitors since they can be viewed as hydantoin and rhodanine bioisosteres potentially free of the hypersensitive reactions which are linked to the presence of the hydantoin system [55] , [236] , [237] . In fact, to date, several thiazolidine-2,4-diones have been patented with dual activity as anti-hyperglycaemic and AR2 inhibitory agents [238] , [239] . The findings suggest that the activity of these compounds might correlate with their AR2 inhibitory ability by preventing the stimulation of PKC or MAPK and the subsequent activation of NF-kB [240] . AR2 plays a pivotal role in mediating oxidative stress-induced inflammation and is implicated in the development of various inflammatory pathologies [232] (see below). This enzyme catalyzes the reduction of lipid peroxidation-derived aldehydes, thus producing metabolites which transduce inflammatory signaling by means of the activation of protein kinases such as PKC and MAPK. This in turn activates NF-kB that is responsible for the transcription of many proinflammatory genes [241] , [242] . Antidiabetic activity of target compounds is often related to the inhibition of PTP 1B (protein tyrosine phosphatase 1B), which is an intracellular PTP and a key negative regulator of the insulin signaling pathway as well as to LMW-PTP (low molecular weight protein tyrosine phosphatase). This possible mode of action may be regarded as an attractive approach to the design of new therapeutic agents for the treatment of type 2 diabetes mellitus, obesity and, therefore, of the states associated with complex metabolic disorders known as metabolic syndrome. 4-(5-Arylidene-2,4-dioxothiazolidin-3-yl)methylbenzoic acids 101 were indicated as inhibitors of both PTP 1B and LMW-PTP ( Scheme 48 ). One of the directions in this study is the example of utilization of phosphotyrosine-mimetics to identify effective low molecular weight nonphosphorus inhibitors of PTPs ( p -methylbenzoic acid residue at N-3 position of the 5-arylidene-2,4-thiazolidinedione scaffold can act as a monoanionic pTyr-mimetic group replicating the interactions of pTyr with the catalytic site of the enzyme) [243] . In this study the authors also noted the importance of the 5-arylidene fragment. Alkylated/acylated phenolic groups and methoxy groups are desirable for the inhibitory effect. Scheme 48 5-Ene-4-thiazolidinones with antidiabetic activity. Scheme 48 5-Substituted 2-cyanimino-4-thiazolidinone ( FPFS-410 ) – a compound related to pioglitazone (PPAR-agonist) also possesses antidiabetic activity. In vivo data revealed the ability to reduce blood glucose and triglycerides levels and reduce the obesity [244] . Screening study allowed identifying and confirming the activity of 5-arylidene derivatives as inhibitors of glycogen synthase kinase-3 ( GSK-3 ) which, in turn, has been emerging as a key therapeutic target not only for type 2 diabetes, but also for Alzheimer'sdisease, cancer and chronic inflammation [245] . Simple 5-benzylidene-1,3-thiazolidine-2,4-dione derivatives 102 were presented as a new class of α-glucosidase inhibitors [246] ( Scheme 49 ). α-Glucosidases (α-D-glucosideglucohydrolases) are membrane bound exo-acting enzymes responsible for catalyzing the final step in the digestive process of carbohydrate metabolism. α-Glucosidases are the enzymes that hydrolyze O-and S-glycosyl residues and are involved in the biosynthesis and processing of oligosaccharide chains of N-linked glycoproteins in the endoplasmic reticulum [247] . Scheme 49 Structures of α-glucosidase and glycogen phosphorylase inhibitors. Scheme 49 5-Benzylidenethiazolidine-2,4-diones carrying 2,3-dihydrobenzo[1,4]dioxine fragment 103 were proved to be glycogen phosphorylase inhibitors ( Scheme 49 ) ( glycogen phosphorylase is a key enzyme in the regulation of blood sugar level and it catalyzes the formation of glucose-1-phosphate from glycogen) [248] that may be relevant to the control of blood glucose concentrations in type 2 diabetes [249] . 3.4 Anticancer agents Search for new anticancer agents is the most dynamically developing area of medicinal chemistry. A lot of papers and patents are devoted to the search for efficient anticancer agents among 4-thiazolidinones, including 5-ene derivatives (selected row of patents see) [5(ESI),250]. Research in this area could be divided into the next groups: i) in vitro screening of highly active/selective hit-compounds for further optimization, mainly with unknown mode of action (including the international programs, mainly Developmental Therapeutics Program, NCI, NIH - https://dtp.cancer.gov ); ii) design of the high-affinity ligands to the "validated" anticancer biotargets; iii) creation of the hybrid molecules that combine several pharmacologically attractive scaffolds [158] , [159] , [251] , [252] ; iv) search for antitumor agents among the compounds with known biological activity (anti-inflammatory, antidiabetic, anti- or prooxidant etc.). 4-Thiazolidinones are known to possess good activity against different types of cancer including relatively simple 5-ene derivatives as well as complex or hybrids/conjugates bearing non-fused 5-ene-4-thiazolidinone fragment. Mentioned compounds don't belong to any "classic" anticancer agent types [8] , [94] , [253] , [26] , [27] , [254] , [255] , [256] . Despite the diversity of 5-ene-4-thiazolidinones, the search for antileukemic agents is one of the most promising directions. It was found the tendency of the maximum sensitivity of leukemia cell lines to various subtypes of thiazolidinones [8] , [25] , [26] , [27] , [34] , [35] , [66] , [94] , [138] , [142] , [143] , [162] , [100] , [257] , [258] , [259] , [260] . The crucial role of the C5-(ylid)ene fragment is also confirmed and complication of ene-fragment is considered as a benefit. But anti-leukemic effects of such compounds have been less documented in comparision with all other cancer types [261] . Among 4-thiazolidinone-3-carboxylic acids derivatives, amides 104 – 107 ( Scheme 50 ) with anticancer activity were identified including samples possessing strong antileukemic activity ( 107 ). The SAR study revealed that anticancer activity has significantly decreased or disappeared after transformation into related isosteric compounds (replacement of C5 arylidene moiety by the C5-carboxymethylidene or 5-carboxymethyl fragments) or in case of compounds based on the related heterocyclic cores (e.g. 2,4-imidazolidinedione) [26] , [94] , [253] Moreover, novel 4-thiazolidinone-3-carboxylic acid amides 107 bearing furan moiety exhibited significant cytotoxicity and induction of apoptosis in human leukemia cells [259] . Scheme 50 Structures of 5-ene-4-thiazolodinones with anticancer activity. Scheme 50 Among 5-arylidene-2-arylamino(imino)-4-thiazolidinones 108 a set of active compounds with micromolar IC 50 levels has been detected (log GI 50 ∼ −5.77) [143] . Based on the obtained results the compounds with bulky C5-ene fragment and benzothiazole core were designed: compound 109 (pGI 50 = 4.97) selectively inhibited growth of the HOP-92 cell lines (CNS cancer, pGI 50 = 6.34) [138] . Study of the anticancer activity of 4-thiazolidinones with pyrazoline moiety in the C2 position of the main core has revealed high anticancer potential of the mentioned compounds 110 ( Scheme 50 ). SAR study confirmed the dependence of anticancer activity from the structure of C5 fragment [142] . The isomeric 4-amino thiazolidinones possessed much less activity levels [25] . Thiazolidinone-isatin conjugates are good examples of polycyclic compounds with high anticancer potential and their design illustrates the molecular hybridization approach when two "pharmacophores" are combined into a single molecule [262] . For instance, 111 possesed micromolar activity level and also selectively inhibited the leukemia cell lines [27] . Further structure optimization led to increasing of the anticancer activity which was reflected in the design of thiazolidinone-isatin-pyrazole hybrids 112 , 113 [34] , [35] ( Scheme 51 ). Similar isatin-pyrazoline conjugates (without thiazolidinone core) didn't show the anticancer activity. Scheme 51 Thiazolidinone-isatin conjugates with anticancer activity. Scheme 51 Another example of anticancer isatin-thiazolidinone hybrides are compounds 114 [263] and 115 [264] ( Scheme 51 , Scheme 52 ). The latter are the result of the structure optimization of sunitinib (SU11248, Sutent TM; Pfizer Inc) which is currently used in the clinics as a multi-targeting tyrosine kinase inhibitor with antiangiogenic activity [265] , [266] . The same approach of molecular hybridization was used in the design of thiazolidine-2,4-diones 116 bearing α-bromoacryloylamido moiety [267] , [268] . Scheme 52 4-Thiazolidinones hybrids as anticancer agents. Scheme 52 Compound MKT077 is known anticancer agent [269] ( Scheme 53 ) with apoptosis related mechanism of action and is an illustrative example of above mentioned compounds, though, its study has been proceeding till now. Thus, it is shown that it acts through differential interaction with the Hsp70 ( Heat shock protein 70 kDa ) allosteric states and reactivation of p53 function. MKT-077 therefore can be treated as an "allosteric drug" [270] , [271] , [272] , [273] . Scheme 53 Structure of MKT 077. Scheme 53 One more example of the hybrid approach in anticancer agents design is the combination of the natural compound, e.g. oleanane scaffold, with thiazolidinone core ( Scheme 54 ). Among oleanane-thiazolidinone hybrids 3-(2,4-thiazolidinedione-5-ylidene)-carboxyimino]olean-12-en-28-oic acid methyl ester ( 117 ) was identified as the most active substance (pGI 50 = 5.57, pTGI = 5.13 and pLC 50 = 4.64 NCI NIH protocol) with low toxicity and moderate activity level in in vivo Hollow Fiber Assay [162] . Another example of this approach is 5-arylidene-2,4-thiazolidinedione bearing cholesterol fragment 118 . Moreover, the level of activity and selectivity of antimitotic effects depends on the substituents in the C5 position of thiazolidinone core [161] . The maximum effect such compounds exhibited towards HeLa cancer cells line. Scheme 54 Combination of natural compounds and 5-ene-4-thiazolidinones fragments. Scheme 54 Promising active molecules based on the combination of thiazolidinone and chromene cores 119 were also found [160] ( Scheme 55 ). Scheme 55 Combination of thiazolidinone and chromene cores. Scheme 55 Once again it should be emphasized that for majority of mentioned compounds the tendency towards maximum sensitivity of leukemia and lung cancer cell lines is shown. This, most probably, may be a platform for the effective anti-leukemic agents design. For example, 5-isopropylidene-3-ethylrhodanine and 5-benzylidene-3-ethylrhodanine demonstrated cytotoxicity towards leukemic cell line CEM by inducing apoptosis [38] , [274] . Besides this, the modification of known drugs, for example assorafenib, with 5-ene-4-thiazolidinone fragments 120 is reported [275] , [276] ( Scheme 56 ). Scheme 56 Modification of known drug with 5-ene-4-thiazolidinone fragment. Scheme 56 Whereas, a lot of potential biotargets for new anticancer agents design are known at present, there have been almost no attempts to systematize the 4-thiazolidinones as anticancer agents in the review papers. For example, three aspects of the PPARs -independent antitumor activities of thiazolidinones had been outlined [10] : i) inhibition of Bcl-2/Bcl-x function, ii) proteasomal degradation of target proteins, iii) transcriptional repression of AR through Sp1 degradation [277] . Though, the latter definitely does not represent the whole spectrum of experimental data and is not enough for the design of novel hit- and lead-compounds. Taking into account the great number of papers, we tried to outline the most referred biotargets and their ligands among 5-ene-4-thiazolidinones. It should be noted that the majority of high-affinity ligands belong to 4-thiazolidinone-3-carboxylic acids, 5-benzylidenerhodanines, 2-amino(imino)derivatives and 2,3-disubstituted-4-thiazolidinones. Based on the established role of PPARs in the cancer and inflammation progress [219] , [278] , [279] , [280] considerable part of investigations is dedicated to the study of glitazones and related derivatives as possible anticancer agents. However, 5-ene analogs as synthetic precursors of glitazones that do not contain exocyclic double bond in the C5 position possess less expressed affinity to PPARs as their agonists. PPARγ antagonists in turn represent a new drug class that holds promise as a broadly applicable therapeutic approach for cancer treatment [229] , [281] . Among them 5-ene derivatives occupy their deserved place. For instance, the novel 3-thiazolidine-modified benzoic acid derivative HL005 ( Scheme 57 ) being a potent PPARγ-specific antagonist inhibits the proliferation of the MCF-7 cell line at the concentration-dependent manner, induces cell cycle arrest at the G2/M phase and interferes with cell adhesion [282] . Scheme 57 Structure of HL005 – PPARγ-specific antagonist. Scheme 57 PPARs indepented anticancer effects of troglitazone are mediated mainly via the repression of cyclin D1 (MCF-7 breast cancer cells) by facilitating proteasome-facilitated proteolysis [283] and partial depletion of intracellular Ca 2+ stores that leads to inhibition of translation initiation [284] . The troglitazone structure modification led to the STG28 identification – the first small-molecule agent mediating the proteasomal degradation of cyclin D1 with high specificity (exposure to STG28 did not cause any appreciable change in the expression levels of a series of other cyclins and CDK-dependent kinases ) [283] ( Scheme 58 ). Scheme 58 Structure of 5-ene-thiazolidinones with anticancer activity derived from troglitazone structure modification. Scheme 58 Following the structure optimization and screening data it was shown that the 5-arylidenerhodanines 121 are equipotent to troglitazone in Ca 2+ releasing activity, induction of eIF2a phosphorylation and more potent in inhibiting cancer cells proliferation [285] . Further modification allowed to state that a series of 5-benzylidene-2,4-thiazolidinediones and -thiones inhibited cell growth at low micromolar concentrations via the inhibition of translation initiation which involves partial depletion of intracellular Ca 2+ stores and strong phosphorylation of eIF2a . Structurally similar benzylidenerhodanines 122 showed high inhibition of protein tyrosine phosphatase PRL-3 (IC 50 = 0.9 μM), which is one of the probable prognostic markers of metastatic cells [286] . Among 5-ene-rhodanines 123 was identified as a promising and selective inhibitor of enzymes of dual-specificity phosphatases group – phosphatases JSP-1 ( JNK-stimulating phosphatase-1 ). Mentioned compounds are as well perspective agents that can be explored in the treatment of the inflammatory disorders [140] ( Scheme 59 ). Scheme 59 3-[5-(4-Fluorobenzylidene)-rhodanin-3-yl]-benzoic acid - JSP-1 inhibitor. Scheme 59 Necrostatin-7 ( Nec-7 ) ( Scheme 60 ) and related heterocycles belong to new class of "small molecules" – inhibitors of necroptosis that is regulated caspas-independent pathway of the cell death, morphological features of which are close to necrosis. This may be used in FADD-changed variant of the treatment of JurkatT cancer cells under the use of TNFα [287] , [288] . Scheme 60 5-[3-(4-Fluorophenyl)-1 H -pyrazol-4-ylmethylene]-2-imino-3-(thiazolyl-2)-4-thiazolidinone – necroptosis inhibitor. Scheme 60 On the other hand, derivatives of hydantoin-5-acetic acid are non-hydroxamate inhibitors of TNFα converting enzyme [289] that allows to treat them as promising anti-inflammatory agents and gives the prospects to establish anticancer potential of the given class of compounds. Moreover, the ability of hydantoincarboxylic acids, especially with the thiazole fragments, to inhibit activity of Ras farnesyl transferase (Ftase) was proved [290] . Ras proteins play an essential role in the processes of cell growth and differentiation and need post translational modification including farnesylation catalyzed by Ras farnesyl transferase . That is the reason why Ras-Ftase inhibitors are considered as potential anticancer agents [291] , [292] , [293] . Structural analogs of the given substances are the ligands for neuro-immunophilin FK506-binding protein (FKBP) [294] . 5-Substituted rhodanine-3-carboxylic acids 124 ( Scheme 61 ) are the inhibitors of protein-protein interaction of antiapoptotic proteins of the Bcl-2 and Bax family and their binding to the appropriate receptor domains [257] , [295] , [296] , [297] , [298] , [299] . Scheme 61 5-Arylidenerhodanine-3-carboxylic acids - Bcl-2 inhibitor. Scheme 61 Apart from the directions shown above, 5-aryl(hetyryl)idenethiazolidine-2,4-diones 125, 126 ( Scheme 62 ) were identified as potent and selective insulin-like growth factor-1 receptor (IGF-1R) inhibitors [230] . Scheme 62 Structure of insulin-like growth factor-1 receptor inhibitors. Scheme 62 IGF-1R is a growth factor receptor of tyrosine kinase family, acting as a critical mediator of cell proliferation and survival. Although, being highly related to insulin receptor, it plays a different role in organism development, being responsible for normal growth and development as opposed to glucose homeostasis. Epidemiological studies indicate that the IGF-1R is overexpressed in human cancer and is primarily responsible for tumor genesis. Signaling through IGF-1R includes the activation of PI3K and Raf pathways [300] . Inhibition of both these pathways makes IGF-1R kinase a promising target for cancer therapy. Moreover, series of 3,5-disubstituted thiazolidine-2,4-dione analogs [301] and 3-(2-aminoethyl)-5-(3-phenylpropylidene)-thiazolidine-2,4-dione [302] were shown to be potential anticancer agents via the inhibition of the Raf/MEK/ERK and PI3K/Akt signaling cascades. The inhibitors of extracellular signal-regulated kinases-1 and 2 (ERK1/2) – (e.g. ( Z )-3-(2-aminoethyl)-5-(4-ethoxybenzylidene)thiazolidine-2,4-dione 127 ) are the examples of compounds with greater selectivity for inhibiting the proliferation of melanoma cells [303] ( Scheme 63 ). Scheme 63 Structure of signal-regulated kinases inhibitor. Scheme 63 Polo-like kinase 1 (Plk1) is a key regulator of mitotic progression and cell division. In eukaryotes it acts in concert with cyclin-dependent kinase 1 – cyclin B1 and Aurora kinases to conduct a wide range of critical cell cycle events. Because Plk1 has been preclinically validated as a cancer target, small-molecule inhibitors of Plk1 have become attractive candidates for anticancer drugs development [304] . Thiazolidinone 128 ( Scheme 64 ) selectively inhibits human Plk1 (IC 50 19 nM) and various human and mouse tumor cell lines (IC 50 0.2–1.3 μM) and cause a prometaphase-like mitotic arrest [305] . Scheme 64 Structure of Polo-like kinase 1 inhibitor. Scheme 64 Simple 5-ene-rhodanines ( 129 , 130 ) ( Scheme 65 ) may also be considered as "Myc-Max compounds" that inhibit or reverse the association between c-Myc and its obligated HLH-LZ heterodimerization partner Max [306] , [307] , [308] . Scheme 65 Simple 5-ene-rhodanines as "Myc-Max compounds". Scheme 65 Among 5-ene derivatives of 2-iminothiazolidine, 2,4-thiazolidinedione and rhodanine the estrogen-related receptor-α (ERR-α) modulators were identified. They can be useful for the prophylaxis or treatment of ERR-α associated diseases such as breast cancer [309] , [310] . Following the modification of 5-(3-trifluoromethylbenzylidene)thiazolidine-2,4-dione (high selectivity Pim-1 inhibitor) a series of substituted thiazolidine-2,4-dione derivatives were identified as highly active and selective Pim-1 and Pim-2 inhibitors (nanomolar values of IC 50 for Pim-1 and ∼2.0 μ M for Pim-2) [54] . Pim-1 and Pim-2 are serine/threonine protein kinases frequently overexpressed in prostate cancer and certain forms of leukemia and lymphoma [311] ; Pim can phosphorylate the proapoptotic protein BAD (the Bcl-2-associated death promoter) leading to sequestration of 14-3-3 proteins and inhibits the apoptosis [312] , [313] . As it was already mentioned, 5-ylidene-2-amino(imino)-4-thiazolidinones 131 ( Scheme 66 ) comprise one of the most investigated thiazolidinone subtype as compounds with anticancer activity [138] , [142] , [143] , [314] , [315] . Among them the derivative MMPT was identified as a hit-compound being able to effectively inhibit the growth of some lung cancer cell lines (H460 and H460/TaxR) and at the same time do not influence normal fibroblasts in a dose response manner [151] , [316] , [317] . Study of the structure-activity relationship in a group of 372 compounds revealed some structural peculiarities: i) the nitrogen atom of thiazolidine core has to be unsubstituted; ii) dimethylaminobenzylidene fragment is the best in the C5 position; iii) 2-phenylamino moiety may contain substituents in different positions of aromatic ring [151] . Scheme 66 Structure of simple 5-ylidene-2-amino(imino)-4-thiazolidinones with significant anticancer activity. Scheme 66 5-[(4-Methylphenyl)methylene]-2-(phenylamino)-4(5 H )-thiazolone ( MMPT ) and 5-(2,4-dihydroxybenzylidene)-2-(phenylimino)-1,3-thiazolidine ( DBPT ) are effective in the treatment of multidrug-resistant (MDR) cancer. Multidrug-resistance is a phenotype of cross-resistance to multiple drugs with diverse chemical structures. One of the well-documented MDR mechanisms is the overexpression of the MDR-1gene that encodes the transmembrane, ATP-dependent drug efflux transporter P-glycoprotein (P-gp) in response to chemotherapy [318] , [319] , [320] . P-gp prevents the intracellular accumulation of many cancer drugs by increasing thei refflux out of cancer cells as well as through hepatic, renal, or intestinal clearance pathways [319] . Attempts to coadminister P-gp modulators or inhibitors to increase cellular availability by blocking the actions of P-gp have met with limited success [321] , [322] , [323] . Therefore, a more promising approach lies in the design and discovery of novel compounds that are not substrates of P-gp and are effective against drug-resistant cancer while at the same time exhibit minimal toxicity to normal cellular functions. MMPT inhibited the growth of human non-small-cell lung cancer and colon cancer cells (drug-sensitive (H460) and drug-resistant (H460/TaxR cell lines) independently of P-glycoprotein and p53 status [151] , [316] . Besides this, it selectively killed drug resistant cancer cells and induced apoptosis [317] . Among 5-ylidene-2-arylamino-4-thiazolidinone derivatives effective growth inhibitors of HT29 cell line (with a high COX-2 expression) were identified [324] . Among mentioned 4-thiazolidinone derivatives efficient CDK1 132 [325] , CDK1/cyclinB inhibitors [326] , [327] , [328] and SHP-2 inhibitors 133 and 134 ( Scheme 67 ) ( SHP-2 is a non receptor protein tyrosine phosphatase that mediates cell signaling by growth factors and cytokines acting via the RAS/MAP kinase pathway) [329] , [330] were discovered. Benzo[ d ]thiazole derivatives turned out to be more potent than their thiazole analogs. Scheme 67 Structures of CDK1 and SHP-2 inhibitors. Scheme 67 2-Substituted thiazolidinone and oxazolidinone derivatives were established as the inhibitors of phosphatases and anticancer agents [331] . Modifying the imidazolidine derivatives the new cell division cycle 7 kinase inhibitors were designed 135 and ( Z )-2-(benzylamino)-5-(1 H -pyrolo2,3- b ]pyridin-3-ylmethylene)-1,3-thiazol-4(5 H )-one] was selected as a lead compound [332] ( Scheme 68 ). Scheme 68 Structure of cycle 7 kinase inhibitors. Scheme 68 2-N,N′-Disubstituted diamines bearing 5-arylidene-4-thiazolidinone moiety 136 ( Scheme 69 ) had shown nanomolar inhibition potency (IC 50 40 nM) towards tyrosine phosphorylation-regulated kinases 1A [333] . This result prompted to explore the symmetric 1,2-diamino-linker grafted on N-3 position of two different 5-arylidenerhodanine platforms in order to modulate potential biological activity and led to the synthesis of unsymmetrical linked bis-5-arylidenerhodanine derivatives with anticancer effects [334] . Scheme 69 Structure of tyrosine phosphorylation-regulated kinases 1A inhibitors. Scheme 69 One of the established modes of action of 4-thiazolidinones anticancer activity is their antagonistic activity towards α v β 3 receptors . The value of inhibition of the α v β 3 activity is proved by the establishing of the correlation between progression of factor's expression and cancer growth. Search for effective inhibitors was held based on the pharmacophore model and virtual screening of 600000 compounds following by further confirming of the activity in vitro . This allowed to select promising compounds including 4-azolidinones. 2-Substituted 4-thiazolidinone derivatives among selected ones appeared to be efficient integrin α v β 3 antagonists 137 [335] ( Scheme 70 ). Scheme 70 General structure of integrin α v β 3 antagonists. Scheme 70 One of the aspects of anticancer effects realization of 4-azolidinone-3-alkanecarboxylic acids is the ability to influence the oxidative homeostasis of the cancer cells and the impact on the reactive oxygen species production. The modulating effect of the 5-ene-rhodanines on the proliferation and apoptosis of the cancer cells was identified as well as the influence on metabolism of NO alone with antioxidant and immunomodulating actions [336] . In this context interesting is the combining anticancer activity with anti-inflammatory effects; with antioxidant and/or other types of activity in some of the compounds [337] that is important in the development of "classic triad": "oxidative stress – inflammation – cancer". Additionally, the pharmacological profile of 5-arylidene-4-thiazolidinones derivatives is related to free radicals scavenging activity [106] , [338] , [339] . Besides this, notes about prooxidant effect of thiazolidinones with anticancer activity are frequently found in the papers. It has been shown that the derivatives of 5-ylidene-2-imino-4-thiazolidinone increased levels of reactive oxygen species (ROS) (HT29 cells), moreover, the ROS increasing was blocked by the ascorbic acid addition [340] . 5-Benzilidene-3-ethylrhodanine treatment led to increased level of ROS production and DNA strand breaks suggesting activation of apoptosis for induction of the cell death (leukemic cell line, CEM) plus by inducing a block at S phase [274] . Cell exposure to the selected compound was associated with the production of ROS, and the induction of autophagy which could lead to cell death [341] . Recent observation has been made in favour of the alternative chemotherapeutic strategy against cancer that consists in increasing the production of reactive oxygen species [340] , [342] . Moreover, generation of ROS could activate the NF-2E related factor 2/Kelch like ECH-associated protein 1 ( Nrf2/Keap1 ) pathway involved in protection of the cells against oxidative stress [343] . Activation of the Nrf2/Keap1 pathway leads to the stabilization of Nrf2 which translocates in nucleic db in ds to Maf protein. The complex, after binding to specific DNA sequence, defined as antioxidant response element localized in the promoter of Nfr2/Maf target genes, increases the transcription of genes involved in reactive species-mediated response such as HemeOxygen-1 (HO-1) , NADH quinoneoxidase-1 (NQO1) or glutamylcysteyl ligase. It could also indicate the indirect impact of the given class of compounds on the pro/antioxidant balance [156] . ROS play a key role in mitochondria mediated apoptosis. Mitochondria are the prime source of ROS which are byproducts of aerobic respiration [344] , [345] , [346] . High levels of ROS in mitochondria can result in free radical attack of membrane phospholipids and cause mitochondrial membrane depolarization. This is an irreversible step associated with the release of mitochondrial factors triggering caspase cascades [347] , [348] . On the other hand, immunoblotting analysis showed that 5-(4-hydroxy-3-dimethoxybenzylidene)-2-thioxo-4-thiazolidinone (RD-1) significantly elevated the Parkin and Miro 2 expression levels in acute MPTP treated mice and improved mitochondrial membrane potential and ATP synthesis in MPP + -treated Neuro-2a cells. Moreover, RD-1 attenuated impaired mitochondrial transport and vesicle release dysfunction has been evoked by MPP + cytotoxicity in cultured primary mesencephalic neurons. Taken together, the results indicate that improving the mitochondrial dysfunction may be a good choice to delay the neurodegenerative progression commonly associated with pakinson diseases as well as cancer [349] . Among the proposed mechanisms for the antitumor effect of target compounds apoptosis induction, cell cycle arrest and differentiation have been extensively reported [38] , [66] , [252] , [274] , [277] . Anticancer effect of such compounds can be also achieved by blocking the cell cycle progression at the G2/M phases border (not affect the G1 to S phase transition) in reversible manner and by the induction of apoptosis [252] , [325] . FACS (fluorescence-activated cell sorting) analysis of 5-ylidene-rhodanine-3-carboxylic acids, which possessed strong antiproliferative activity against human leukemia cells, showed a remarkable accumulation of subploid cells, the sub G1 phase (G0/G1) followed by the decline of both G1 and G2/M phases. Morerover, the presented findings suggest that the observed growth inhibition could be due to apoptosis [259] . The related 5-ylidene thiazolidinones with heterocyclic fragments induced cell cycle arrest of leukemic cells at G2/M phase and induced cell death that resulted in increased level of SubG1 phase population, thereby affecting its proliferation and causing depolarization of mitochondrial membrane potential [260] . The similar results were found for 2-heterylamino 4-thiazolidinones that induced cell accumulation at G2/M and in sub G0/G1 phases of the cell cycle. Furthermore, dissipation of mitochondria membrane potential was observed as well as red/ox changes in the treated cells (HT29 cell lines) [340] . Treatment of cancer cells with mentioned MMPT and DBPT led to a time-dependent accumulation of cells arrested in G2/M phase [316] . 5-Ene-4-thiazolidinones caused a prometaphase-like mitotic (G 2 –M) arrest (as the ATP-competitive inhibitors of Plk1) [305] , [350] . Despite the diversity of presented biotargets, the notes about thiazolidine derivatives being inductors of apoptosis in various cancer cells have become more frequent [351] , [352] , [353] , [354] , [355] , [356] , [357] . Above all, apoptosis-related mechanism is associated with the inhibition of Bcl-2/Bcl-x function. However, there has been described another feasibility of target compounds to affect diverse apoptotic signaling pathways, namely, induction of peroxisome proliferator-activated receptor-associated caspase-dependent [250] , [317] , [358] mitochondria-mediated apoptosis. 4-Thiazolidinones can decrease mitochondrial membrane potential which is one of the important mechanisms for the mitochondrial mediated apoptotic cell death [359] (an example of such compound is MKT 077 ) [269] , [271] . 3.5 Anti-inflammatory agents Progress in the anti-inflammatory agents search is reached via both high affinity ligand design and pharmacological screening of the compounds with unknown molecular mode of action. Anti-inflammatory activity of 4-thiazolidinones is associated, primarily, with their ability to inhibit isoforms of cyclooxygenase ( COX ) and lipoxygenase ( LOX ). The most successful representative of anti-inflammatory agents on the base of 5-ene-4-thiazolidinones is darbufelone – (5-(3,5-ditertbutyl-4-hydroxybenzylidene)-thiazol-4-one [33] , [360] , [361] ) ( Scheme 71 ) that has attracted broad attention as dual inhibitor of cellular prostaglandin and leukotriene production via cyclooxygenase activity inhibition including inhibition of prostaglandin endoperoxide synthase-1 (PGHS-1) and PGHS-2 [362] . Scheme 71 Darbufelone structure and directions of the structure modification. Scheme 71 Chemical structure of darbufelon has been considered as lead-compound for novel anti-inflammatory agents design till now. Optimization of the latter and application of the bioisostere concept allowed the elaboration of new bioactive compounds 138 (GS26) and 139 (GS28) ( Scheme 71 ) with significant anti-inflammatory activity as well as antiedematogenic and ulcerogenic activity in vivo [363] . Further studies of related compounds allowed identifying selective inhibitors of COX-2 as promising anti-inflammatory agents with fewer side effects. Obtained data and SAR analysis resulted in identifying the most active compound 140 ( Scheme 71 ) from the group of thiazolones and oxazolones with di- tert -butylphenol fragment. Compound 140 inhibited human recombinant COX-2 (IC 50 1.7 μM) and inhibited the COX-2 activity in cell line J774A.1 (IC 50 0.17 μM). This compound turned out to be inactive towards COX-1 at the concentration of 100 μM and did not inhibit the COX-1 in thrombocytes at 20 μM [360] . Among 5-ene-4-thiazolidinones CT-8 ( Scheme 72 ) was found to be potent inhibitor of 15-hydroxyprostaglandin dehydrogenase ( 15-PGDH ) [364] . 15-PGDH type I is NAD + -dependent (while Type II is NADP + preferred) cytosolic enzyme which catalyzes oxidation of prostaglandins to 15-ketoprostaglandins. Inhibition of this enzyme is related to prostaglandin E2 action and can lead not only to anti-inflammatory effects but, for instance, to reduce hair loss [365] . Structure-activity analysis indicated that the N-methylation of CT-8 abolished the inhibitory activity. It was also established that the nature of the moiety linked to benzylidenethiazolidine-2,4-dione through another linkage plays an important role in its inhibitory activity [364] . Further modification led to the new related compound 141 identification (IC 50 = 0.031 μM) [366] ( Scheme 72 ). Scheme 72 Structures of 15-hydroxyprostaglandin dehydrogenase inhibitors. Scheme 72 Another mechanism of anti-inflammatory activity has been the inhibition of phosphodiesterase (PDE4 and PDE2). Phosphodiesterase is responsible for the hydrolysis of secondary messanger cAMP, the level of which increases under inflammation processes. Besides this, phosphodiesterase inhibitors may be effective agents in astma and obstructive lung diseases treatment. It has been established the ability of 5-arylidenerhodanines, including 3-substituted carboxylic acids [367] to inhibit above mentioned enzyme. Moreover, unlike other rhodanine derivatives, carboxylic derivatives selectively inhibit only phosphodiesterase-4 that may be an argument for their study as probable agents for Parkinson's and Alzheimer's diseases. The antigenerative effect of 5-benzylidene-4-thiazolidinones with benzothiazole 142 and isothiazole 143 moieties ( Scheme 73 ) on human chondrocytes culture was estimated when studying potent inhibitors of metalloproteinase (MMP). Given experimental model reproduces the mechanisms involved in osteoarthritis [240] , [368] . The compound 143 (Ar = 4-MeO-C 6 H 4 ) showed MMP-13 inhibition activity at nanomolar concentration (IC 50 0,036 μМ). Scheme 73 Structures of 15-hydroxyprostaglandin dehydrogenase inhibitors. Scheme 73 Moreover, among anti-inflammatory 5-ene-4-thiazolidinones a series of furan-2-ylmethylene-2,4-thiazolidinediones have been investigated as the ATP-competitive PI3Ks inhibitors 144 ( Scheme 74 ) ( IC 50 0.2–0.9 μM). Class I phosphoinositide 3-kinases ( PI3Ks ), in particular PI3Kγ , have become attractive drug targets in inflammatory and autoimmune diseases treatment [369] . An acidic NH group in the thiazolidinedione core and a hydroxyl group in the furan-2-yl-phenyl part of the molecule play crucial roles in binding to PI3K and contribute to class IBPI3K selectivity. Besides the compound 145 inhibited PI3Kγ (IC 50 33 nM) (but not PI3Kα , PI3Kβ ) dependent pathways inside cells and in a murine peritonitis model, it produced a similar decrease in leukocyte infiltration. Scheme 74 Structure of ATP-competitive PI3Ks inhibitors. Scheme 74 The anti-inflammatory properties of 5-arylidene-2-phenylimino-4-thiazolidinones 146 ( Scheme 75 ) are related to their ability to block the production or action of the degenerative factors induced by IL-1b [370] and possessing antidegenerative activity on human chondrocyte cultures. Scheme 75 5-Arylidene-2-amino-4-thiazolidinones with anti-inflammatory activity. Scheme 75 5-Arylidine-2,4-thiazolidinedione derivatives without substituent in the N3 position [371] as well as N3-substituted analogs 147 [372] ( Scheme 76 ) possessed anti-inflammatory activity and related type of activities, such as analgesic activity and in vitro antioxidant activity or even antimicrobial [167] . The structural affinity to PPARγ agonists ( in silico correlation studies and PPAR-competition binding assay) indicates the possibility of PPAR mediated mode of anti-inflanmatory action [372] . Scheme 76 5-Ene-4-thiazolidinones as active anti-inflammatory agents. Scheme 76 Similarly, 5-arylidene-2-imino-4-thiazolidinones 148 ( Scheme 76 ) exhibited anti-inflammatory effect (carrageenan-induced paw edema model) causing the reduction of PGE2 level together with insignificant COX-2 inhibition when compared with celecoxib. It should be noted that 5-unsaturated thiazolidinones possess much less expressed effect or are inactive [373] . Applying procedures of virtual and highthroughput screening for 2-(thiazole-2-ylamino)-thiazol-4-ones study led to identification of compounds 149 ( Scheme 76 ) with anti-inflammatory activity. In the in vivo cyclooxygenase and lipoxygenase inhibition assays as well as anti-inflammatory assay the compounds turned out to be promising agents for further studies in this field [374] . The row of hybrids 138 [363] ( Scheme 71 ), 150 [375] , 151 [376] ( Scheme 77 ) based on indometacine, roziglitazone and darbufelone molecules was designed. The molecules possesed in vivo anti-inflammatory activity in air pouch and peritonitis models or carrageenan induced paw edema model and inhibited the cyclooxygenase-1 and 2 . Scheme 77 5-Ene-4-thiazolidinones designed based on indometacine modification. Scheme 77 A number of related 3-substituted-5-arylidenethiazolidine-2,4-diones ( 152, 153 ) ( Scheme 78 ) was screened for anti-inflammatory effects. Compounds exhibited significant levels of anti-inflammatory activity in the assay of induced edema in mouse paws [377] . Scheme 78 5-Ene-4-thiazolidinones being active in induced edema paws model. Scheme 78 Biphenyl-4-carboxylic acid 5-(arylidene)-2-(aryl)-4-oxothiazolidin-3-yl amides 154 ( Scheme 78 ) had also shown significant activity in carrageenan test [378] . Application of polypharmacological approach and the advances in the pathophysiological role of various biotargets had resulted in the great interest in search for new anti-inflanmatory agents. For example, it was established that aldose reductase (AR) is critically involved in inflammatory processes under both normoglycemic and diabetic conditions. This enzyme, which is overexpressed under various oxidative conditions, intervenes in multiple signaling pathways leading to inflammation and tissue degeneration. Accordingly, it was ascertained that AR inhibition prevents multiple inflammatory pathways [379] , [380] , [381] and, therefore, new more effective and safer AR inhibitors were designed not only as antidiabetic (see above) but also antinflammatory agents [232] . 2-(Benzylamino)-5-((thiophen-2-yl)methylene)-thiazol-4-(5 H )-one showed the highest anti-inflammatory response on PBMCs (peripheral blood mononuclear cells) exerted through the NF-kB inhibition. This and related compounds also had antioxidant activity and xanthine oxidase inhibitory potency [152] . 3.6 Antiviral agents Investigation of antiviral activity of 4-thiazolidinone derivatives has been carried out mainly in two directions: search for anti HIV agents and search for agents used in hepatitis C treatment. Tested compounds are small molecule inhibitors of validated targets such as HCV NS3 and 5 proteases [39] , [382] , HIV RT [383] etc. Recently, rhodanine derivatives, namely 2-aryl-5-(4-oxo-3-phenethyl-2-thioxothiazolidinylidenemethyl)-furans 155 ( Scheme 79 ) were reported to exhibit anti-HIV-1 activity [384] . Scheme 79 5-Ene-4-thiazolidinone derivatives with antiviral activity. Scheme 79 Among 5-arylidenerhodanine derivatives the first small molecule able to inhibit HIV replication by targeting a cellular enzyme – the RNA helicases DDX3 had been identified [385] . The precise combination of functional groups on the rhodanine scaffold was shown to be responsible for the DDX3 inhibitory activity and selectivity of the hit compound 156 (FE15) ( Scheme 79 ). The 2-thioxo-4-thiazolidinones and salicylic acid containing compounds were the most potent HIV-1 integrase inhibitors ( HIV-1 integrase catalyzes the integration of proviral DNA in to the host genome, an essential step for viral replication) among several compounds retrieved from a database of small-molecules. Compounds 157 ( Scheme 79 ) inhibited strand transfer activities of HIV-1 integrase with similar IC 50 values (10–20 μM). The replacement of either the rhodanine or salicylic acid fragments is found to reduce HIV-1 integrase inhibitory potency [386] . Rhodanine-3-alkanecarboxylic acids, their amides and bicyclic analogs 158, 159 ( Scheme 80 ) represent a new class of inhibitors of HIV-1 Integrase as antiviral agents [387] . They were identified in in silico studies when modeling HIV-1 integrase inhibitors as well as in in vitro investigations. Scheme 80 Structures of inhibitors of HIV-1 integrase. Scheme 80 Employing HCV proteins as targets, directly acting antiviral agents were identified and collectively described as specifically targeted antiviral therapy for HCV (STAT-C) [388] , [389] , [390] . Most approaches to small molecule inhibitors search for Hepatitis C virus treatment have mainly focused upon inhibition of essential viral targets, such as the NS3-4A protease (analogous to HIV protease ), NS5B polymerase , NS3 helicase and NS5A [391] , [392] . A literature survey on HCV NS5B polymerase inhibitors clearly indicates that 4-thiazolidinones could inhibit this enzyme and might be promising candidates for the development of novel antiviral agents against HCV. Screening of in-house library allowed identifying derivative 160 ( Scheme 81 ) as hit-compound (IC 50 2.0 μM) acting towards NS5B-polymerase of HCV. Based on the results of rational design and virtual screening (GOLD docking) novel 2-imino-4-thiazolidinone derivatives were synthesized and compound 161 ( Scheme 81 ) with the IC 50 (3.0 μM) level close to that of 160 was identified [393] , [394] , [395] . Scheme 81 Structures of HCV NS5B polymerase inhibitors. Scheme 81 Further investigation led to N-substituted (aryl)alkylidene-rhodanines synthesis which inhibit HCV NS3 protease and also are good inhibitors of other serine proteases ( chymotrypsin and plasmin ). But, the selectivity of some compounds ( 162, 163 ) ( Scheme 82 ) with bulkier C5 fragments bearing hydrophobic functional groups as well as simple analogs ( 164 – 166 ) ( Scheme 82 ) was increased by ten fold towards HCV NS3 protease respectively [39] . Scheme 82 Structures of HCV NS3 protease inhibitors. Scheme 82 The 4-thiazolidinones from the groups of 2,3-diaryl-4-thiazolidinones [382] , [396] ; 2-amino-5-arylidene-4-thiazolinones [397] , [398] ; and 5-arylidene-3-substituted rhodanines [399] , [400] (167 – 171) ( Scheme 83 ) were the initial structures for new inhibitors of HCV NS5B polymerase design, mainly 2-heteroarylimino-5-arylidene-4-thiazolidinones. In all cases the 5-ylidenethiazolidininones inhibited NS5B at lower IC 50 values ranging between 19.8 and 64.9 mM. Moreover, the authors argued about the contribution of these (ar)ylidene groups in stabilizing the binding mode to NS5B active site. Scheme 83 5-Ene-4-thiazolidinones as inhibitors of HCV NS5B. Scheme 83 Further investigation led to new lead-compound identification 172 [401] . Thus the derivatives of 5-ene-rhodanine-3-acetic acids 173, 174 were active towards HCV inhibiting virus protease [402] . Additionally, it was established the ability of the derivatives of this compounds' class to inhibit human serine proteases . SAR analysis showed that substitution of arylidene moiety by the carboxy methylidene fragment led to insignificant loss of the activity ( Scheme 84 ). Scheme 84 Structures of HCV protease inhibitors. Scheme 84 Based on the previous data [403] a series of 2(4)-pyrazolyl-4(2)-thiazolidinones 175, 176 ( Scheme 85 ) were synthesized and their antiviral activity against Influenza viruses (Type A and B), Corona virus SARS, Tacaribe virus, Dengue virus, Rift Valley Fever virus, Respiratory Syncytial virus, Vaccinia virus and Venezuelan Equine Encephalitis virus was tested in vitro . Scheme 85 Structures of hit-compounds with antiviral activity. Scheme 85 Molecular hybridization method allowed obtaining thiazolidinone–peptide hybrids 177 ( Scheme 85 ) that inhibited Dengue virus protease . Moreover, it was determined that thiazolidinone core (2,4-thiazolidinedione and rhodanine) and the peptides should be accomplished by relatively rigid arylidene moieties using para-substitution [404] . 3.7 Anti- and prooxidant agents Violation of the antioxidant defence system and the balance of the pro/antioxidants are described under majority of pathologies including above mentioned. Therefore, a lot of papers are dedicated to the study of the 4-thiazolidones antioxidant activity [339] . Most of the articles describe investigation of the antiradical properties of 5-ylidenethiazolidinones in the model systems and evaluation of their efficiency as free radical scavengers (e.g. scavenging activity to DPPH (1,1-diphenyl-2-picrylhydrazyde) [106] ) combined with the study of other types of activity, in particular, antitumor and anti-inflammatory (see above) [405] . Such data concern all the classes of 4-thiazolidinones bearing different types of ylidene fragment. 4-Thiazolidinone derivative 178 ( Scheme 86 ) possesses anti-radical activity including the ability to inhibit superoxide anion formation [406] and it is known that antioxidant activity increases when additional carboxylic group is introduced in the substituent at the position C5 of the heterocyclic core in both N3 substituted and N3 unsubstituted derivatives [407] . Scheme 86 Structures of 4-thiazolidinones with anti-radical activity. Scheme 86 For the simple 5-benzylidene rhodanines antioxidant activity in various LDL oxidation models, such as TBARS assay, conjugated diene formation, REM of ox-LDL, fragmentation of apo β-100 by SDS-PAGE, radical DPPH scavenging activity, and macrophage-mediated LDL oxidation was showed [405] . It should be noted that the branched aliphatic fragments prevalence (spatially-screened phenols) largely contributes to the imitation of the known agents, e.g. BHT or probucol [336] . Antioxidant properties of a series of 2,4-dichlorothiazolylthiazolidine-2,4-dione and 4-chloro-2-benzylsulfanylthiazolyl-thiazolidine-2,4-dione derivatives 179 ( Scheme 86 ) were reported in two different in vitro assays: superoxide anion radical formation and DPPH radical scavenging activity [408] . The presence of the benzylsulfanyl group at the second position of the thiazole ring plays a significant role in increasing the superoxide anion scavenging activity. No correlation was observed in results on superoxide radical and DPPH radical scavenger capacity. Simple thiazolidinedione derivative ( Z )-5-(2,4-dihydroxybenzylidene)-thiazolidine-2,4-dione 180 ( Scheme 87 ) plays a crucial role in UV-induced melanogenesis, which is known to be related to the induction of tyrosinase enzyme . Compound inhibited nitroprusside-induced NO generation dose-dependently and suppressed tyrosinase activity and melanin synthesis (B16F10 melanoma cells) [409] . Scheme 87 Structures of 4-thiazolidinones with dual activity. Scheme 87 Series of 5-arylidene-2,4-thiazolidinediones 181 , 182 ( Scheme 86 ) and their 2-phenylimino analogs inhibiting AR [236] , [410] , [411] , [412] possess dual activity as antidiabetic and antioxidant agents for the treatment of diabetic complications [413] . The results indicate that latter possess excellent antioxidant properties inhibiting the production of TBARS in the test of compounds effect on hydroxyl radical-dependent lipoperoxidation induced in rat brain homogenate by the oxidant system Fe 3+ /ascorbic acid. It was observed that the presence of electron releasing substituents on the distal phenyl ring differentiates the antioxidant activity of the compounds. Detailed mechanisms of antioxidant effects of compounds of this class is still unclear and can be explained not only by the ability to inhibit superoxide anion (or other ROS) production, but also by influencing other parts of the pro/antioxidants system. Because of the changes in the classic approaches to the evaluation of pro/antioxidants [414] , emerging of new approaches to the interpretation of antioxidant effects [415] and study of other (not related to free radical scavenging) mechanisms of the antioxidants [156] , more attention is paid to the study of the molecular mechanisms of stimulation of antioxidant defence system [416] , [417] including the possibility of these substances pro-oxidant impact with subsequent stimulation of antioxidant defence system (see above – anticancer agents). 3.8 Other types of activity A large number of bio-targets of 5-ene-4-thiazolidinones are involved in different pathologies and there are a lot of data on the efficacy of 5-ene-thiazolidinones in different experimental models [5] , [6] , [7] , [100] , [418] . For instance, molecular docking studies indicated that the 5-ene-thiazolidinedione moiety was a likely candidate for its selectivity to monoamine oxidase-B . This hypothesis was confirmed by the identification of two lead-compounds 183 ( Scheme 88 ) with IC 50 13–20 μM [419] . Scheme 88 Examples of 5-ene-4-thiazolidinones with different types of activity. Scheme 88 Some 5-ene-rhodanine-3-carboxylic acid derivatibes, namely 4-(5 Z )-5-{5-(4-bromophenyl)-2-furyl]methylene}-4-oxo-2-thioxo-1,3-thiazolidin-3-yl-butanoic acid and 6-((5 Z )-5-{5-(4-bromophenyl)-2-furylmethylene}-4-oxo-2-thioxo-1,3-thiazolidin-3-yl)hexanoic acid are efficient inhibitors of apoptosis signal-regulating kinase 1 [420] which is involved in the variety of cellular processes [421] and can be considered as a promising target for the treatment of the various cardiovascular diseases, neurodegenerative diseases, several liver and kidney diseases etc. At the same time several other kinases, for example a cyclin-dependent kinase family are attractive targets for the design of high affinity inhibitors based on 5-alkylideenrhodanines [52] . 5-Ylidenes 184, 185 ( Scheme 88 ) were identified in the virtual screening as representatives of novel glycogen synthase kinase-3 ( GSK-3b ) inhibitors and their activity was confirmed experimentally [245] (ІС 50 1.56–5.56 μМ). GSK-3b has been emerging as a key therapeutic target for type-2 diabetes melitus, Alzheimer's disease, cancer and chronic inflammation. There is experimental evidence that GSK-3b inhibitors activate negative regulation of NF-kB activity, p53-dependent apoptosis, and enhance the TRAIL-induced cell death [422] , [423] . 3-(3-Trifluoromethyl)phenyl]-5-arylidene-2-thioxo-4-thiazolidinone was identified as a potent and selective blocker of the cystic fibrosis transmembrane conductance regulator ( CFTR ) Cl-channel . Moreover, it acted differently from other known blockers of the CFTR channel. Rather than blocking the channel pore, it affects the mechanism of channel gating, possibly by binding to nucleotide binding domain (NBD-1) [424] , [425] , [426] . CFTR , a member of the ATP-binding cassette transporter superfamily, is an epithelial chloride channel that plays a critical role in fluid absorption and secretion. Defective CFTR function causes cystic fibrosis, the most common lethal genetic disease in caucasians that produces severe lung disease, pancreatic insufficiency, neonatal intestinal obstruction and infertility where the hyperfunction of CFTR chloride channels , usually resulted from bacterial enterotoxins, constitutes the basic cause for secretory diarrhea. Thiazolidinone CFTR inhibitors may be useful in developing large-animal models of cystic fibrosis and in reducing intestinal fluid loss in cholera and other secretory diarrheas. Benzylidenerhodanine derivatives showed good inhibitory activity against recombinant human PRL-3 (phosphatase of regenerating liver, the representant of protein tyrosine phosphatases ). Compounds 186 and 187 ( Scheme 88 ) were the most active in vitro in this series and showed the ability to reduce the invasiveness of tumor ( 186 ) [286] . Following virtual screening of potent serotonin N-acetyltransferase (member of the GCN5N-acetyltransferase superfamily catalyzing the penultimate step in the biosynthesis of melatonin) inhibitors the class of 5-ene-rhodanine-3-carboxylic acids 188 ( Scheme 89 ) was identified that exhibited low micromolar competitive inhibition against acetyl-CoA and proved to be effective in blocking melatonin production in pineal cell [427] . Scheme 89 Structures of 5-ene-4-thiazolidinones with action on different targets. Scheme 89 5-Arylidene-4-thiazolidinone derivatives 189 ( Scheme 89 ) exhibited antidegenerative activity and could block multiple cartilage destruction during the osteoarthritic process by reducing NO release and restoring normal levels of glycosaminoglycans in chondrocytes treated with beta L-1, and possessed MMP-13 inhibition activity ( MMPs are a large family of calcium dependent, zinc containing endopeptidases responsible for the tissue remodeling and an extracellular matrix degradation) [240] . These compounds were found to be endowed with interesting activity levels in models of acute inflammation, such as carrageenan-induced paw and pleurisy edema in rats and to be inhibitors of COX isoforms [373] . 5-Benzylidene 2,4-thiazolidinedione derivatives are proposed for the treatment of Non-Alcoholic Fatty Liver Disease [428] . Compound 190 ( Scheme 89 ) was found to upregulate the adiponectin protein expression and down regulate the secretion of leptin protein in 3T3-L1 adipocytes at a respective concentration of 10 mM. 4-{( Z )-(2 Z )-2-(2-Fluorobenzylidene)-4-oxo-1,3-thiazolidin-5-ylidene]methyl}benzoic acid 191 ( Scheme 89 ) was described as the most potent full agonist to human GPR35 [429] . A class of orphan receptors GPR35 has been described as a potentially novel drug target [430] for the design of compounds with therapeutic application in a number of diseases including inflammation, metabolic disorders, nociception and cardiovascular diseases. The olefinic thiazolidinedione 192 ( Scheme 90 ) was found to be potent and selective b3-adrenergic receptor ( b3-AR ) (b3-Ar EC 50 0.006 mM, IA = 1.03) agonists (96-fold vs b2 and 492-fold vs b1 ) being more active than primary hits (5-saturated analogs) [431] . b3-AR Agonists are potential drugs for the treatment of obesity, type II diabetes, frequent urination and related diseases [432] . Related thiazolidinones with a cyanamide ( 193 ), hydroxylamine ( 194 ) ( Scheme 90 ) or 4-amino-1-benzylpiperidine substituents were generally very potent towards the b3 receptor, however, they were not very selective against both the b2 and b1 receptors. Related benzimidazolones ( 195 ) ( Scheme 90 ) displayed a good b3 agonist selectivity profile. Scheme 90 Structures of 5-ene-4-thiazolodinone b3-adrenergic receptor agonists. Scheme 90 5-Benzylidene-2-thioxo-4-thiazolidinones 196, 197 were described as inhibitors of Aggrecanase-2 that are members of the ADAMTS ( adisinte grin and metalloprotease with thrombospondin motifs ) family of zinc metalloproteases. The inhibition of Aggrecanase-2 , therefore, protect cartilage from damage and provide the first potential therapy to halt and/or reverse the progression of osteoarthritis and the compounds were at least 3-fold more active than 2,4-thiazolidinedione analogs [433] . In the result of virtual screening 198 ( Scheme 91 ) was found as potent canabinoid receptor type 1 antagonist with promising binding affinity (IC 50 125 nM) and also displayed good pharmacokinetic profile in rats as well as in human plasma [434] , [435] . Scheme 91 Structures of 5-ene-4-thiazolodinone b3-adrenergic receptor agonists and canabinoid receptor type 1 antagonist. Scheme 91 Novel class of sphingosine-1-phosphate ( S1P1 ) receptor agonists based on the 2-iminothiazolidin-4-one scaffold 199 and ACT-128800 ( Scheme 92 ) was found as the most active compound in in vitro and in vivo assays. Selective agonists of S1P1 receptor are of therapeutic interest for their ability to halt the exit of lymphocytes from lymph nodes. This interruption of lymphocyte migration promises a new immunomodulatory therapeutic principle for a variety of autoimmune diseases [436] . Scheme 92 Novel class of sphingosine-1-phosphate receptor agonists. Scheme 92 New class of benzisothiazolylimino-5-benzylidene-4-thiazolidinones was identified as new metalloproteinase-3 i nhibitors and chondroprotective agents. The mode of action is also related to the decreasing of NF-kB level [437] . Optovin ((5 E )-5-[(2,5-Dimethyl-1-pyridin-3-ylpyrrol-3-yl)methylidene]-2-sulfanylidene-1,3-thiazolidin-4-one) ( Scheme 93 ) was detected as a small molecule that enables repeatable photo activation of motor behaviors in wild type animals. Optovin acts as a light-sensitive ligand for anion channel involved in the detection of painful sensory stimuli. It activates human TRPA1 via structure-dependent photochemical reactions with redox-sensitive cysteine residues. Optovin treatment enables control of motor activity in the paralyzed extremities by localized illumination. These studies identify a light-based strategy for controlling endogenous TRPA1 [146] , [147] . It should be mentioned that optovin is a rare example of the E -isomers of 5-heterylidenerhodanines. Scheme 93 Structure of optovin. Scheme 93 Among rhodanine-3-carboxylic (sulfonic) acids the new inhibitors of histone acetyltransferases ( 200 – 202 ) ( Scheme 94 ) were identified. Histone acetyltransferases ( HATs ) are promising epigenetic drug targets involved in the pathogenesis of a wide range of diseases [427] , [438] . Scheme 94 Structures of inhibitors of histone acetyltransferases. Scheme 94 Summarizing all the above, 5-ene-4-thiazolidinones deservedly can be considered as privileged heterocycles with the wide spectrum of biological activity. Assigning 5-ene-thiazolidinones as PAINS due to possible Michael acceptor functionality is not so unambiguous and requires further study; therefore, most compounds can not be rejected per se . There are two main directions of the presented literature data: i) screening of the compounds activity without confirmed mode of action and ii) search for highly active and selective ligands to so-called "validated" bio-targets, which definitely don't negate each other. Search for antitumor, hypoglycemic, antimicrobial, antiviral and anti-inflammatory agents are the main fields of the 5-ene-4-thiazolidinones study. Moreover, most of the works are dedicated to the investigation of the antitumor potential of these compounds. Despite a series of established bio-targets, a detailed mode of anti-cancer effect undoubtedly depends on the structure of a particular derivative and needs further study. However, for the majority of 5-ene-thiazolidinones apoptosis-dependent as well as the ROS mediated modes of action are established. The main directions of the active 5-ene-thiazolidinones optimization can be devided as follows: i ) complications of the C5 fragment (bulky-substituents), while N3 unsubstituted position is preferred; ii ) introduction of substituents in N3 position (especially fragments bearing carboxylic group or its derivatives); iii ) annulation of 5-ene-thiazolidinones in complex fused heterocyclic systems; iv ) combination with other pharmacologically attractive fragments within hybride pharmocophore approach. 4 5-Ene-4-thiazolidinones based synthesis Following the diversity of the chemical and pharmacological properties of 5-ene-4-thiazolidinones they can be treated as the building blocks for the synthesis of various derivatives and as a step of so-called privileged substructure-based diversity oriented synthesis strategy [15] , [16] , [17] , [439] , which is proved to be a fruitful tool to rapidly discover biologically active lead-compounds. Moreover, recently the following thesis was put forward: the compounds based on 5-ene-4-thiazolidinones, mainly annelated systems (e.g. thiopyrano[2,3- d ]thiazoles) possess the similar to 5-ene-4-thiazolidinones pharmacological profiles and can be considered as cyclic bio-mimetics of their synthetic precursors [440] , [441] , [442] , [443] , [444] . Regardless of the type of thiazolidinone core, we tried to outline the reactions common for all 5-ene-4-thiazolidinones and in particular for those involving the C5 exocyclic double bond [445] . 4.1 Addition reactions Reduction. The C5 exocyclic double bond of 5-ene-4-thiazolidinones which formed the planar conjugated systems could be relatively easy reduced leading to the formation of corresponding unconjugated structures ( Scheme 95 ). Challenging problem of such type of non-conjugated systems (that are basically most studied on the example of glitazones (see above)) is the relative simplicity of enolization at the 5-position under physiological conditions, which makes the stereochemistry difficult to maintain at this position [446] . Scheme 95 General scheme of reduction and enolization. Scheme 95 Lithium and sodium borohydride and sodium hydrides [234] , [303] , [447] , H 2 –Pd/C in dioxane [223] , Na/Hg, THF/H 2 O [431] , Mg in methanol or zinc in acetic acid [448] etc. have been used as reducing agents. This reaction is mainly studied as a phase of glytazone synthesis [449] . An appropriate reaction conditions allow selective reducing of the double bond leaving intact other molecular fragments, e.g. in the case of the rhodanine-3-carboxylic acids. Under reduction of the complex 5-heterylidene-4-thiazolidinones 203 ( Scheme 96 ) the formation of fused heterocyclic systems 204 was described. The treatment of pyrazole–thiazolidinone derivatives with sodium hydride in N,N-dimethylformamide caused dimerisation reaction to give the corresponding spiro compounds 204 [450] . Scheme 96 Synthesis of spiro compounds under complex 5-heterylidene-4-thiazolidinones reduction. Scheme 96 The exocyclic double bond of 5-arylidene-3-methyl-2-thioxo-4-thiazolidinones is a good olefin for variety addition reactions, e.g. Michael addition [451] . The reaction conditions (e.g. microwave irradiation) confirm the doubts about the realization of such reaction under physiological conditions (see above) considering 5-ene-thiazolidinones as Michael acceptors. While the Michael addition (followed the Knoevenagel condensation) [452] is a phase of domino or tandem multicomponent reactions. Reaction of 3-aroylrhodanines, aromatic aldehydes and N -dithiocarbaminates involves the consecutive Knoevenagel condensation and Michael addition ( Scheme 97 ). The reaction is stereoselective and is carried out in dry conditions under microwave irradiation. Dithioester 205 under intermolecular cyclization and iodine ( 207 ), montmorillonite Li + clay ( 206 ) or montmorillonite K-10 clay ( 208 ) action gave the appropriate heterocycles [453] . Scheme 97 Michael addition as a phase of multicomponent reaction in the synthesis of fused heterocycles. Scheme 97 Utilization of the 5-carboxymethylidene-rhodanines 209 (as [C 2 ] 2+ synthones) in the known [2 + 3]-cyclocondensation reaction led to bis-thiazolidinones 210 formation. The latter can be easely reduced (adding zinc in acetic acid) to bis-thiazolidinones 211 which can readily be oxidized to starting 209 (in the presence of catalitic ammounts of triethylamine) [454] , [455] ( Scheme 98 ). Scheme 98 Synthesis of bis-thiazolidinones. Scheme 98 Similarly, 5-heterylidene-4-thiazolidinones 213 ( Scheme 99 ) have been synthesized based on 5-methoxycarbonylmethylidene-4-thiazolidones 209 in the reaction with o -aminothiophenol and following dehydrogenation of appropriate 5-alkyl-4-thiazolidinones ( 212 ) in the presence of triethylamine in DMF [456] . Scheme 99 Synthetic approach to C5 fragment complication. Scheme 99 Three-component, one-pot procedure for the synthesis of pyran-annulated thiazoles 216 involving the in situ generation of azlactone ( 215 ) and formation of pyranothiazoles is best explained by Michael addition of azlactone to 5-ene-rhodanine ( 214 ) affording the corresponding Michael adducts which underwent ring transformation to yield the final products [457] ( Scheme 100 ). Scheme 100 Synthesis of 5-oxo-2-thioxo-2,3,6,7-tetrahydropyrano[2,3- d ]thiazol-5-ones. Scheme 100 5-Oxo-2-thioxo-2,3,4,5,6,7-hexahydro[1,3]thiazolo[4,5- b ]pyridine-6-carbonitriles 217 ( Scheme 101 ) were obtained by refluxing 5-ene-rhodanines with ethyl cyanoacetate or malonodinitrile in acetic acid in the presence of ammonium acetate. First stage of the process is base-catalyzed condensation of ethyl cyanoacetate and the α,β-unsaturated ketone fragment of 5-ene-rhodanine ( 214 ). The subsequent intramolecular cyclization of Michael adduct with elimination of water leads to the formation of final product [133] , [458] , [459] , [460] . Scheme 101 Synthesis of 5-oxo-2-thioxo-2,3,4,5,6,7-hexahydro[1,3]thiazolo[4,5- b ]pyridines. Scheme 101 5-Ene-rhodanines were successfully applied in cascade reactions for the construction of spiro rhodanines 218 ( Scheme 102 ) with multiple consecutive chiral centers catalyzed by a simple diamine (the most efficient is 2S-N 1 -cyclohexyl-4,4-dimethylpentane-1,2-diamine and N-Boc- l -tryptophan), providing products with high stereoselectivities [461] . Scheme 102 Synthesis of spiro rhodanines. Scheme 102 5-Ene-2-spirothiazolidine has been used as a component of Michael addition with hydroxylamine hydrochloride to yield the novel spiro (cyclohexane-isoxazolo-thiazole) derivative 219. In addition, α,β-unsaturated ketone (5-ene-2-spirothiazolidine) was used for the synthesis of corresponding oxiranyl derivatives 220 by treatment with hydrogen peroxide in the presence of sodium hydroxide. Also, starting thiazolidinone was treated with ethylcyanoacetate (medium of acetic acid in the presence of sodium acetate) under reflux to afford the corresponding pyrano-thiazole-carbonitrile derivative 221 [462] ( Scheme 103 ). Scheme 103 Utilization of 5-ene-2-spirothiazolidine. Scheme 103 Similar heterocycles based on the same starting compound (5-ylidene-3-phenyl-cyclohexane(1′-2)thiazolidin-4-one) are described in the reaction with the difunctional nucleophiles (thiourea, hydrazine hydrate derivatives, malononitrile and ethylcyanoacetate) that yielded thiazolo[4,5- d ]pyrimidine, pyrazolo[3,4- d ]thiazole, thiazolo[4,5- b ]pyridine derivatives 222 [463] ( Scheme 104 ). Scheme 104 Reactions of 5-ene-2-spiro-4-thiazolidinones with difunctional nucleophiles yielding different fused heterocycles. Scheme 104 The bromination of the double bond is the efficient approach to the synthesis of compounds 223 with high antimicrobial activity [464] , [465] ( Scheme 105 ). Scheme 105 Bromination of 5- ylidene-4-thiazolidinones. Scheme 105 The 5-arylidene derivatives reacted with Grignard reagent via 1,4-conjugate addition to the exocyclic double bond to afford 5-alkylaryl-4-thiazolidinones 224 [466] , [467] , [468] , [469] ( Scheme 106 ). Such reaction has been rarely found in the current literature data [78] . Scheme 106 Reaction of 5-arylidene-thiazolidin-4-one with Grignard reagent. Scheme 106 5-Benzylidene thiazolidinone derivatives having a conjugated carbonyl group reacted with such difunctional nucleophile as phenylhydrazine in ethanol in the presence of sodium acetate to give tetrahydro-5 H -pyrazolo[3,4- d ]1,3]thiazole-5-(thi)ones 225 [458] , [459] , [460] ( Scheme 107 ). Scheme 107 Synthesis of pyrazolothiazole starting from 5-benzylidene thiazolidinone. Scheme 107 4.2 Hydrolytic cleavage 4-Thiazolidinone core is not stable in the alkali medium and this feature has been often explored in their cyclization reactions. 5-Ylidene-rhodanines are attractive starting reagents for the synthesis of 3-substituted-2-mercaptoacrylic acids 226 [43] , [470] . Alkaline hydrolysis of 5-ylidenerhodanines leads to the mentioned acids formation [471] ( Scheme 108 ). Scheme 108 Synthesis of 3-substituted-2-mercaptoacrylic acids under the alkaline hydrolysis of 5-ene-rhodanines. Scheme 108 The possibility of 3-substituted-2-mercaptoacrylic acids utilization as thiolic agents in the one-pot, three-component reaction for the synthesis of 5-ene-2,3-disubstituted-4-thiazolidinones was proposed based on the retro-synthetic approach [43] . The 5-arylidene-4-thiazolidinones containing substituents in the ortho position of arylidene fragment are of special interest. The efficient approach to the synthesis of isothiocoumarin derivatives 227 and related compounds 228 – 231 (structures of which depends on the nature of substituent in the ortho- position of aryl(heter)ylidene fragment) is based on the above mentioned 5-ene-4-thiazolidinones [472] , [473] , [474] , [475] , [476] , [477] ( Scheme 109 ). Scheme 109 Synthesis of different polycyclic compounds based on 5-arylidene-4-thiazolidinones (adapted from [477] ). Scheme 109 In the base hydrolysis of rhodanines two-step reaction occurs in which intermediates of the hydrolysis undergo heterocyclization with the formation of the 1-oxo-1 H -2-benzothiopyran-3-carboxylic acids, 3-mercaptocoumarins [478] , [479] , ketocinchoninic acids, 2-indole-carboxylic acids etc. 5-Arylidene-4-thiazolidinones under the action of hydrazine/thiosemicarbazone yield pyrazoline 232 and 1,2,4-triazine 233 derivatives [480] ( Scheme 110 ). Scheme 110 Transformation of 5-arylidene-thiazolidinones under the action of hydrazine/thiosemicarbazone. Scheme 110 The synthesis of similar pyrazole derivatives 234 is also described starting from 2-ylidene-substituted 4-thiazolidinones [481] ( Scheme 111 ). Scheme 111 Synthesis of pyrazoles from 2-ylidene-substituted 4-thiazolidinones. Scheme 111 Simple synthesis of 2-thioxo-6-azauracils was developed based on the reaction of 5-substituted 2,4-thiazolidinediones with thiosemicarbazide in alkaline medium [482] . Based on the above mentioned approach to C5-moiety optimization, the introduction of bulky heterocyclic fragments into 4-thiazolidinone molecule via the recyclization of C5 ylidene fragment containing highly reactive groups (mainly in ortho position) was proposed ( Scheme 112 ). Scheme 112 Examples of the reactions leading to introduction of bulky heterocyclic moieties into C5 of 4-thiazolidinone core. Scheme 112 Though, presented protocols for 235 – 237 [75] , [483] are not persuasive, taking into consideration the transformation (instability) of 4-thiazolidinone core in the basic medium (see below) [1] , [2] , [6] . 4.3 Hetero-Diels-Alder reaction and related processes The thionation is one of the effective approaches to the modification of 4-thiazolidinones (mainly in the position 4) and one of the steps of the multicomponent reactions in the synthesis of fused heterocycles. For example, thionation of 5-aroylmethylidene- or 5-aroylmethylrhodanines led to 2,3-dihydro-2-thioxothieno[2,3- d ]thiazoles 238 [93] ( Scheme 113 ). Scheme 113 Synthesis of 2,3-dihydro-2-thioxothieno[2,3- d ]thiazoles. Scheme 113 In order to synthezise the pyrazolo[3,4- d ]thiazole system the condensation of 5-ene-4-thioxothiazolidine with hydrazines were studied [484] . Thiazolidine-thione condensation with hydrazine derivatives in DMF yielded a single product in regioselective synthesis of 1,3,5-trisubstituted-5,6-dihydro-2 H -pyrazolo[3,4- d ]thiazoles 239 [76] ( Scheme 114 ). Scheme 114 Utilization of 5-ene-4-thioxothiazolidine in the reaction with hydrazines. Scheme 114 4-Thioxothiazolidines are efficient heterodienes in hetero -Diels–Alder cycloaddition being one of the most powerful methods in the construction of fused heterocycles 240 bearing thiazolidinone fragment ( Scheme 115 ). Scheme 115 Synthesis of fused thiopyranothiazole system. Scheme 115 Thiopyrano[2,3- d ][1,3]thiazoles are usually synthesized via [4 + 2]-cycloaddition in hetero -Diels-Alder reaction of 5-arylidene-2,4-dithioxothiazolidines (5-arylidenethiorhodanines) or 5-arylidene-4-thioxo-2-thiazolidinones (5-arylideneisorhodanines) which contain С = С 5 -С 4 = S group in their structure and are active heterodienes [21] , [440] , [441] , [442] , [443] , [444] , [482] . Thiopyrano[2,3- d ][1,3]thiazoles are of a special interest as cyclic isosteric mimics of their synthetic precursors 5-arylidene-4-thiazolidinones without Michael accepting functionalities (see above) [440] , [441] , [442] , [443] , [444] . Maleic acid and its derivatives, acrylic acid and its derivatives, acrylonitrile, β-nitrostyrene, norbornene, 5-norbonene-2,3-dicarboxylic acid derivatives, cynnamic acids, propiolic acid, aroylpyruvic acid etc. have been studied in such heterodiene condensation as dienophiles [21] , [440] , [441] , [442] , [443] , [444] , [485] , [486] , [487] , [488] . The cycloadditions are highly regio- and srereoselective and form products according to Frontier Orbital Theory [67] , [69] , [489] , [490] . Interesting results were described when 5-benzylidenerhodanines with substituent in ortho -position in phenyl ring were utilized in such type of reactions. The reactions between the latter and crotonic acid, its anhydride or 4-chlorophenylamide, cynnamic acids [490] , acrylic acid and its ester involved tandem hetero -Dielse-Alder and acylation processes, affording tetracyclic fused heterocycles (chromeno[4′,3′:4,5]thiopyrano[2,3- d ]thiazoles) 241 ( Scheme 116 ). Scheme 116 Synthesis of chromeno[4′,3′:4,5]thiopyrano[2,3- d ]thiazoles. Scheme 116 Similar 2-hydroxybenzylidene rhodanine derivatives reacted with malononitrile to give fused chromeno[4′,3′:4,5]pyrano[2,3- d ]thiazol-6-ones 242 . Due to instability of rhodanines in basic medium at high temperatures the yields of the cyclocondensation products are low [458] ( Scheme 117 ). Scheme 117 Synthesis of chromeno[4′,3′:4,5]pyrano[2,3- d ]thiazol-6-ones. Scheme 117 5-Ene-4-thiazolidinones are often the intermediates in the above mentioned tandem reactions involving the hetero -Dielse-Alder reaction [67] , [69] , [483] , [484] , [490] ( Scheme 118 ). Scheme 118 Examples of 5-ene-4-thiazolidinones as intermediates in tandem reactions. Scheme 118 Rare are the reports about the utilization of 2-imino-4-thiazolidinones [491] and rhodanine derivatives in the hetero -Diels-Alder reaction. For example, in the reaction of 2-imino-4-thiazolidinones and norbornene fused tiopyranothiazole derivatives are formed 244 ( Scheme 119 ). Scheme 119 Utilization of 2-imino-4-thiazolidinones in the hetero -Diels-Alder reaction. Scheme 119 The reactions of arylidenerhodanines with maleic anhydride, N -phenylmaleimide, and DMAD were performed under MW irradiation and have led to pyrano[2,3- d ][1,3]thiazoles formation [492] . Besides 5-arylidene derivatives, other types of 5-ene-4-thioxothiazolidines were described. For instance, 5-ethoxy-4-thioxothiazolidines are starting materials for polycyclic systems 245 – 247 formation [64] ( Scheme 120 ). The final product formation is accompanied by the elimination of alcohol molecule and oxidation in the case of 247 . Scheme 120 Synthesis of polycyclic molecules based on 5-ene-4-thioxothiazolidines. Scheme 120 Rhodanine derivatives were successfully used as the dienophiles to react with the various 2,4-dienals in the asymmetric Diels-Alder reaction leading to the construction of structurally complex compounds 248 containing the rhodanine motif [493] ( Scheme 121 ). The reaction is also diastereo- and enantioselective. Scheme 121 Rhodanine derivatives as dienophiles in the Diels-Alder reaction. Scheme 121 4.4 Other cycloaddition reactions Three-component 1,3-dipolar cycloaddition reaction of isatin, sarcosine and 5-arylidene-1,3-thiazolidine-2,4-dione or 5-arylidene-4-thioxo-1,3-thiazolidine-2-one in ethanol under ultrasound irradiation is an efficient protocol for the synthesis of dispiropyrrolidine derivatives 249 [494] ( Scheme 122 ). Scheme 122 One-pot approach for the synthesis of spiro compounds using 5-arylidene-thiazolidinones. Scheme 122 The analougues two step condensation reactions affording similar complex spiro derivatives 250 are described based on 5-arylidene-thiazolo[3,2-b][1,2,4]triazol-6(5 H )-ones which contain 'fixed' 5-arylidenethiazolidinone fragment in their structure [495] ( Scheme 122 ). Related dispiroindole-[3,3′-pyrrolidine-[4′,5″-1,3]thiazolidine]-2,4"(1 H )-diones 251 were synthesized starting from isatin-thiazolidione conjugates in cyclocondensation reaction ( Scheme 123 ). Mentioned compounds were formed as racemic mixture of two enantiomers resulting from the addition of azomethine ylide (the reaction proceeded diastereoselectively – following NMR data) [496] . Scheme 123 Synthesis of dispiroindole-[3,3′-pyrrolidine-[4′,5″-1,3]thiazolidine]-2,4"(1 H )-diones. Scheme 123 The 5-arylidene-4-thiazolidinones in the Michael reaction with 5-aminopyrazole in acetic acid afforded fused heterocyclic systems 252 containing anazolone moiety. The reaction involved the intramolecular cyclization of the appropriate intermediates and oxidative aromatization [497] ( Scheme 124 ). Scheme 124 Example of 5-arylidene-4-thiazolidinones utilization in Michael reaction. Scheme 124 4.1 Addition reactions Reduction. The C5 exocyclic double bond of 5-ene-4-thiazolidinones which formed the planar conjugated systems could be relatively easy reduced leading to the formation of corresponding unconjugated structures ( Scheme 95 ). Challenging problem of such type of non-conjugated systems (that are basically most studied on the example of glitazones (see above)) is the relative simplicity of enolization at the 5-position under physiological conditions, which makes the stereochemistry difficult to maintain at this position [446] . Scheme 95 General scheme of reduction and enolization. Scheme 95 Lithium and sodium borohydride and sodium hydrides [234] , [303] , [447] , H 2 –Pd/C in dioxane [223] , Na/Hg, THF/H 2 O [431] , Mg in methanol or zinc in acetic acid [448] etc. have been used as reducing agents. This reaction is mainly studied as a phase of glytazone synthesis [449] . An appropriate reaction conditions allow selective reducing of the double bond leaving intact other molecular fragments, e.g. in the case of the rhodanine-3-carboxylic acids. Under reduction of the complex 5-heterylidene-4-thiazolidinones 203 ( Scheme 96 ) the formation of fused heterocyclic systems 204 was described. The treatment of pyrazole–thiazolidinone derivatives with sodium hydride in N,N-dimethylformamide caused dimerisation reaction to give the corresponding spiro compounds 204 [450] . Scheme 96 Synthesis of spiro compounds under complex 5-heterylidene-4-thiazolidinones reduction. Scheme 96 The exocyclic double bond of 5-arylidene-3-methyl-2-thioxo-4-thiazolidinones is a good olefin for variety addition reactions, e.g. Michael addition [451] . The reaction conditions (e.g. microwave irradiation) confirm the doubts about the realization of such reaction under physiological conditions (see above) considering 5-ene-thiazolidinones as Michael acceptors. While the Michael addition (followed the Knoevenagel condensation) [452] is a phase of domino or tandem multicomponent reactions. Reaction of 3-aroylrhodanines, aromatic aldehydes and N -dithiocarbaminates involves the consecutive Knoevenagel condensation and Michael addition ( Scheme 97 ). The reaction is stereoselective and is carried out in dry conditions under microwave irradiation. Dithioester 205 under intermolecular cyclization and iodine ( 207 ), montmorillonite Li + clay ( 206 ) or montmorillonite K-10 clay ( 208 ) action gave the appropriate heterocycles [453] . Scheme 97 Michael addition as a phase of multicomponent reaction in the synthesis of fused heterocycles. Scheme 97 Utilization of the 5-carboxymethylidene-rhodanines 209 (as [C 2 ] 2+ synthones) in the known [2 + 3]-cyclocondensation reaction led to bis-thiazolidinones 210 formation. The latter can be easely reduced (adding zinc in acetic acid) to bis-thiazolidinones 211 which can readily be oxidized to starting 209 (in the presence of catalitic ammounts of triethylamine) [454] , [455] ( Scheme 98 ). Scheme 98 Synthesis of bis-thiazolidinones. Scheme 98 Similarly, 5-heterylidene-4-thiazolidinones 213 ( Scheme 99 ) have been synthesized based on 5-methoxycarbonylmethylidene-4-thiazolidones 209 in the reaction with o -aminothiophenol and following dehydrogenation of appropriate 5-alkyl-4-thiazolidinones ( 212 ) in the presence of triethylamine in DMF [456] . Scheme 99 Synthetic approach to C5 fragment complication. Scheme 99 Three-component, one-pot procedure for the synthesis of pyran-annulated thiazoles 216 involving the in situ generation of azlactone ( 215 ) and formation of pyranothiazoles is best explained by Michael addition of azlactone to 5-ene-rhodanine ( 214 ) affording the corresponding Michael adducts which underwent ring transformation to yield the final products [457] ( Scheme 100 ). Scheme 100 Synthesis of 5-oxo-2-thioxo-2,3,6,7-tetrahydropyrano[2,3- d ]thiazol-5-ones. Scheme 100 5-Oxo-2-thioxo-2,3,4,5,6,7-hexahydro[1,3]thiazolo[4,5- b ]pyridine-6-carbonitriles 217 ( Scheme 101 ) were obtained by refluxing 5-ene-rhodanines with ethyl cyanoacetate or malonodinitrile in acetic acid in the presence of ammonium acetate. First stage of the process is base-catalyzed condensation of ethyl cyanoacetate and the α,β-unsaturated ketone fragment of 5-ene-rhodanine ( 214 ). The subsequent intramolecular cyclization of Michael adduct with elimination of water leads to the formation of final product [133] , [458] , [459] , [460] . Scheme 101 Synthesis of 5-oxo-2-thioxo-2,3,4,5,6,7-hexahydro[1,3]thiazolo[4,5- b ]pyridines. Scheme 101 5-Ene-rhodanines were successfully applied in cascade reactions for the construction of spiro rhodanines 218 ( Scheme 102 ) with multiple consecutive chiral centers catalyzed by a simple diamine (the most efficient is 2S-N 1 -cyclohexyl-4,4-dimethylpentane-1,2-diamine and N-Boc- l -tryptophan), providing products with high stereoselectivities [461] . Scheme 102 Synthesis of spiro rhodanines. Scheme 102 5-Ene-2-spirothiazolidine has been used as a component of Michael addition with hydroxylamine hydrochloride to yield the novel spiro (cyclohexane-isoxazolo-thiazole) derivative 219. In addition, α,β-unsaturated ketone (5-ene-2-spirothiazolidine) was used for the synthesis of corresponding oxiranyl derivatives 220 by treatment with hydrogen peroxide in the presence of sodium hydroxide. Also, starting thiazolidinone was treated with ethylcyanoacetate (medium of acetic acid in the presence of sodium acetate) under reflux to afford the corresponding pyrano-thiazole-carbonitrile derivative 221 [462] ( Scheme 103 ). Scheme 103 Utilization of 5-ene-2-spirothiazolidine. Scheme 103 Similar heterocycles based on the same starting compound (5-ylidene-3-phenyl-cyclohexane(1′-2)thiazolidin-4-one) are described in the reaction with the difunctional nucleophiles (thiourea, hydrazine hydrate derivatives, malononitrile and ethylcyanoacetate) that yielded thiazolo[4,5- d ]pyrimidine, pyrazolo[3,4- d ]thiazole, thiazolo[4,5- b ]pyridine derivatives 222 [463] ( Scheme 104 ). Scheme 104 Reactions of 5-ene-2-spiro-4-thiazolidinones with difunctional nucleophiles yielding different fused heterocycles. Scheme 104 The bromination of the double bond is the efficient approach to the synthesis of compounds 223 with high antimicrobial activity [464] , [465] ( Scheme 105 ). Scheme 105 Bromination of 5- ylidene-4-thiazolidinones. Scheme 105 The 5-arylidene derivatives reacted with Grignard reagent via 1,4-conjugate addition to the exocyclic double bond to afford 5-alkylaryl-4-thiazolidinones 224 [466] , [467] , [468] , [469] ( Scheme 106 ). Such reaction has been rarely found in the current literature data [78] . Scheme 106 Reaction of 5-arylidene-thiazolidin-4-one with Grignard reagent. Scheme 106 5-Benzylidene thiazolidinone derivatives having a conjugated carbonyl group reacted with such difunctional nucleophile as phenylhydrazine in ethanol in the presence of sodium acetate to give tetrahydro-5 H -pyrazolo[3,4- d ]1,3]thiazole-5-(thi)ones 225 [458] , [459] , [460] ( Scheme 107 ). Scheme 107 Synthesis of pyrazolothiazole starting from 5-benzylidene thiazolidinone. Scheme 107 4.2 Hydrolytic cleavage 4-Thiazolidinone core is not stable in the alkali medium and this feature has been often explored in their cyclization reactions. 5-Ylidene-rhodanines are attractive starting reagents for the synthesis of 3-substituted-2-mercaptoacrylic acids 226 [43] , [470] . Alkaline hydrolysis of 5-ylidenerhodanines leads to the mentioned acids formation [471] ( Scheme 108 ). Scheme 108 Synthesis of 3-substituted-2-mercaptoacrylic acids under the alkaline hydrolysis of 5-ene-rhodanines. Scheme 108 The possibility of 3-substituted-2-mercaptoacrylic acids utilization as thiolic agents in the one-pot, three-component reaction for the synthesis of 5-ene-2,3-disubstituted-4-thiazolidinones was proposed based on the retro-synthetic approach [43] . The 5-arylidene-4-thiazolidinones containing substituents in the ortho position of arylidene fragment are of special interest. The efficient approach to the synthesis of isothiocoumarin derivatives 227 and related compounds 228 – 231 (structures of which depends on the nature of substituent in the ortho- position of aryl(heter)ylidene fragment) is based on the above mentioned 5-ene-4-thiazolidinones [472] , [473] , [474] , [475] , [476] , [477] ( Scheme 109 ). Scheme 109 Synthesis of different polycyclic compounds based on 5-arylidene-4-thiazolidinones (adapted from [477] ). Scheme 109 In the base hydrolysis of rhodanines two-step reaction occurs in which intermediates of the hydrolysis undergo heterocyclization with the formation of the 1-oxo-1 H -2-benzothiopyran-3-carboxylic acids, 3-mercaptocoumarins [478] , [479] , ketocinchoninic acids, 2-indole-carboxylic acids etc. 5-Arylidene-4-thiazolidinones under the action of hydrazine/thiosemicarbazone yield pyrazoline 232 and 1,2,4-triazine 233 derivatives [480] ( Scheme 110 ). Scheme 110 Transformation of 5-arylidene-thiazolidinones under the action of hydrazine/thiosemicarbazone. Scheme 110 The synthesis of similar pyrazole derivatives 234 is also described starting from 2-ylidene-substituted 4-thiazolidinones [481] ( Scheme 111 ). Scheme 111 Synthesis of pyrazoles from 2-ylidene-substituted 4-thiazolidinones. Scheme 111 Simple synthesis of 2-thioxo-6-azauracils was developed based on the reaction of 5-substituted 2,4-thiazolidinediones with thiosemicarbazide in alkaline medium [482] . Based on the above mentioned approach to C5-moiety optimization, the introduction of bulky heterocyclic fragments into 4-thiazolidinone molecule via the recyclization of C5 ylidene fragment containing highly reactive groups (mainly in ortho position) was proposed ( Scheme 112 ). Scheme 112 Examples of the reactions leading to introduction of bulky heterocyclic moieties into C5 of 4-thiazolidinone core. Scheme 112 Though, presented protocols for 235 – 237 [75] , [483] are not persuasive, taking into consideration the transformation (instability) of 4-thiazolidinone core in the basic medium (see below) [1] , [2] , [6] . 4.3 Hetero-Diels-Alder reaction and related processes The thionation is one of the effective approaches to the modification of 4-thiazolidinones (mainly in the position 4) and one of the steps of the multicomponent reactions in the synthesis of fused heterocycles. For example, thionation of 5-aroylmethylidene- or 5-aroylmethylrhodanines led to 2,3-dihydro-2-thioxothieno[2,3- d ]thiazoles 238 [93] ( Scheme 113 ). Scheme 113 Synthesis of 2,3-dihydro-2-thioxothieno[2,3- d ]thiazoles. Scheme 113 In order to synthezise the pyrazolo[3,4- d ]thiazole system the condensation of 5-ene-4-thioxothiazolidine with hydrazines were studied [484] . Thiazolidine-thione condensation with hydrazine derivatives in DMF yielded a single product in regioselective synthesis of 1,3,5-trisubstituted-5,6-dihydro-2 H -pyrazolo[3,4- d ]thiazoles 239 [76] ( Scheme 114 ). Scheme 114 Utilization of 5-ene-4-thioxothiazolidine in the reaction with hydrazines. Scheme 114 4-Thioxothiazolidines are efficient heterodienes in hetero -Diels–Alder cycloaddition being one of the most powerful methods in the construction of fused heterocycles 240 bearing thiazolidinone fragment ( Scheme 115 ). Scheme 115 Synthesis of fused thiopyranothiazole system. Scheme 115 Thiopyrano[2,3- d ][1,3]thiazoles are usually synthesized via [4 + 2]-cycloaddition in hetero -Diels-Alder reaction of 5-arylidene-2,4-dithioxothiazolidines (5-arylidenethiorhodanines) or 5-arylidene-4-thioxo-2-thiazolidinones (5-arylideneisorhodanines) which contain С = С 5 -С 4 = S group in their structure and are active heterodienes [21] , [440] , [441] , [442] , [443] , [444] , [482] . Thiopyrano[2,3- d ][1,3]thiazoles are of a special interest as cyclic isosteric mimics of their synthetic precursors 5-arylidene-4-thiazolidinones without Michael accepting functionalities (see above) [440] , [441] , [442] , [443] , [444] . Maleic acid and its derivatives, acrylic acid and its derivatives, acrylonitrile, β-nitrostyrene, norbornene, 5-norbonene-2,3-dicarboxylic acid derivatives, cynnamic acids, propiolic acid, aroylpyruvic acid etc. have been studied in such heterodiene condensation as dienophiles [21] , [440] , [441] , [442] , [443] , [444] , [485] , [486] , [487] , [488] . The cycloadditions are highly regio- and srereoselective and form products according to Frontier Orbital Theory [67] , [69] , [489] , [490] . Interesting results were described when 5-benzylidenerhodanines with substituent in ortho -position in phenyl ring were utilized in such type of reactions. The reactions between the latter and crotonic acid, its anhydride or 4-chlorophenylamide, cynnamic acids [490] , acrylic acid and its ester involved tandem hetero -Dielse-Alder and acylation processes, affording tetracyclic fused heterocycles (chromeno[4′,3′:4,5]thiopyrano[2,3- d ]thiazoles) 241 ( Scheme 116 ). Scheme 116 Synthesis of chromeno[4′,3′:4,5]thiopyrano[2,3- d ]thiazoles. Scheme 116 Similar 2-hydroxybenzylidene rhodanine derivatives reacted with malononitrile to give fused chromeno[4′,3′:4,5]pyrano[2,3- d ]thiazol-6-ones 242 . Due to instability of rhodanines in basic medium at high temperatures the yields of the cyclocondensation products are low [458] ( Scheme 117 ). Scheme 117 Synthesis of chromeno[4′,3′:4,5]pyrano[2,3- d ]thiazol-6-ones. Scheme 117 5-Ene-4-thiazolidinones are often the intermediates in the above mentioned tandem reactions involving the hetero -Dielse-Alder reaction [67] , [69] , [483] , [484] , [490] ( Scheme 118 ). Scheme 118 Examples of 5-ene-4-thiazolidinones as intermediates in tandem reactions. Scheme 118 Rare are the reports about the utilization of 2-imino-4-thiazolidinones [491] and rhodanine derivatives in the hetero -Diels-Alder reaction. For example, in the reaction of 2-imino-4-thiazolidinones and norbornene fused tiopyranothiazole derivatives are formed 244 ( Scheme 119 ). Scheme 119 Utilization of 2-imino-4-thiazolidinones in the hetero -Diels-Alder reaction. Scheme 119 The reactions of arylidenerhodanines with maleic anhydride, N -phenylmaleimide, and DMAD were performed under MW irradiation and have led to pyrano[2,3- d ][1,3]thiazoles formation [492] . Besides 5-arylidene derivatives, other types of 5-ene-4-thioxothiazolidines were described. For instance, 5-ethoxy-4-thioxothiazolidines are starting materials for polycyclic systems 245 – 247 formation [64] ( Scheme 120 ). The final product formation is accompanied by the elimination of alcohol molecule and oxidation in the case of 247 . Scheme 120 Synthesis of polycyclic molecules based on 5-ene-4-thioxothiazolidines. Scheme 120 Rhodanine derivatives were successfully used as the dienophiles to react with the various 2,4-dienals in the asymmetric Diels-Alder reaction leading to the construction of structurally complex compounds 248 containing the rhodanine motif [493] ( Scheme 121 ). The reaction is also diastereo- and enantioselective. Scheme 121 Rhodanine derivatives as dienophiles in the Diels-Alder reaction. Scheme 121 4.4 Other cycloaddition reactions Three-component 1,3-dipolar cycloaddition reaction of isatin, sarcosine and 5-arylidene-1,3-thiazolidine-2,4-dione or 5-arylidene-4-thioxo-1,3-thiazolidine-2-one in ethanol under ultrasound irradiation is an efficient protocol for the synthesis of dispiropyrrolidine derivatives 249 [494] ( Scheme 122 ). Scheme 122 One-pot approach for the synthesis of spiro compounds using 5-arylidene-thiazolidinones. Scheme 122 The analougues two step condensation reactions affording similar complex spiro derivatives 250 are described based on 5-arylidene-thiazolo[3,2-b][1,2,4]triazol-6(5 H )-ones which contain 'fixed' 5-arylidenethiazolidinone fragment in their structure [495] ( Scheme 122 ). Related dispiroindole-[3,3′-pyrrolidine-[4′,5″-1,3]thiazolidine]-2,4"(1 H )-diones 251 were synthesized starting from isatin-thiazolidione conjugates in cyclocondensation reaction ( Scheme 123 ). Mentioned compounds were formed as racemic mixture of two enantiomers resulting from the addition of azomethine ylide (the reaction proceeded diastereoselectively – following NMR data) [496] . Scheme 123 Synthesis of dispiroindole-[3,3′-pyrrolidine-[4′,5″-1,3]thiazolidine]-2,4"(1 H )-diones. Scheme 123 The 5-arylidene-4-thiazolidinones in the Michael reaction with 5-aminopyrazole in acetic acid afforded fused heterocyclic systems 252 containing anazolone moiety. The reaction involved the intramolecular cyclization of the appropriate intermediates and oxidative aromatization [497] ( Scheme 124 ). Scheme 124 Example of 5-arylidene-4-thiazolidinones utilization in Michael reaction. Scheme 124 5 Conclusions Class of 5-ene-thiazolidinones is an illustrative example of the so-called privileged heterocycles in modern medicinal chemistry, possessing wide range of pharmacological activities. The major achievements in their study have been related to antitumor, hypoclycemic, antiviral and antimicrobial activities. Despite 5-ene-4-thiazolidinone derivatives are assigned as pan-assay interference compounds (with possible low affinity to several biotargets and not sufficient selectivity), their properties of Michael acceptors in vivo should be studied more precisely and objectively. The affinity toward different targets may be considered as an advantage and basis for further modification aiming to increase the selectivity. Another benefit of the described compounds is relative simplicity of their synthesis – sufficiently important argument for the needs of medicinal chemistry. A lot of data indicates saving the biological activity of 5-ene-4-thiazolidinones in the molecules synthesized on their basis. Thereby, polyfunctional 5-ene-4-thiazolidinones are useful starting building blocks for the synthesis of complex pharmacologically attractive heterocyclic systems. Given the data presented in the article, 5-ene-thiazolidinone scaffolds are the powerful tool in medicinal chemistry and undoubtedly should not be regarded as problematic compounds per se .

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8709489/

Bacteriophages as an Alternative Method for Control of Zoonotic and Foodborne Pathogens

The global increase in multidrug-resistant infections caused by various pathogens has raised concerns in human and veterinary medicine. This has renewed interest in the development of alternative methods to antibiotics, including the use of bacteriophages for controlling bacterial infections. The aim of this review is to present potential uses of bacteriophages as an alternative to antibiotics in the control of bacterial infections caused by multidrug-resistant bacteria posing a risk to humans, with particular emphasis on foodborne and zoonotic pathogens. A varied therapeutic and immunomodulatory (activation or suppression) effect of bacteriophages on humoral and cellular immune response mechanisms has been demonstrated. The antibiotic resistance crisis caused by global antimicrobial resistance among bacteria creates a compelling need for alternative safe and selectively effective antibacterial agents. Bacteriophages have many properties indicating their potential suitability as therapeutic and/or prophylactic agents. In many cases, bacteriophages can also be used in food quality control against microorganisms such as Salmonella , Escherichia coli , Listeria , Campylobacter and others. Future research will provide potential alternative solutions using bacteriophages to treat infections caused by multidrug-resistant bacteria. 1. Introduction Zoonotic pathogens cause problems all over the world, including diseases such as anthrax, brucellosis, bovine tuberculosis, hydatid disease, echinococcosis, trichinellosis, rabies, highly pathogenic avian influenza, Nipah/Hendra disease and bovine spongiform encephalopathy. In 2015, the WHO reported that more than 600 million people (1 in 10) worldwide fell ill as a result of foodborne infections [ 1 , 2 ]. According to the European Food Safety Authority (EFSA), the most frequently reported human foodborne diseases were campylobacteriosis and salmonellosis. However, the most dangerous pathogens for humans were identified as foodborne pathogenic bacteria found in livestock products, including enterohaemorrhagic Escherichia coli (EHEC; O157:H7), Shigella sp., Enterococcus spp. or Listeria spp. Multidrug-resistant pathogens isolated from human outbreaks, cattle, swine, and poultry were most often S. aureus , Streptococcus spp., Vibrio sp. and Yersinia spp. [ 3 , 4 ]. According to Niu et al. [ 5 ], these bacteria can also be transmitted to food products by direct contact with animals or indirectly by vectors such as insects, rodents, wild birds, or irrigation water. The global increase in multidrug-resistant infections and antibiotic failures in control of pathogens has raised concerns in human and veterinary medicine. An official report of the European Food Safety Authority (EFSA) regarding zoonotic and indicator bacteria isolated from humans, animals, and food showed that a high proportion (28.6%) of human Salmonella strains were resistant to three or more antimicrobials, and 34.9% of E. coli strains isolated from pigs were resistant to more than six antibiotics [ 6 ]. There has been a marked increase in the antibiotic resistance of Gram-negative bacteria via a variety of mechanisms, such as antibiotic target modification, antibiotic degradation, and modulation of permeability through the bacterial membrane. These mechanisms have limited the development of novel antibiotics. The most resistant strains of bacteria are carbapenem-resistant Enterobacteriaceae , extensively drug-resistant (XDR) Pseudomonas aeruginosa , and XDR Acinetobacter baumannii . Understanding the mechanisms of resistance of multidrug-resistant bacteria is the main goal in the development of modern antibacterial agents [ 7 ]. Global livestock production is faced with an alarming increase in bacterial resistance, including among zoonotic pathogens. For example, Donkor et al., [ 8 ] showed higher antimicrobial resistance in livestock than in humans, with animal E. coli isolates exhibiting high levels of resistance to tetracycline and penicillin. This has led to renewed interest among scientists to develop alternative methods to antibiotics, including the use of bacteriophages, since the beginning of the 21st century [ 9 ]. Widespread multidrug resistance among bacteria necessitates the search for alternative methods of controlling infections, including pre- and probiotics, vaccines, bacteriophages, nanoparticles, antimicrobial peptides (AMPs) and others. An example is the use of bacteriophages to reduce or eliminate pathogenic bacteria in livestock production, as biocontrol agents to control foodborne pathogens and to reduce contamination on food-contact surfaces [ 9 ]. An important contribution to research on the use of bacteriophages to control bacteria, including zoonotic pathogens, is the development and implementation of new legal regulations in the EU regarding restrictions or complete bans on the use of selected groups of chemotherapeutics in individual sectors of animal production. An example of such legislative action is the EU Council Directive 2019/6 [ 10 ] coming into force in January 2022. 1.1. General Characteristics of Bacteriophages Due to the widespread nature of bacteriophages (phages) associated with crops, live animals, and human intestinal environments, humans have direct and indirect contact with them. Many studies have demonstrated the common presence of bacteriophages in various fermented foods, such as yogurt and cheese. The application of specific bacteriophages to foods helps to reduce foodborne pathogenic bacteria [ 5 ]. Bacteriophages are bacterial viruses, causing complete lysis of a susceptible bacterial culture [ 11 ]. Interactions between phages and bacteria can be regarded as parasitism, as most virulent phage replication necessarily results in bacterial death. Certain interactions can be termed mutualistic, while some temperate phages encode benefits for the phenotypic properties of the host bacteria [ 12 ] According to Batinovic et al. [ 13 ], the prevalence of bacteriophages in the environment has been a natural phenomenon for billions of years, resulting in a balance of commensal and pathogenic bacteria. Phages and bacteria are the oldest and most ubiquitous microorganisms on Earth, likely having originated approximately 3 billion years ago [ 14 , 15 ]. Phages are prevalent in a variety of environments, including water, forest groundcover, food products, wastewater, and animal and human waste [ 16 ]. Bacteriophages have also been detected in commercial products, such as sera and human vaccines, as well as inside the human mouth (dental plaque and saliva) and in the gastrointestinal tracts of animals and humans [ 17 ]. Although bacteriophages may be present autonomously outside the host, all phages require the bacterial cell as a host for multiplication. Most phages are highly specific for host cell surface receptors such as receptor binding proteins (RBPs) or LPS [ 18 , 19 ]. 1.2. History of Bacteriophages Bacteriophages were first discovered more than 100 years ago by two microbiologists, Frederick Twort from England and the French Canadian Felix d'Herelle [ 20 , 21 ]. The first experimental and successful phage therapy was carried out by D'Herelle in the control of fowl typhoid in chickens (95–100% survival) [ 22 ]. He also coined the term 'bacteriophage', meaning 'bacteria eater'. Finally, in 1940, electron microscopes were used to identify the viral nature and morphology of phages [ 23 ]. Bacteriophages have been used in various types of therapies in humans, e.g., in dermatological, ophthalmological, urological, paediatric, otolaryngology and surgical infections. The significant therapeutic success of these treatments had a major impact on the development of phage therapy in the pre-antibiotic era. This was crucial, as the only treatment available in the first two decades of the 20th century was serum therapy (e.g., for pneumococci or the diphtheria bacterium), so bacteriophage therapy began to dominate in human medicine [ 24 ]. The discovery of the antimicrobial properties of Penicillium notatum in 1928 by Alexander Fleming culminated in the successful development of the first major antibiotic, penicillin, in 1941 [ 25 ], which marked the beginning of the antibiotic era and naturally inhibited the development of bacteriophage therapy. At present, as bacterial resistance to antibiotics is increasing significantly worldwide, phages are one of the factors with potential to replace them [ 26 ]. The best known bacteriophage centres in the world are the Eliava Institute of Bacteriophages, Microbiology, and Virology (EIBMV) of the Georgian Academy of Sciences, in Tbilisi, Georgia, and the Hirszfeld Institute of Immunology and Experimental Therapy (HIIET) of the Polish Academy of Sciences, in Wroclaw, Poland. Both institutes offer phage therapy against many bacterial and fungal pathogens, e.g., Staphylococcus spp., Klebsiella sp., Proteus sp., E. coli , and Pseudomonas sp., as well as other enteric pathogens [ 27 , 28 , 29 ]. 1.3. Classification of Bacteriophages Bacteriophages are the most widespread life forms on Earth. By 2018 year more than 650 strains of bacteriophages had been deposited in the American Type Culture Collection (ATCC) and >27,000 bacteriophage nucleotide sequences had been deposited in the International Nucleotide Sequence Database Collaboration (INSDC) [ 30 ]. The total number of these bacterial viruses has been estimated at 10 32 , which is 10 times the number of characterized bacteria. In water, the total count of bacteriophages has been estimated at 10 4 to 10 8 virions/mL −1 [ 31 ]. The classification of bacteriophages is based on the type of nucleic acid (ssRNA, dsRNA, ssDNA, dsDNA), the structure of the capsid (e.g., helical, pleomorphic, icosahedral, filamentous/thread-like, complex/polyhedral), which is built of structural proteins, and their life cycle, bacterial target, and site ( Figure 1 ). The phage taxonomy criteria applied by the International Committee on Taxonomy of Viruses (ICTV) were nucleic acid composition and virion morphology [ 9 ]. In 2015 the Bacterial and Archaeal Viruses Subcommittee (BAVS) classified phages into 873 species, 204 genera and 14 subfamilies [ 32 ]. The classification of bacteriophages is still ongoing, and in 2018 the ICTV presented a new classification of these bacterial viruses into 142 families, 81 subfamilies and 4978 species [ 9 ]. Most bacteriophages (96%) belong to the order Caudovirales , which is grouped into three main families: Myoviridae , Podoviridae and Siphoviridae [ 32 ]. Most bacteriophages contain double-stranded DNA, and the nucleic acid is coated with a protein capsid. Some phages have an additional layer (envelope) [ 17 ]. As new bacteriophages are continually being detected, their classification is constantly modified. The latest classification of viruses, based on the virus taxonomy proposed by the ICTV, was presented in our previous paper [ 33 ]. 1.4. Life Cycles of Phages The life cycle of phages is an important element of infections of bacteria. Phages can be categorized into types based on their virulence: lytic (virulent, productive) and lysogenic (temperate, dormant). Virulent phages follow a lytic cycle in the bacterial cell and lyse it to release a newly created population of phages [ 34 ]. The lytic cycle includes the adsorption, penetration, biosynthesis, assembly and release of bacteriophages from the infected bacterium. During this process many phages use specific proteins located on the surface of the bacterial cell as receptors. During the adsorption phase, the bacteriophage adheres to the bacterial cell, and phage proteins bind to specific receptors, such as teichoic and lipoteichoic acid for Gram-positive bacteria or LPS for Gram-negative bacteria [ 35 ]. The next phase, penetration, consists of destruction of the bacterial wall by bacteriophage enzymes and insertion of the genetic material into the bacterial cell. This is followed by the formation of capsid structures for nucleic acid and protein replication, accompanied by inhibition of replication of bacterial DNA. The phage genetic material is transcribed in the bacterial cell by RNA polymerases to produce mRNA, which supresses host intracellular synthesis as a consequence of bacteriophage multiplication [ 36 ]. Tens, hundreds or thousands of replicated phages are released by means of lysis of the bacterial cytoplasmic membrane by a phage protein (holin) and the formation of pores by endolysin encoded by double-stranded phage DNA and peptidoglycan hydrolases. The duration of the entire lytic cycle may be 20–40 min or up to 1–2 h [ 9 , 37 ]. Lysogenic infection via phages involves integration of their genetic material into the chromosome of the infected bacteria (prophage), which does not destroy the bacterial cell or produce a new population of bacteriophages. It leads to the integration of the phage genetic material with the bacterial DNA and its transmission into a new population of bacteria. This kind of bacteriophage is called a temperate phage, and in cells carrying a prophage it is referred to as lysogenic. Nevertheless, the viral prophage, also called an endogenous phage (a latent form of phage), can become activated by abnormal environmental conditions and other external stress factors that can damage the bacterial genetic material, including sunlight, UV radiation, some alkylating cytostatics (chlorambucil, cyclophosphamide, ifosfamide, estramustine or chlormethine) or mutagenic antibiotics such as mitomycin C. In some cases, the prophage is excised incorrectly from the chromosome, taking with it neighbouring bacterial genes. This is one of the main means of horizontal gene transfer (HGT) among bacteria, which is also one of the main methods in molecular biology [ 9 ]. Phages which have been defined as temporary include E. coli Lambda [ 38 ], with activity against E. coli and other Enterobacteriaceae ; phage Mu, specific for Salmonella , Citrobacter and Erwinia ; MM1 Streptococcus pneumoniae ; and φ11 S. aureus [ 39 ]. In another type of life cycle involving chronic infection, bacteriophages infect the bacterial cell, in which new phage populations arise without destroying the bacteria. The chronic infection lifestyle is found in rod-shaped (filamentous) single-stranded DNA phages and in plasmaviruses that infect mycoplasmas. In the chronic infection lifestyle, phages are gradually eliminated from the bacteria over a long period without destruction of the cell [ 40 ]. 1.1. General Characteristics of Bacteriophages Due to the widespread nature of bacteriophages (phages) associated with crops, live animals, and human intestinal environments, humans have direct and indirect contact with them. Many studies have demonstrated the common presence of bacteriophages in various fermented foods, such as yogurt and cheese. The application of specific bacteriophages to foods helps to reduce foodborne pathogenic bacteria [ 5 ]. Bacteriophages are bacterial viruses, causing complete lysis of a susceptible bacterial culture [ 11 ]. Interactions between phages and bacteria can be regarded as parasitism, as most virulent phage replication necessarily results in bacterial death. Certain interactions can be termed mutualistic, while some temperate phages encode benefits for the phenotypic properties of the host bacteria [ 12 ] According to Batinovic et al. [ 13 ], the prevalence of bacteriophages in the environment has been a natural phenomenon for billions of years, resulting in a balance of commensal and pathogenic bacteria. Phages and bacteria are the oldest and most ubiquitous microorganisms on Earth, likely having originated approximately 3 billion years ago [ 14 , 15 ]. Phages are prevalent in a variety of environments, including water, forest groundcover, food products, wastewater, and animal and human waste [ 16 ]. Bacteriophages have also been detected in commercial products, such as sera and human vaccines, as well as inside the human mouth (dental plaque and saliva) and in the gastrointestinal tracts of animals and humans [ 17 ]. Although bacteriophages may be present autonomously outside the host, all phages require the bacterial cell as a host for multiplication. Most phages are highly specific for host cell surface receptors such as receptor binding proteins (RBPs) or LPS [ 18 , 19 ]. 1.2. History of Bacteriophages Bacteriophages were first discovered more than 100 years ago by two microbiologists, Frederick Twort from England and the French Canadian Felix d'Herelle [ 20 , 21 ]. The first experimental and successful phage therapy was carried out by D'Herelle in the control of fowl typhoid in chickens (95–100% survival) [ 22 ]. He also coined the term 'bacteriophage', meaning 'bacteria eater'. Finally, in 1940, electron microscopes were used to identify the viral nature and morphology of phages [ 23 ]. Bacteriophages have been used in various types of therapies in humans, e.g., in dermatological, ophthalmological, urological, paediatric, otolaryngology and surgical infections. The significant therapeutic success of these treatments had a major impact on the development of phage therapy in the pre-antibiotic era. This was crucial, as the only treatment available in the first two decades of the 20th century was serum therapy (e.g., for pneumococci or the diphtheria bacterium), so bacteriophage therapy began to dominate in human medicine [ 24 ]. The discovery of the antimicrobial properties of Penicillium notatum in 1928 by Alexander Fleming culminated in the successful development of the first major antibiotic, penicillin, in 1941 [ 25 ], which marked the beginning of the antibiotic era and naturally inhibited the development of bacteriophage therapy. At present, as bacterial resistance to antibiotics is increasing significantly worldwide, phages are one of the factors with potential to replace them [ 26 ]. The best known bacteriophage centres in the world are the Eliava Institute of Bacteriophages, Microbiology, and Virology (EIBMV) of the Georgian Academy of Sciences, in Tbilisi, Georgia, and the Hirszfeld Institute of Immunology and Experimental Therapy (HIIET) of the Polish Academy of Sciences, in Wroclaw, Poland. Both institutes offer phage therapy against many bacterial and fungal pathogens, e.g., Staphylococcus spp., Klebsiella sp., Proteus sp., E. coli , and Pseudomonas sp., as well as other enteric pathogens [ 27 , 28 , 29 ]. 1.3. Classification of Bacteriophages Bacteriophages are the most widespread life forms on Earth. By 2018 year more than 650 strains of bacteriophages had been deposited in the American Type Culture Collection (ATCC) and >27,000 bacteriophage nucleotide sequences had been deposited in the International Nucleotide Sequence Database Collaboration (INSDC) [ 30 ]. The total number of these bacterial viruses has been estimated at 10 32 , which is 10 times the number of characterized bacteria. In water, the total count of bacteriophages has been estimated at 10 4 to 10 8 virions/mL −1 [ 31 ]. The classification of bacteriophages is based on the type of nucleic acid (ssRNA, dsRNA, ssDNA, dsDNA), the structure of the capsid (e.g., helical, pleomorphic, icosahedral, filamentous/thread-like, complex/polyhedral), which is built of structural proteins, and their life cycle, bacterial target, and site ( Figure 1 ). The phage taxonomy criteria applied by the International Committee on Taxonomy of Viruses (ICTV) were nucleic acid composition and virion morphology [ 9 ]. In 2015 the Bacterial and Archaeal Viruses Subcommittee (BAVS) classified phages into 873 species, 204 genera and 14 subfamilies [ 32 ]. The classification of bacteriophages is still ongoing, and in 2018 the ICTV presented a new classification of these bacterial viruses into 142 families, 81 subfamilies and 4978 species [ 9 ]. Most bacteriophages (96%) belong to the order Caudovirales , which is grouped into three main families: Myoviridae , Podoviridae and Siphoviridae [ 32 ]. Most bacteriophages contain double-stranded DNA, and the nucleic acid is coated with a protein capsid. Some phages have an additional layer (envelope) [ 17 ]. As new bacteriophages are continually being detected, their classification is constantly modified. The latest classification of viruses, based on the virus taxonomy proposed by the ICTV, was presented in our previous paper [ 33 ]. 1.4. Life Cycles of Phages The life cycle of phages is an important element of infections of bacteria. Phages can be categorized into types based on their virulence: lytic (virulent, productive) and lysogenic (temperate, dormant). Virulent phages follow a lytic cycle in the bacterial cell and lyse it to release a newly created population of phages [ 34 ]. The lytic cycle includes the adsorption, penetration, biosynthesis, assembly and release of bacteriophages from the infected bacterium. During this process many phages use specific proteins located on the surface of the bacterial cell as receptors. During the adsorption phase, the bacteriophage adheres to the bacterial cell, and phage proteins bind to specific receptors, such as teichoic and lipoteichoic acid for Gram-positive bacteria or LPS for Gram-negative bacteria [ 35 ]. The next phase, penetration, consists of destruction of the bacterial wall by bacteriophage enzymes and insertion of the genetic material into the bacterial cell. This is followed by the formation of capsid structures for nucleic acid and protein replication, accompanied by inhibition of replication of bacterial DNA. The phage genetic material is transcribed in the bacterial cell by RNA polymerases to produce mRNA, which supresses host intracellular synthesis as a consequence of bacteriophage multiplication [ 36 ]. Tens, hundreds or thousands of replicated phages are released by means of lysis of the bacterial cytoplasmic membrane by a phage protein (holin) and the formation of pores by endolysin encoded by double-stranded phage DNA and peptidoglycan hydrolases. The duration of the entire lytic cycle may be 20–40 min or up to 1–2 h [ 9 , 37 ]. Lysogenic infection via phages involves integration of their genetic material into the chromosome of the infected bacteria (prophage), which does not destroy the bacterial cell or produce a new population of bacteriophages. It leads to the integration of the phage genetic material with the bacterial DNA and its transmission into a new population of bacteria. This kind of bacteriophage is called a temperate phage, and in cells carrying a prophage it is referred to as lysogenic. Nevertheless, the viral prophage, also called an endogenous phage (a latent form of phage), can become activated by abnormal environmental conditions and other external stress factors that can damage the bacterial genetic material, including sunlight, UV radiation, some alkylating cytostatics (chlorambucil, cyclophosphamide, ifosfamide, estramustine or chlormethine) or mutagenic antibiotics such as mitomycin C. In some cases, the prophage is excised incorrectly from the chromosome, taking with it neighbouring bacterial genes. This is one of the main means of horizontal gene transfer (HGT) among bacteria, which is also one of the main methods in molecular biology [ 9 ]. Phages which have been defined as temporary include E. coli Lambda [ 38 ], with activity against E. coli and other Enterobacteriaceae ; phage Mu, specific for Salmonella , Citrobacter and Erwinia ; MM1 Streptococcus pneumoniae ; and φ11 S. aureus [ 39 ]. In another type of life cycle involving chronic infection, bacteriophages infect the bacterial cell, in which new phage populations arise without destroying the bacteria. The chronic infection lifestyle is found in rod-shaped (filamentous) single-stranded DNA phages and in plasmaviruses that infect mycoplasmas. In the chronic infection lifestyle, phages are gradually eliminated from the bacteria over a long period without destruction of the cell [ 40 ]. 2. The Spectrum of Use of Bacteriophages The specificity of phage activity means that they infect only the bacteria specific for them (called the host) via external receptors, which determines the phage host range. Therefore, the use of phage therapy relies on a detailed and accurate characterization of the bacteria, including pathotypes and serotypes. Bacteriophages can be used in a variety of forms and methods to control and eliminate bacteria, including therapy, food protection and sanitation procedures [ 1 , 9 ]. Examples of the scope of the use of bacteriophages are presented in Table 1 . Bacteriophages can potentially be used as biological control agents, especially in the reduction and elimination of bacterial contamination in foods, e.g., by Salmonella , Listeria monocytogenes , Campylobacter spp. or E. coli O157:H7 [ 15 , 50 ]. The high efficacy and safety of bacteriophage therapy is due in part to their specificity for selected bacteria: a single species, serotype, or strain. This is beneficial because the commensal gut microbiota is not destroyed. Another advantage is that, due to the self-replication of bacteriophages at the site of application, repetition of the application is often unnecessary. In many cases, no side effects of clinical treatment are observed, indicating a high level of safety that has been confirmed in many studies [ 51 ]. However, the application of bacteriophages in live animals or humans induces a cellular immune response, which could lead to the inactivation of phages, rendering them ineffective in eliminating bacteria [ 52 , 53 , 54 ]. In many experimental phage therapies a beneficial effect was observed as a significant reduction in bacterial content or elimination of the pathogens. Phages have been used to control Shiga-toxin-producing E. coli (ETEC) infections in newborn ruminants, including calves and lambs, or other livestock species, such as piglets [ 41 , 55 ]. They have been exploited to control bacterial infections in humans in many countries, including Poland, Georgia, Russia, France, Belgium, Switzerland and the USA [ 56 , 57 , 58 ]. Phage therapies have been applied against infections caused by numerous pathogens, especially multidrug-resistant bacteria, e.g., Acinetobacter , Burkholderia , Citrobacter , Enterobacter , Enterococcus , Escherichia coli , Klebsiella , Morganella , Proteus , Pseudomonas , Shigella spp., Shigella flexneri , Staphylococcus , Salmonella , Serratia and Stenotrophomonas . According to the Eliava Phage Therapy Centre, bacteriophage therapy against Enterococcus faecalis , E. coli (O11, O18, O20, O25, O26, O44, O55, O113, O125 and O128), Proteus vulgaris , Proteus mirabilis , Pseudomonas aeruginosa , Salmonella spp., and Shigella spp. showed positive results in 35–50% of human patients [ 59 , 60 ]. Bacteriophage Interactions during Phage Therapy Bacteriophages are regarded as the most applicable ecological and alternative means of elimination of pathogens (control and prevention of infections) due to their natural origin and numerous advantages, including the following: lysis of bacteria usually highly resistant to antibiotic therapy, living in a biofilm; high degree of safety for commensal and symbiotic flora; possibility of use with other bacteriophages as a cocktail or with other antibacterials; complete biodegradability of bacteriophages, making them safe for the organism and the environment [ 17 ]. However, phage therapy may carry a risk of immunological reactions, which is linked to the protein structures of bacteriophages. The immune response to bacteriophages depends on the location of the bacterial infection and the route of administration of the phages. The activity of bacteriophages also relies on their ability to penetrate epithelial cells and potentially spread to the bloodstream, lymph and internal organs such as the lung, liver, kidney and brain [ 61 ]. Bacteriophages can activate dendritic cells to synthesize pro-inflammatory factors (including IL-6, IL-1α, IL-1β and TNF-α) and to induce changes in the expression profile of these cell surface proteins and activation of the NF-κB signalling pathway [ 62 ]. The results of many studies confirm that bacteriophages can be phagocytosed by mammalian cells [ 63 ]. For example, Geier et al. [ 64 ] demonstrated rapid removal of wild-type phage λ from the circulatory system in humans. According to the authors, phagocytosis via immune cells is the main process of elimination of bacteriophages in mammals, and this mechanism takes place during lysis of bacteria by bacteriophages, which increases the activity of phagocytic cells, including PMN cells. The higher number of neutrophils at the site of infection is necessary to remove phage-resistant bacteria; this neutrophil-phage cooperation process has been confirmed in the resolution of P. aeruginosa infections [ 53 , 65 ]. However, some studies [ 66 ] have confirmed that bacteriophages can also express anti-inflammatory properties by decreasing the expression of TNFα and monocyte chemoattractant protein-1, which reduces ROS production by neutrophils and protects the epithelia against damage. Some bacteriophages can also be a natural component of the intestinal microbiota and consumed food [ 67 ]. The oral administration of phages against Staphylococcus , Klebsiella , Escherichia , Proteus and Pseudomonas also induces the production of antibodies [ 68 ]. There has been no evidence of immunological disorders following phage ingestion per os at any concentration [ 69 ]. Topical application of phages to animals and humans also caused no side effects [ 70 ]. Minor problems have been observed in the case of other internal organs and blood vessels, which are not natural environments for phages. Here the immunogenic and immunomodulatory effects of phages can be observed. Bacteriophages can have non-specific effects on the immunological functions of various immune cells, including PMNCs, as well as on cytokine production and the induction of specific antibodies against non-phage antigens [ 71 ]. For example, resident liver macrophages (also called Kupffer cells) are able to eliminate bacteriophages by phagocytosis four times faster than spleen macrophages. The natural innate immune response is usually sufficient to eliminate pathogens before the activation of adaptive immune mechanisms. Bacteriophages can activate immune mechanisms and thus affect the metabolic activity of immune cells. However, bacteriophages can inhibit the production and release of reactive oxygen species in response to pathogens, which could decrease innate antibacterial immunity. [ 72 ]. Phages can induce antibodies that neutralize them, which can inhibit the antibacterial effect of phages in the form of lysis of targeted bacteria [ 71 , 73 ]. It is not currently clear how long this type of antibody will remain in the body, as knowledge of the kinetic aspect of bacteriophage activity is insufficient. Moreover, the titre of these antibodies depends on many factors, including the route of application (local application causes a minor increase in antibodies) and its frequency [ 74 ]. Some information about the influence of bacteriophages on immune responses in animal's model has been presented in Table 2 . Antiphage antibodies are probably one of the most important factors influencing the efficacy of phage therapy. However, the activation of the production of neutralizing antibodies by phages need not be a problem during the initial phase of treatment of bacterial infections, because bacteriophage activity is much faster than the production of phage-neutralizing antibodies [ 27 ]. However, these antibodies can affect the efficacy of treatment during the second phase of therapy. This necessitates the implementation of additional solutions, such as the following: repeating phage administration two or more times, because bacteriophages can multiply at the site of application during infection of the host bacteria; increasing the phage concentration in the solution, because a high level of phages protects against complete destruction by neutralizing antibodies; using different phages, especially during the second and subsequent cycles of application during treatment, because resistance differs from one phage to another [ 27 ]. In addition to the increase in neutralizing antibodies during phage therapies, the concentration of class M and G immunoglobulins increases as well and continues to increase with subsequent applications of phage preparations [ 83 , 84 ]. Besides activating humoral response mechanisms, phages also play a significant role in the modulation of cellular immunity against them. For example, s.c. application of MS-2 phages induced a hypersensitivity reaction in guinea pigs [ 85 ]. It has been suggested that the cellular response plays only a minor role in phage inactivation, as observed in the case of phage T7 in T cell-deficient mice [ 79 ]. In another study [ 86 ], phages had an immunosuppressive effect by suppressing the activation of T lymphocytes during the development of transplantation tolerance. While bacteriophage therapies have been an effective tool in control of bacterial infections in various animal species, phages are also currently used for typing and diagnosis of specific bacterial species and for control of foodborne pathogens in food. Bacteriophage Interactions during Phage Therapy Bacteriophages are regarded as the most applicable ecological and alternative means of elimination of pathogens (control and prevention of infections) due to their natural origin and numerous advantages, including the following: lysis of bacteria usually highly resistant to antibiotic therapy, living in a biofilm; high degree of safety for commensal and symbiotic flora; possibility of use with other bacteriophages as a cocktail or with other antibacterials; complete biodegradability of bacteriophages, making them safe for the organism and the environment [ 17 ]. However, phage therapy may carry a risk of immunological reactions, which is linked to the protein structures of bacteriophages. The immune response to bacteriophages depends on the location of the bacterial infection and the route of administration of the phages. The activity of bacteriophages also relies on their ability to penetrate epithelial cells and potentially spread to the bloodstream, lymph and internal organs such as the lung, liver, kidney and brain [ 61 ]. Bacteriophages can activate dendritic cells to synthesize pro-inflammatory factors (including IL-6, IL-1α, IL-1β and TNF-α) and to induce changes in the expression profile of these cell surface proteins and activation of the NF-κB signalling pathway [ 62 ]. The results of many studies confirm that bacteriophages can be phagocytosed by mammalian cells [ 63 ]. For example, Geier et al. [ 64 ] demonstrated rapid removal of wild-type phage λ from the circulatory system in humans. According to the authors, phagocytosis via immune cells is the main process of elimination of bacteriophages in mammals, and this mechanism takes place during lysis of bacteria by bacteriophages, which increases the activity of phagocytic cells, including PMN cells. The higher number of neutrophils at the site of infection is necessary to remove phage-resistant bacteria; this neutrophil-phage cooperation process has been confirmed in the resolution of P. aeruginosa infections [ 53 , 65 ]. However, some studies [ 66 ] have confirmed that bacteriophages can also express anti-inflammatory properties by decreasing the expression of TNFα and monocyte chemoattractant protein-1, which reduces ROS production by neutrophils and protects the epithelia against damage. Some bacteriophages can also be a natural component of the intestinal microbiota and consumed food [ 67 ]. The oral administration of phages against Staphylococcus , Klebsiella , Escherichia , Proteus and Pseudomonas also induces the production of antibodies [ 68 ]. There has been no evidence of immunological disorders following phage ingestion per os at any concentration [ 69 ]. Topical application of phages to animals and humans also caused no side effects [ 70 ]. Minor problems have been observed in the case of other internal organs and blood vessels, which are not natural environments for phages. Here the immunogenic and immunomodulatory effects of phages can be observed. Bacteriophages can have non-specific effects on the immunological functions of various immune cells, including PMNCs, as well as on cytokine production and the induction of specific antibodies against non-phage antigens [ 71 ]. For example, resident liver macrophages (also called Kupffer cells) are able to eliminate bacteriophages by phagocytosis four times faster than spleen macrophages. The natural innate immune response is usually sufficient to eliminate pathogens before the activation of adaptive immune mechanisms. Bacteriophages can activate immune mechanisms and thus affect the metabolic activity of immune cells. However, bacteriophages can inhibit the production and release of reactive oxygen species in response to pathogens, which could decrease innate antibacterial immunity. [ 72 ]. Phages can induce antibodies that neutralize them, which can inhibit the antibacterial effect of phages in the form of lysis of targeted bacteria [ 71 , 73 ]. It is not currently clear how long this type of antibody will remain in the body, as knowledge of the kinetic aspect of bacteriophage activity is insufficient. Moreover, the titre of these antibodies depends on many factors, including the route of application (local application causes a minor increase in antibodies) and its frequency [ 74 ]. Some information about the influence of bacteriophages on immune responses in animal's model has been presented in Table 2 . Antiphage antibodies are probably one of the most important factors influencing the efficacy of phage therapy. However, the activation of the production of neutralizing antibodies by phages need not be a problem during the initial phase of treatment of bacterial infections, because bacteriophage activity is much faster than the production of phage-neutralizing antibodies [ 27 ]. However, these antibodies can affect the efficacy of treatment during the second phase of therapy. This necessitates the implementation of additional solutions, such as the following: repeating phage administration two or more times, because bacteriophages can multiply at the site of application during infection of the host bacteria; increasing the phage concentration in the solution, because a high level of phages protects against complete destruction by neutralizing antibodies; using different phages, especially during the second and subsequent cycles of application during treatment, because resistance differs from one phage to another [ 27 ]. In addition to the increase in neutralizing antibodies during phage therapies, the concentration of class M and G immunoglobulins increases as well and continues to increase with subsequent applications of phage preparations [ 83 , 84 ]. Besides activating humoral response mechanisms, phages also play a significant role in the modulation of cellular immunity against them. For example, s.c. application of MS-2 phages induced a hypersensitivity reaction in guinea pigs [ 85 ]. It has been suggested that the cellular response plays only a minor role in phage inactivation, as observed in the case of phage T7 in T cell-deficient mice [ 79 ]. In another study [ 86 ], phages had an immunosuppressive effect by suppressing the activation of T lymphocytes during the development of transplantation tolerance. While bacteriophage therapies have been an effective tool in control of bacterial infections in various animal species, phages are also currently used for typing and diagnosis of specific bacterial species and for control of foodborne pathogens in food. 3. Commercial Products with Bacteriophages for Elimination of Foodborne Zoonotic Pathogens Foodborne infections are the most important global health problem, contributing significantly to hospitalizations and deaths worldwide despite many advances in pathogen surveillance. Traditional food sanitation techniques using antimicrobial methods (including pasteurization, high pressure, irradiation or chemical treatment) can reduce pathogens in foods in varying degrees. However, these methods may damage equipment and adversely affect the organoleptic qualities (and possibly the nutritional value) of foods. The most important problem with using chemicals is that they eliminate 'good' microbes, which are beneficial in natural preservation of foods [ 87 ]. Therefore, it seems preferable to use an effective natural and ecological alternative such as bacteriophages for biocontrol of foodborne pathogens. Bacteriophages are mainly used in three sectors of the food industry to ensure food safety: primary production, biopreservation and biosanitization. As components of commercial products, they are currently finding application in the elimination of pathogens from food products of animal origin (meat products, milk and dairy products) or plant origin (fresh fruits and vegetables). The number of commercial bacteriophage products approved for use in food safety in various countries is continually increasing. Many commercial companies around the world have shown interest in information about the use of bacteriophages as antibacterial tools to control foodborne pathogens, e.g., in the United States (AmpliPhi Biosciences (VI, USA), Enbiotix (Boston, MA, USA); Intralytix), the United Kingdom (Novolytics, Sarum Biosciences and Fixed Phage, Bolton, UK), EU countries such as France (Pherecydes Pharma, Romainville, Ile-de-France, France) and Portugal (Technophage and InnoPhage, Lisbon, Portugal), and other countries [ 88 ]. Detailed information about commercial bacteriophage products used for biocontrol of foodborne pathogens in food is presented in Table 3 . 4. Advantages and Disadvantages of Bacteriophage Therapy Phages have several advantages over antibiotics as therapeutic agents, such as activity against all types of bacteria, including MDR-pathogens. Their narrow antibacterial spectrum (which protects the natural microbiome), the low level of side effects, and their extensive distribution when administered systemically are also worth noting. They also may exert an effect on the inflammatory response, and their low production cost and high efficacy are significant benefits [ 31 , 99 ]. Many studies have confirmed the beneficial effects of the use of bacteriophages, shown as follows: Bacteriophages show high specificity for their target pathogens and kill only pathogens without destroying the physiological saprophytic flora; the narrow host range of phages is also a useful feature in prophylaxis of infections caused by enteric bacteria [ 100 ]. The distribution of phages in the body following systemic administration is much more extensive than in the case of antibiotics, in part due to the lack of or very low level of resistance of bacteria [ 31 ]. Because the mechanism of action of phages against the host bacteria is different to that of antibiotics, they are highly effective against many pathogens, especially against multidrug-resistant bacteria [ 36 ]. Phages replicate at the site of infection even after a single application, because they multiply inside the bacterial cell and therefore are released at the site of infection [ 101 ]. Bacteriophages are resistant to stress factors during food production [ 91 ]. Phage therapy is theoretically cheaper than antibiotic therapy due to the simplicity of production [ 99 ]. The unit costs of production as well as the costs of isolation and characterization are comparable or even lower than the costs of classical chemotherapeutic products [ 102 ]. There is no withdrawal period in livestock due to the lack of residue in tissues as soon as therapy is completed [ 103 ]. There are no side effects or allergic reactions because most bacteriophages consist mainly of proteins and genetic material (DNA or RNA). However, in addition to the positive effects of phage therapies, widespread use of bacteriophages is limited by obstacles such as the following: Due to their high specificity for a single type of bacteria, bacteriophages have a narrow host range [ 104 ]. Bacteriophages may neutralize antibodies, which may prevent a portion of the administered phage dose from adhering to the target bacteria [ 104 ]. Bacteriophages have poor stability in the environment, e.g., sunlight, UV, low pH 50°C [ 17 , 105 ]. Only lytic phages are admissible in phage therapy because lysogenic (temporary) phages may be a source of horizontal transfer of bacterial toxins or antibiotic resistance [ 31 ]. The duration of survival of phages is varied, depending in part on the presence of the host bacteria. Their activity is also influenced by the environment within the organism in which it is administered, and therefore the survival of phages must be monitored at the site of administration in order to assess their antimicrobial activity [ 99 ]. Information about the kinetics of phages remains insufficient, especially the degree of adsorption, the number of replications necessary for a therapeutic effect, the latent period, and their elimination from the body by phagocytic cells [ 106 ]. 5. Bacteriophage Efficacy in Experimental Models There are many methods of application of phages in therapies for humans and animals, including intraperitoneal, subcutaneous or intramuscular injection or oral, intragastric, rectal, topical or intranasal administration. Forms of administration of phages during therapy include sprays, aerosols, lozenges, compresses, mouthwash, suppositories, throat rinses, bandages, eye or ear drops and tampons [ 107 ]. In many studies in humans and animals, the form of application and type of administration has been associated with the type and location of the disease. In earlier studies [ 108 , 109 , 110 , 111 ], the best therapeutic effect was observed after direct application of phages to the target bacteria, as in the case of bacterial dysentery caused by Shigella ; intestinal dysbacteriosis caused by E. coli and Proteus spp.; lung and pleural infections caused by Staphylococcus ; suppurative skin infections caused by Pseudomonas , Staphylococcus , Klebsiella , Proteus , and E. coli ; and infections of the skin or nasal mucosa caused by Klebsiella spp. Some studies have explored the use of phages for control and treatment of neonatal enterotoxigenic E. coli infections in cattle, poultry and pigs [ 40 , 112 ]. Bacteriophages have also been used in controlling systemic infections with foodborne pathogens, including Salmonella spp., E. coli , Campylobacter spp., Vibro spp., Pseudomonas aeruginosa , and other pathogens, such as Staphylococcus spp., Streptococcus spp., Klebsiella spp., Acinetobacter spp., and even Mycobacterium spp. These experiments were carried out in experimental mouse or rat models, as well as in chickens, rabbits, calves, pigs and sheep. Examples of the effects of experimental phage therapies in different animal species and in the control of various pathogens are presented in Table 4 . 6. Conclusions To sum up, bacteriophages have many properties indicating their potential suitability as therapeutic or/and prophylactic agents. Future research on the scope of phages will provide a good picture of their potential to treat infections caused by multidrug-resistant bacteria. However, as bacteriophages are essentially 'living' drugs, the study of their use for therapy or biocontrol spans from purely clinical observations to molecular analysis to considerations of immunology and ecology. Due to the antibiotic resistance crisis, there is a compelling need for alternative safe and selectively effective antibacterial agents.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10389281/

Effects of Molecular Size on Resolution in Charge Detection Mass Spectrometry

Instrumental resolution of FT-CDMS instruments with electrostatic ion trap detection of individual ions depends on the precision with which ion energy is determined. Energy can be selected using ion optic filters or from harmonic amplitude ratios (HAR) that provide Fellgett's advantage and eliminate the necessity of ion transmission loss to improve resolution. Unlike the ion energy filtering method, the resolution of the HAR method increases with charge (improved S/N) and thus with mass. An analysis of the HAR method with current instrumentation indicates that higher resolution can be obtained with the HAR method than the best resolution demonstrated for instruments with energy selective optics for ions in the low MDa range and above. However, this gain is typically unrealized because the resolution obtainable with molecular systems in this mass range are limited by sample heterogeneity. This phenomenon is illustrated with both TMV (0.6 – 2.7 MDa) and AAV9 (3.7 – 4.7 MDa) samples where mass spectral resolution is limited by the sample, including salt adducts, and not by instrument resolution. Nevertheless, the ratio of full to empty AAV9 capsids and the included genome mass can be accurately obtained in a few minutes from 1x PBS buffer solution and an elution buffer containing 300+ mM non-volatile content despite extensive adduction and lower resolution. Empty and full capsids adduct similarly indicating that salts encrust the complexes during late stages of droplet evaporation and that mass shifts can be calibrated in order to obtain accurate analyte masses even from highly salty solutions.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC209510/

Emerging ethical issues in instructions to authors of high-impact biomedical journals

Public interest in issues concerning the maintenance of high ethical standards in the conduct of scientific research and its publication has been increasing. Some of the developments in these issues as reflected in the publication of the medical literature are traced here. This paper attempts to determine whether public interest is reflected in the specific requirements for authors for manuscript preparation as stated in the "Instructions to Authors" for articles being prepared for submission to 124 "high- impact" journals. The instructions to authors of these journals were read on the Web for references to ethical standards or requirements. The ethical issues that the instructions most often covered were specifically related to the individual journal's publication requirements. The results suggest that while the editors and publishers of the biomedical literature are concerned with promoting and protecting the rights of the subjects of the experiments in the articles they publish, and while these concerns are not yet paramount, they are evolving and growing.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8999828/

Preventive Measures against Pandemics from the Beginning of Civilization to Nowadays—How Everything Has Remained the Same over the Millennia

As of 27 March 2022, the β-coronavirus severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has infected more than 487 million individuals worldwide, causing more than 6.14 million deaths. SARS-CoV-2 spreads through close contact, causing the coronavirus disease 2019 (COVID-19); thus, emergency lockdowns have been implemented worldwide to avoid its spread. COVID-19 is not the first infectious disease that humankind has had to face during its history. Indeed, humans have recurrently been threatened by several emerging pathogens that killed a substantial fraction of the population. Historical sources document that as early as between the 10th and the 6th centuries BCE, the authorities prescribed physical–social isolation, physical distancing, and quarantine of the infected subjects until the end of the disease, measures that strongly resemble containment measures taken nowadays. In this review, we show a historical and literary overview of different epidemic diseases and how the recommendations in the pre-vaccine era were, and still are, effective in containing the contagion. 1. Introduction A novel single-stranded RNA virus from the β-Coronaviridae family named severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), initially recognized as the viral cause of a cluster of pneumonia "originally unknown", emerged firstly in China, spread rapidly worldwide, and led to a global pandemic called "coronavirus disease 2019" (COVID-19). β-Coronaviruses are members of a large family of RNA viruses. One-third of the RNA sequence encodes four core structural proteins: spike (S), membrane-associated envelope (E), membrane (M), and nucleocapsid (N) proteins [ 1 ]. COVID-19 is characterized by severe upper respiratory tract infections, respiratory distress, and frequent need for hospitalization. Since the outbreak, COVID-19 has killed more than 6.12 million patients worldwide as of March 27th 2022. The clinical spectrum of COVID-19 ranges from asymptomatic or pauci-symptomatic to severe symptomatic with a large range of clinical manifestations such as cough, fever, myalgia, gastrointestinal symptoms, and anosmia. Admission to intensive care units (ICUs), mechanical ventilation, and mortality are common, especially in subjects with compromised immune systems, in subjects affected by underlying chronic diseases, or in the elderly [ 2 ]. The rapid deterioration of clinical conditions has been described as a typical feature of SARS-CoV-2 infection, driven by the pathogen–host interaction. The angiotensin-converting enzyme 2 (ACE2) is the only confirmed SARS-CoV-2 entry receptor [ 3 ]. SARS-CoV-2 and ACE2 binding allows the virus to enter a cell through the spike (S) proteins that work in concert with the host cell serine protease transmembrane protease serine 2 (TMPRSS2). TMPRSS2 cleaves the spike protein of SARS-CoV-2, facilitating membrane fusion. ACE2 relies on the renin–angiotensin system (RAS) molecular pathway; ACE2 is a crucial counter-regulatory enzyme to ACE by the breakdown of angiotensin II that is involved in blood pressure regulation and electrolyte homeostasis [ 4 ]. ACE2 is expressed in multiple tissues, and thus the existence of multiorgan complications/failure is not surprising [ 5 ]. Recently, clinicians started to report prolonged sequelae of symptoms in the post-acute phase of COVID-19, which may be represented in three categories: (1) Residual symptoms continuing after acute infection recovery; (2) Organ dysfunction continuing after initial recovery; (3) New symptoms/syndromes developing after initial asymptomatic or mild infection [ 6 ]. As a consequence of the long duration of pandemics and the large number of patients that survived the disease, the priority for National Health Systems is progressively shifting towards mid- and long-term effects of COVID-19. Frail and prefrail subjects who recovered after hospital admission represent an excellent model for studying this new scenario. With the COVID-19 pandemic, consideration of frailty hallmarks in dismissing infected older patients is critical to identify those subjects at higher risk of functional decline that may fall into accelerated aging processes. In this contest, there is the vital role of rehabilitation programs for the coming years and the urgent need to develop strategies to assist COVID-19 survivors. Communicable Diseases Communicable diseases, also known as transmissible diseases or infectious diseases, are illnesses that spread, directly or indirectly, from one organism to another through the transfer of a pathogen such as viruses, bacteria, fungi, protozoa, multicellular parasites, and aberrant proteins known as prions. They have been present for all of human history but became important with the rising of agrarian life 10,000 years ago when the demographic expansion and technological innovations made the transmission of infectious diseases more possible. The zoonotic transmission of pathogens from animals to humans is the key mechanism in the process of emerging infections, and the probability of cross-species transmission is enhanced by increasing interactions between humans and animals [ 7 ]. According to Wolfe et al. [ 7 ], the passage of a pathogen from a different species to humans is characterized by five steps: (1) The pathogen infects only animals under natural conditions; (2) The pathogen evolves to be transmitted to humans without continuous human-to-human transmission; (3) The pathogen undergoes secondary transmission to humans; (4) The disease exists in animals but different secondary human-to-human transmission occurs without the involvement of animal hosts; (5) The disease occurs exclusively in humans. The risk of zoonotic transmission is related to the animal species harboring the pathogen, the nature of human–animal interaction, and the frequency of these interactions. One of the first pieces of evidence of the insurgence of an epidemic is derived from molecular studies that unambiguously identified the presence of DNA from Yersinia pestis ( Y. pestis ), the causative agent of plague, in different individuals at Fralsegarden in Gokhem parish, Falbygden, western Sweden, dated to 4900 BCE [ 8 ]. Interestingly, the analysis of both the archaeological context and the human genomes supports the notion that the rise and the spread of the above plague were related to lifestyle (high population densities in close contact with animals), population growth, and expanding trade networks. It has been proposed that the Y. pestis plague may have contributed to the Neolithic decline [ 8 ]. Communicable Diseases Communicable diseases, also known as transmissible diseases or infectious diseases, are illnesses that spread, directly or indirectly, from one organism to another through the transfer of a pathogen such as viruses, bacteria, fungi, protozoa, multicellular parasites, and aberrant proteins known as prions. They have been present for all of human history but became important with the rising of agrarian life 10,000 years ago when the demographic expansion and technological innovations made the transmission of infectious diseases more possible. The zoonotic transmission of pathogens from animals to humans is the key mechanism in the process of emerging infections, and the probability of cross-species transmission is enhanced by increasing interactions between humans and animals [ 7 ]. According to Wolfe et al. [ 7 ], the passage of a pathogen from a different species to humans is characterized by five steps: (1) The pathogen infects only animals under natural conditions; (2) The pathogen evolves to be transmitted to humans without continuous human-to-human transmission; (3) The pathogen undergoes secondary transmission to humans; (4) The disease exists in animals but different secondary human-to-human transmission occurs without the involvement of animal hosts; (5) The disease occurs exclusively in humans. The risk of zoonotic transmission is related to the animal species harboring the pathogen, the nature of human–animal interaction, and the frequency of these interactions. One of the first pieces of evidence of the insurgence of an epidemic is derived from molecular studies that unambiguously identified the presence of DNA from Yersinia pestis ( Y. pestis ), the causative agent of plague, in different individuals at Fralsegarden in Gokhem parish, Falbygden, western Sweden, dated to 4900 BCE [ 8 ]. Interestingly, the analysis of both the archaeological context and the human genomes supports the notion that the rise and the spread of the above plague were related to lifestyle (high population densities in close contact with animals), population growth, and expanding trade networks. It has been proposed that the Y. pestis plague may have contributed to the Neolithic decline [ 8 ]. 2. History of "Plagues" The written texts and artworks of past times are useful as they provide information about the cultural contexts and enhance the understanding of the spread of communicable diseases in the past, centuries before the modern scientific literature. Literary plagues involve different forms, from true-to-life to completely imaginary. The term "plague" derives from the Latin plaga, meaning "blow or wound" [ 9 ]. Although plague refers to infection by an extremely contagious Gram-negative bacterium, Y. pestis (formerly called Pasteurella pestis ), the word "plague" is used not only as a generic term for an epidemic or pandemic but also as a metaphor for a wide range of calamities. Interestingly, the first book of Western literature, Homer's Iliad, starts with the story of a plague that strikes the Greek army at Troy ( Table 1 ) [ 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 26 , 27 , 28 ]. It is important to remember that no medicine or doctor had been able to cure or prevent infectious diseases for millennia, and the only way to stay safe was to avoid contact with infected people and contaminated objects. The great terror of epidemics was also fueled by the belief in their supernatural origin, as they were often believed to be caused by irate gods. In the Bible (i.e., Exodus (Hebrew: יציאת מצרים) 9:14, Numbers (Hebrew: בְּמִדְבַּר) 11:33, Former Prophets 1 Samuel (Hebrew: × ×‘×™××™× ×¨××©×•× ×™×, שְׁמוּאֵל]) 4:8, Psalms (Hebrew: תְּהִלִּים) 89:23, the Latter Prophets Isaiah (Hebrew: × ×‘×™××™× ××—×¨×•× ×™×; ספר ישעיהו) 9:13), the plague was regarded as a divine punishment for sins, and hence the frightening description of its diffusion was a warning to behave morally (reviewed also by Freemon [ 29 ]). This supposed causal relationship between plague and sin is also present in Greek literary texts, such as Homer's Iliad and Sophocles' King Oedipus (429 BCE) ( Table 1 ). This attitude changed completely with Thucydides (460–395 BCE), a Greek historian, who in his History of the Peloponnesian War, refused a supernatural origin of the disease (the Plague of Athens), describing it with a scientific approach, reporting the origin of the plague, the symptoms, the incapacity of doctors to cure it, and the uncontrolled fear of contagion among the public ( Table 1 ). The plague originated in early May 430 BCE, with a second wave in the summer of 428 BCE and a third in the winter of 427–426 BCE, and spared no segment of the population, including the statesman Pericles and Thucydides himself. The Plague of Athens killed between 75,000 and 100,000 people, ~25% of the population of Athens (429 BCE) ( Table 1 and Table 2 and Figure 1 ) [ 16 , 30 ]. The Roman historian Titus Livius, known as Livy in English ( Table 1 ), writing many centuries later, reported that epidemics occurred also in Rome in 433 and 438 BCE, suggesting the same origin of the disease. DNA analysis of skeletal remains recovered in 2001 in a mass burial pit in Greece and dating back to the plague years revealed ancient microbial typhoid ( Salmonella enterica serovar Typhi ), supporting the hypothesis that the Plague of Athens was a typhoid disease [ 32 ]. Figure 1 reports the emergence of the most important pandemics in the world until today with a possible estimation of the number of casualties, whereas Figure 2 reports the origin of the pandemics, considering their onset for the first time. Centuries later, between 165 and 180 CE, the Roman empire was challenged by the Antonine Plague. It is supposed that the disease was brought by armies from what is currently Iraq and eventually spread into the entire Roman empire, which comprised Central and Southern Europe. It is estimated that the diseases killed about 5 million people [ 20 ]. The symptoms were described by Galen and suggested that it was probably smallpox, as its presentation was characterized by rashes, hemorrhagic pustules, bloody diarrhea, fever, and sometimes hemoptysis [ 33 ]. The Justinian Plague (starting in 541 CE with subsequent outbreaks until 750–1000 CE) is the first confirmed pandemic plague caused by Y. pestis based on the analysis of the teeth of people buried at that time [ 22 ]. Accordingly, the reported symptoms were those distinctive of Y. pestis infection: fever, cough, and dyspnea in pneumonic plague and groin or axillary buboes in bubonic plague. The transmission was mediated from infected rats to humans through flea bites, but there was also human–human transmission. It has been estimated that the Plague of Justinian killed 60% of people in the Mediterranean area [ 34 ]. The Justinian Plague may have contributed to the end of the Roman empire, characterizing the transition from the Classical to the Medieval era [ 22 ]. It has been shown that the strain of Y. pestis linked with the Plague of Justinian is different from those associated with later human plague pandemics, and it seems that this strain is extinct or poorly present in wild rodent reservoirs [ 22 ]. The first scientific description of the etiology of plague and of the possible way of spreading of the disease did not become available until the 19th century. In 1894, in Hong Kong, Alexander Yersin (1863–1943) isolated in culture and identified the causative bacterial agent (i.e., Yersinia pestis ) [ 35 ]. After examination of the lymph glands of dead rats, he found the same bacteria described previously in human tissues; consequently, he made the causal connection between rat mortality and human epidemics. Surprisingly, the earliest recorded role of mice in plague is in the Bible that ascribed the pestilence among the Philistines to "the mice that marred the land" (1 שְׁמוּאֵל Samuel. 6:4–18). For the next centuries, after the different outbreaks of Justinian's plague, different epidemics occurred frequently. The next great pandemic plague was the dreaded "Black Death" of Europe in the 14th century ( Figure 1 and Figure 2 and Table 1 ). The Black Death, also known as the Black Plague, spread from the Caspian Sea by the town of Caffa, now Feodosia, a Genoese colony, to almost all European countries, causing the death of one-third of the European population [ 36 ] over the next few years, and persisted in Europe until 1750. Following traditional beliefs promoted by Hippocrates and Galen, the poisoned air ("miasma") was believed to be the causative agent of the epidemics; as already mentioned, an association between the plague and dead rats was not observed until the beginning of the 20th century [ 30 ]. The name Black Death derives from a typical symptom of the disease, called acral necrosis, of black color due to subdermal hemorrhages. Although there was a great debate on the etiology of this disease, DNA and protein signatures specific to Y. pestis were found in human skeletons from mass graves associated archaeologically with the Black Death throughout Europe, confirming it as the causative agent. Furthermore, the analysis of 17 single nucleotide polymorphisms (SNPs) and the absence of a deletion in the glpD gene (aerobic glycerol-3-phosphate dehydrogenase) identified two previously unknown but related clades of Y. pestis associated with distinct medieval mass graves, suggesting that the plague arrived in Europe at different times, through distinct routes [ 37 ]. The third wave of Y. pestis started in 1772 in Yunnan Province, Southwest China. In this case, Europe was not significantly affected, but some cases were reported in Malta in June 1945 [ 35 , 38 ]. A great nightmare and terror worldwide was represented by smallpox, which was a highly contagious and lethal disease with remarkably high death rates until its global eradication, declared by the WHO on 8 May 1980. The causative agent of smallpox is the variola virus (VARV), a member of the genus Orthopoxvirus , family Poxviridae , and subfamily Chordopoxvirina . Analysis of viruses' DNA collected from different individuals living both in Eurasia and in the Americas between 200 and 150 years ago revealed that their DNA resembled modern variola. The young pharaoh Rameses V in the 12th century BCE was probably killed by VARV [ 39 ]. The etiological cause of the Antonine Plague, which killed the emperor himself, was probably VARV. Queen Elizabeth I of England, at the age of 29 years in October 1562, survived smallpox, which left her without hair and with permanent disfiguring facial scars [ 40 ]. In 1694, VARV killed Queen Mary of England; six years later, it killed her son, the Duke of Gloucestershire, and then it went on to kill Emperor Joseph I of Austria (17 April 1711), King Luis I of Spain (31 August 1724), Tsar Peter II of Russia (30 January 1730), Queen Ulrika Eleonora of Sweden (24 November 1741), and the French King Louis XV (10 May 1774) [ 41 ]. During the American Revolutionary War (1775–1783), one of the greatest threats to the army was smallpox. In 1775, General George Washington recognized smallpox as a very serious problem for his army, especially after the smallpox outbreak in the city of Boston in the winter of 1775 [ 42 ]. Table 2 shows the scientific aspect of the history of infectious diseases [ 43 , 44 , 45 , 46 , 47 , 48 , 49 , 50 , 51 , 52 , 53 , 54 , 55 , 56 , 57 , 58 , 59 , 60 , 61 , 62 , 63 , 64 , 65 , 66 , 67 , 68 , 69 , 70 , 71 , 72 , 73 , 74 , 75 , 76 , 77 , 78 , 79 , 80 ], and Table 3 shows infectious diseases studied using molecular techniques, including metagenomics. Different reviews as well as two important books explore this field [ 81 , 82 , 83 , 84 , 85 , 86 ]. Moreover, Spyrou et al. [ 87 ] reviewed the ancient pathogen genomic data recovered from archaeological or historical specimens, including the method of retrieval. DNA analysis from archaeological sites has been a powerful tool for deciphering the steps of the evolution of pathogens [ 81 ]. Combining the data obtained from molecular analysis with disease modeling and the genetic history of the human population, together with the information offered by the archaeological, historical, and palaeopathological records, is helpful to define a comprehensive picture of host–pathogen interactions during centuries. Artistic and literary works, on the other hand, are helpful in understanding what kind of measures were taken to deal with pandemic emergencies in ancient times. jcm-11-01960-t002_Table 2 Table 2 Scientific milestones in infectious disease discovery. Years Discovery References 1530 Girolamo Fracastoro expresses his ideas on the origin of syphilis, explaining that this disease is spread by "seeds" distributed by intimate contact [ 43 ] 1683 Anton van Leeuwenhoek observes bacteria under the first microscope [ 44 ] 1701–1714 Giacomo Pilarino and Emmanuel Timoni give the first smallpox inoculations [ 45 ] 1757 Francis Home demonstrates that measles is caused by an infectious agent in the blood of patients [ 46 ] 1796 Edward Jenner develops the process of vaccination for smallpox, the first vaccine for any disease [ 45 ] 1842–1847 Oliver Wendel Holmes describes puerperal fever and Ignaz Semmelweis discovers how to prevent the transmission of puerperal fever [ 44 ] 1857 Louis Pasteur identifies germs as a cause of disease [ 47 ] 1870 Robert Koch and Louis Pasteur establish the germ theory of disease [ 48 ] 1879 First vaccine developed for chicken cholera by Louis Pasteur [ 49 ] 1881 First vaccine developed for anthrax by Louis Pasteur [ 49 ] 1882 First vaccine developed for rabies by Louis Pasteur [ 49 ] 1882 Koch discovers the Tuberculosis bacillus: Mycobacterium tuberculosis [ 50 ] 1890 Emil von Behring discovers antitoxins and develops tetanus and diphtheria vaccines [ 49 ] 1892 Dmitri Ivanovsky shows that sap from a diseased tobacco plant remained infectious to healthy tobacco plants despite having been filtered [ 51 ] 1894 Isolation in culture and microscopic description of causative bacteria [ 52 , 53 ] 1896/1897 Almroth Wright and Richard Pfeiffer develop the first vaccine for typhoid fever [ 54 ] 1897 Waldemar Haffkine tests on himself the first vaccine developed for bubonic plague [ 55 ] 1897 Paul Ehrlich develops a standardized unit of measure for diphtheria antitoxin that would play an important role in future developmental work on sera and vaccines [ 56 ] 1898 Martinus Beijerinck is convinced that filtrate contains a new form of infectious agent called a virus [ 57 ] 1913 Paul Ehrlich develops the first antimicrobial drug, "Salvarsan", against the bacterium Treponema pallidum , the etiological agent of syphilis [ 58 ] 1918 Charles Nicolle and Charles Lebailly advance the hypothesis that the causative agent of the "Spanish" flu is a nonfilterable agent of infinitesimal dimensions: possibly a virus [ 59 , 60 ] 1923 First vaccine developed for diphtheria by Alexander Thomas Glenny [ 49 ] 1924 First vaccine developed for tetanus (tetanus toxoid) by Alexander Thomas Glenny [ 49 ] 1914/1926 First vaccine developed for whooping cough (pertussis) by Leila Denmark [ 49 ] 1927 First vaccine developed for tuberculosis by Albert Calmette and Camille Guérin [ 49 ] 1928 Sir Alexander Fleming discovers penicillin [ 61 ] 1931 German engineers Ernst Ruska and Max Knoll project the electron microscope [ 62 ] 1931 Live attenuated bacterial vaccine developed and tested [ 63 , 64 ] 1932 Gerhard Domagk announces that the red dye prontosil is active against streptococcal infections in humans; afterward, Ernest Fourneau, Jacques and Thérèse Tréfouël, Daniel Bovet, and Fedrico Nitti show that the active antibacterial agent is sulfanilamide [ 65 ] 1935 First vaccine developed for yellow fever by Max Theiler [ 49 ] 1937 First vaccine developed for typhus by Rudolf Weigl [ 49 ] 1938 Jonas Salk and Thomas Francis develop the first vaccine against flu viruses [ 66 ] 1940–1947 Large concentrations of blood bacteria correlated with mortality [ 67 ] 1944 Oswald Avery, Colin MacLeod, and Maclyn McCarty report that DNA is the transforming factor in the experiments of Frederick Griffith where an extract of the pathogenic strain of pneumococcus could transform a harmless strain into a pathogenic one [ 68 ] 1944 Selman Waksman, Albert Schatz, and Elizabeth Bugie announce the discovery of streptomycin and state that it is active against Mycobacterium tuberculosis [ 69 ] 1949 John Franklin Enders, Thomas Weller, and Frederick Robbins grow poliovirus for the first time in cultured human embryo cells [ 70 ] 1953 James Watson and Francis Crick describe the structure of DNA [ 71 ] 1955 Jonas Salk develops the first polio vaccine [ 49 ] 1964 First vaccine developed for measles by John Franklin Enders [ 49 ] 1967 Maurice Hilleman develops the first vaccine for mumps virus [ 49 ] 1970 Maurice Hilleman develops the first vaccine for rubella [ 49 ] 1977 First approved vaccine developed for pneumonia [ 49 ] 1978 First approved vaccine developed for meningitis [ 49 ] 1980 Genetic relatedness of Yersinia pestis and Yersinia pseudotuberculosis [ 72 ] 1981 First approved vaccine developed for hepatitis B [ 49 ] 1983 HIV, the virus that causes AIDS, is identified [ 73 ] 1987 First approval of a drug against AIDS: zidovudine (AZT, ZDV) [ 74 ] 1992 First approved vaccine developed for hepatitis A [ 49 ] 1995 First vaccine developed for varicella (chickenpox) and hepatitis A [ 49 ] 1998 First approved vaccine developed for rotavirus [ 49 ] 2001–2011 DNA sequence of Yersinia pestis samples [ 75 ] 2006 First vaccine developed for human papillomavirus and for herpes zoster (shingles) [ 49 ] 2019 Food and Drug Administration (FDA) approves rVSVΔG-ZEBOV-GP Ebola vaccine [ 76 ] 2020 First vaccine against COVID-19 [ 77 ] 2022 First drugs for COVID-19 (molnupiravir and Paxlovid or nirmatrelvir + Paxlovid) [ 78 , 79 , 80 ] In bold the most important milestones in science discovery. In modern history, the most devastating pandemic was the 1918–1919 influenza called the "Spanish" flu (or the "Spanish Lady") ( Table 1 ). The flu was named "Spanish" since the infection spread in 1918 when the majority of nations were involved in World War I and the newspapers were under censorship rules and only the Spanish reported the pandemic. The first wave of flu appeared in March 1918, at Camp Funston in Kansas (USA), a military training camp. The second, more virulent wave appeared in August. In Europe, it appeared firstly at the seaport of Brest, where the American troops arrived. The confirmed number of casualties is unknown, with an estimated number ranging from 20 to 150 million [ 101 ]. The etiological agent of Spanish flu is named H1N1 for the hemagglutinin (H) and neuraminidase (N) proteins. The H1N1 influenza virus is an orthomyxovirus . The pandemic was characterized by an uncommonly high mortality rate among healthy young adults (from 15 to 34 years), not observed in either prior or subsequent influenza A epidemics. The case fatality rate was >2.5%, compared to 97%) showed a typical, self-limited course of influenza, without any antivirals, antibiotics, or vaccines. Public health officials imposed multiple interventions as disease containment measures, including the closure of schools, churches, and dancing halls; banning of mass gatherings; mandatory mask wearing; case isolation; and disinfection/hygiene. The primary lesson of the 1918 influenza pandemic was that it is critical to intervene early and that viral spread will be renewed upon relaxation of such measures [ 105 ]. Starting in the early 1980s, a new and unusual disease began to emerge worldwide. The disease was recognized as pandemic when on June 5, 1981, the US Centers for Disease Control (CDC) of Atlanta received information of unusually high rates of the uncommon diseases Pneumocystis jirovecii (then called Pneumocystis carinii ) pneumonia (PCP) and Kaposi's sarcoma in young white healthy homosexual men living in Los Angeles. Afterward, in 1982, cases were reported in injection drug users and then in women possibly infected through heterosexual sex. The disease was named acquired immune deficiency syndrome (AIDS) [ 106 ]. The first appearance of the disease was in Kinshasa, in the Democratic Republic of Congo, in around 1920, when HIV, the causal agent, crossed species from chimpanzees to humans. In early 1983, a new human retrovirus, initially named lymphadenopathy-associated virus (LAV), was isolated at the Pasteur Institute, Paris, France, from a culture obtained by a lymph node biopsy from a patient with generalized lymphadenopathy. Then, a similar virus was isolated from patients with AIDS, and in 1985, the virus RNA was sequenced. The main receptor for HIV was identified in the CD4 cell surface molecule [ 73 ], confirming that HIV causes AIDS. In November 2020, the WHO reported that at the end of 2019, 38.0 million people, 25.7 million of which were in the African Region, were living with HIV [ 107 ]. Table 4 summarizes the interventions adopted to prevent the spread of the modern infectious diseases [ 108 , 109 , 110 , 111 , 112 , 113 , 114 , 115 , 116 ]. 3. Transmission and Measures to Contain SARS-CoV-2 Spread After the WHO declared a pandemic [ 117 ] on 11th March 2020, most countries, to avoid the pandemic spread and limit the number of casualties, introduced several strict nonpharmaceutical interventions [ 118 ], namely (1) improved diagnostic testing and contact tracing; (2) isolation and quarantine for infected people; and (3) measures aimed at reducing mobility and creating social distancing (containment, mitigation, and suppression). Most countries decided on the following containment measures: Physical distances > 1.5 m; Wearing masks and gloves; Stay-at-home orders; School and workplace closures and activation of distance learning and smart working; Closure of museums, commercial parks, gyms, and swimming pools; Cancellation of public events; Restrictions on size of crowds; Seat limitation on public transport to ensure the right distance between passengers; Restrictions on internal and international travel; Measurements of body temperature at the entrance of closed areas (<37.5 °C); Ensuring disinfection rules are followed in public areas such as public transport, shopping areas, schools, and universities; Protecting healthcare workers with appropriate personal protection equipment (PPE). The main objective of these interventions is to reduce the reproduction number (Rt) of the virus. The Rt is defined "as the mean number of secondary cases generated by a typical primary case at time t in a population calculated for the whole period over a 5-day moving average" [ 59 ]. Thus, Rt is an indicator measuring the transmission of SARS-CoV-2 before and after the interventions. According to Ecclesiastes 1:9 [קהלת, Qohelet הַשָּֽׁמֶשׁתַּ֥חַת כָּל־חָדָ֖שׁ וְאֵ֥ין Latin Ecclesiastes Nihil sub sole novum ], "There's nothing new under the sun." Table 5 [ 101 , 102 , 103 , 104 , 105 ] reports the containment measures adopted over the millennia to avoid disease spread. In the past, against some infectious diseases, medicine was ineffective [ 85 ] Thucydides in his History of Peloponnesian War (II, vii3-5) wrote "The doctors were unable to cope, since they were treating the disease for the first time and in ignorance: indeed, the more they came into contact with sufferers, the more liable they were to lose their own lives." Indeed, the only possible way to escape the plague was to avoid any contact with infected persons and contaminated objects. The Italian poet Giovanni Boccaccio (1313–1375), in his book The Decameron (1349–1353), tells the story of ten people, seven women and three men, who entertain themselves with novels while in isolation from the plague of Florence in a villa in the countryside. In the first chapter, Boccaccio describes how the plague struck the city of Florence, how people reacted, and the staggering death toll. Boccaccio, echoing Thucydides, also wrote: "Neither a doctor's advice nor the strength of medicine could do anything to cure this illness". Accordingly, the procedure of obligatory quarantine was introduced as a measure to isolate and separate people, animals, foods, and objects that may have been exposed to a contagious disease. Quarantine is from the Italian "quaranta", meaning forty. For millennia, contagious diseases were believed to be a divine punishment for sinners. Thus, in the Old Testament, God destroyed the earth with water for 40 days (בראשית: Genesis 7:4); Noah waited for forty days after the tops of mountains were seen after the flood (בראשית: Genesis 8:5–7). Moses was on Mount Sinai for 40 days (Exodus שְׁמֹות 24:18): "Then Moses entered the cloud as he went on up the mountain. And he stayed on the mountain forty days and forty nights". In the New Testament, Jesus was tempted for 40 days (Matthew 4:2, Mark 1:13, Luke 4:2). There were 40 days between Jesus' resurrection and ascension (Acts 1:3). Eugenia Tognotti [ 121 ] and Gensini et al. [ 124 ] reviewed the origin of quarantine from the time of the Bible to nowadays. Nevertheless, 40 days may derive from the Pythagorean theory of numbers; according to Pythagoreans, the number 40 was considered to be sacred. Hippocratic teaching in the 5th century BCE established that an acute illness only manifested itself within forty days [ 124 ]. Only during the epidemic of 1347–1352 was an organized institutional response to control disease set up. Quarantine was introduced, for the first time, in 1377 by the Rector of the seaport of Ragusa (Dubrovnik, Croatia), and the first stable plague hospital (lazaretto or quarantine station) was built by the Republic of Venice in 1423 on the island of Santa Maria di Nazareth [ 124 ]. The term lazaretto, usually referred to as Nazarethum or Lazarethum, is related to Lazarus, who was brought back to life by Jesus (John 11:1–45) and/or to the Order of Saint Lazarus of Jerusalem, a Catholic military order founded by crusaders around 1119 at a leper hospital in Jerusalem, as a hospital and military order of chivalry [ 125 ]. The Venetian system became a model for other European countries: in 1467, Genoa adopted the Venetian system, and in 1476, in Marseille, France, a hospital for persons with leprosy was converted into a lazaretto [ 126 ]. Afterward, quarantine became the foundation of a coordinated disease-control strategy that included different measures such as isolation, sanitary cordons, bills of health issued to ships (certification assuring the absence of disease), sanification (i.e., fumigation), and disinfection. Girolamo Fracastoro, Latin Hieronymus Fracastorius (1478–1553, Verona), physician, poet, astronomer, and geologist, was the first to propose, in 1546, a scientific germ theory of disease. In his book "On Contagion and Contagious Diseases", he affirmed that each disease is caused by a different type of rapidly multiplying minute body and that these bodies are transferred from the infector to the infected in three ways: by direct contact; by carriers such as soiled clothing and linen; and through the air. 4. Boosting the Immune Response: How Vaccines Changed the Scenario In the millennial history of mankind, vaccination is a relatively young intervention of primary prevention. For about 200 years, vaccination strategies have had a profound effect in shaping the natural history of infectious diseases. Smallpox eradication represents the most impressive success of a vaccination strategy. As discussed above, smallpox represented a dreadful menace throughout the centuries [ 127 ]. It was common knowledge that smallpox survivors acquired immunity to the disease, so the practice of variolation, consisting in having healthy individuals inhale dust from smallpox lesions, become common in Europe and in North America. At the end of the 18th century, there were anecdotes regarding immunity to smallpox in people previously infected with cowpox, a zoonotic pathogen [ 128 ]. In 1798, Edward Jenner published his first observations on the benefits of inoculating biological material from cowpox lesions in humans, to protect from smallpox, and for the first time, the term "vaccination" (from Latin vacca, English cow) was used [ 129 ]. Initially, vaccination was perpetuated by transferring fluids from individual to individual, but this practice reduced the strength and the duration of protection. The next step was to deliberately infect cows to mass-produce sufficient material ("lymph") for vaccination [ 127 ]. However, this practice led to an increase in the frequency of transmission of secondary infections, including syphilis. This issue was resolved after the observation of bacterial inactivation by glycerin made by Robert Koch [ 130 ], so lymph was treated with glycerin before inoculation. At the end of the 19th century, Louis Pasteur made observations that strongly enhanced the development of vaccines. Studying chicken cholera, he noticed that chickens inoculated with cultures left out over a prolonged period were protected from subsequent infection with fresh material, suggesting the existence of protective immunity induced by the inoculation of "aged" material. These observations would lead Louis Pasteur to the production of antirabic vaccination, using material from an infected dog's brain, exposed to dry air [ 131 ]. In the meantime, studies on the immune system contributed to unravel the mechanisms of host defense from pathogens. In particular, studies by Elie Metchnikoff in 1884 [ 132 ] introduced the concept of cellular immunity, and Paul Ehrlich published his theory of receptor of immunity in 1897, paving the way for the development of antitoxins against pathogens such as diphtheria. At the end of the 19th century, five human vaccines were in use: two live virus vaccines (smallpox and rabies) and three dead bacterial vaccines (typhoid, cholera, and plague.). The first half of the 20th century saw the development of passive immunization, with the production of antitoxins for diphtheria and tetanus. At the same time, new vaccines against tuberculosis, bacillus Calmette-Guérin (BCG), yellow fever, typhus, influenza A, and pertussis were developed, and the first combination vaccine, against diphtheria, tetanus, and pertussis was produced in 1948. Progresses in cell culture lead the way to the techniques of virus attenuation through passages on tissues and cellular monolayers, thanks to the studies of Hugh and Mary Maitland in 1928 and Ernest William Goodpasture in 1931, who first used the chorioallantoic membranes of a fertile hen's egg as a culture medium for sterile passage of viruses. In the 1900s, poliomyelitis (caused by the poliovirus) represented another threat to public health. The virus spreads from person to person and can invade an infected person's brain and spinal cord, causing paralysis. Better hygiene conditions led to an increase in the age of infected children, which in the previous centuries were breastfed, protected by maternal antibodies. The older age of infected children led to frequent polio outbreaks. The first effective antipolio vaccine was a formaldehyde-inactivated (or "killed") PV vaccine (IPV) developed by Jonas Salk in 1955. A second vaccine which was demonstrated to be both safe and effective was the oral (or "live") PV vaccine (OPV) was developed by Albert Sabin in 1963 [ 133 ]. Jonas Salk and Albert Sabin decided not to patent their vaccines and therefore sacrificed billions of dollars in potential royalties, approximately USD 500 million for Salk and approximately USD 1.2 billion for Sabin. Nowadays, thanks to the vaccines, the virus remains endemic only in Afghanistan and Pakistan. In more recent years, research started to focus on multiple vaccines, starting from live viruses attenuated by multiple passages on cultured cells, such as the vaccine against measles, mumps, and rubella. Exploiting the newly available techniques of molecular biology, newly designed vaccines started to be produced. Japanese researchers developed an acellular pertussis vaccine based on two of the main protective antigens of Bordetella pertussis . Research on polysaccharide vaccines led to the development of new vaccines against pneumococcus, meningococcus, and Haemophilus Influenzae type B. Recombinant DNA technology and the possibility to produce recombinant protein in vitro paved the way for the release of the anti-hepatitis B vaccine. Under the urgent need to battle COVID-19, different SARS-CoV-2 vaccines, including the inactivated virus vaccine, nucleic acid vaccine, adenovirus vector vaccine, and viral subunit vaccine, have been developed [ 134 ]. In the history of vaccines, COVID-19 vaccines are unique for the extraordinary rapidity of their production. In recent years, mRNA vaccines have started to attract great attention thanks to their potential to: (1) Speed up vaccine development; (2) Simplify vaccine production, scale-up, and quality control; (3) Be produced and scaled up in a predictable and consistent fashion regardless of the antigen; (4) Have improved safety and efficacy; (5) Challenge diseases impossible to prevent with other approaches; (6) Enable precise antigen design; (7) Generate proteins with a "native-like" presentation; (8) Express proteins stabilized in a more immunogenic conformation or expose key antigenic sites; (9) Deliver multiple mRNAs to the same cell; (10) Allow the generation of multiprotein complexes or protein antigens from different pathogens, thus creating a single vaccine against several targets. Moreover, mRNA is characterized as: (11) Being noninfectious; (12) Being nonintegrating; (13) Being degradable by normal cellular processes soon after injection; (14) Decreasing the risk of toxicity and long-term side effects; (15) Not inducing vector-specific immunity; (16) Not competing with pre-existing or newly raised vector immunity that could interfere with subsequent vaccinations. An mRNA vaccine is based on the principle that mRNA is an intermediate messenger to be translated to an antigen after the delivery into host cells via various routes. The mRNA is synthesized in the laboratory by transcribing a DNA template of the genetic sequence encoding the immunogen. In the case of SARS-CoV-2, the spike (S) protein is identified as the immunodominant antigen of the virus. The most important problem is that mRNA is unstable, easily recognized by the immune system, and rapidly degraded by nucleases after entering the body. mRNA vaccines do not enter the nucleus but need to pass through the cell membrane, a negatively charged phospholipid bilayer, to enter the cytoplasm and then be translated into the target protein. Different delivery systems for mRNA vaccines, such as viral and nonviral vector delivery systems, may be utilized. Vectors based upon lipids or lipid-like compounds are the most common nonviral gene carriers. 5. Conclusions The deep knowledge of the history of the "plagues" that have struck humanity is not only precious in understanding the long-term sociological and demographic changes but in better understanding the evolution of infectious diseases over the centuries. Reading all the literary works carefully, we can recognize our present journeys under the COVID-19 pandemic, including the risks of inappropriate responses. Confinement measures such as social distancing and/or quarantine still remain the most efficient measure to contain the spread of the virus. As in the past, an "infodemia" is present, generating chaos and fear among the population. Moreover, technical knowledge on agents, hosts, and the environment alone, although essential, is not enough. The COVID-19 pandemic is an example of how diseases unknown to medical science and to human immune systems may develop and spread quickly in a highly connected world. Even it is not possible to avoid every risk, there are ways to reduce or mitigate the chances of a future pandemic, such as investing in research and preparation, funding and implementing vaccine programs, and strengthening health systems. Along with vaccines and specific therapies, the best course of action in facing new pandemics remains social distancing, the practice of good hygiene, and the use of quarantine. "And Darkness and Decay and the Red Death held illimitable dominion over all." ("The Mask of the Red Death: A Fantasy" E. A. Poe, 1842)

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3922542/

Comparison of three next-generation sequencing platforms for metagenomic sequencing and identification of pathogens in blood

Background The introduction of benchtop sequencers has made adoption of whole genome sequencing possible for a broader community of researchers than ever before. Concurrently, metagenomic sequencing (MGS) is rapidly emerging as a tool for interrogating complex samples that defy conventional analyses. In addition, next-generation sequencers are increasingly being used in clinical or related settings, for instance to track outbreaks. However, information regarding the analytical sensitivity or limit of detection (LoD) of benchtop sequencers is currently lacking. Furthermore, the specificity of sequence information at or near the LoD is unknown. Results In the present study, we assess the ability of three next-generation sequencing platforms to identify a pathogen (viral or bacterial) present in low titers in a clinically relevant sample (blood). Our results indicate that the Roche-454 Titanium platform is capable of detecting Dengue virus at titers as low as 1X10 2.5 pfu/mL, corresponding to an estimated 5.4X10 4 genome copies/ml maximum. The increased throughput of the benchtop sequencers, the Ion Torrent PGM and Illumina MiSeq platforms, enabled detection of viral genomes at concentrations as low as 1X10 4 genome copies/mL. Platform-specific biases were evident in sequence read distributions as well as viral genome coverage. For bacterial samples, only the MiSeq platform was able to provide sequencing reads that could be unambiguously classified as originating from Bacillus anthracis. Conclusion The analytical sensitivity of all three platforms approaches that of standard qPCR assays. Although all platforms were able to detect pathogens at the levels tested, there were several noteworthy differences. The Roche-454 Titanium platform produced consistently longer reads, even when compared with the latest chemistry updates for the PGM platform. The MiSeq platform produced consistently greater depth and breadth of coverage, while the Ion Torrent was unequaled for speed of sequencing. None of the platforms were able to verify a single nucleotide polymorphism responsible for antiviral resistance in an Influenza A strain isolated from the 2009 H1N1 pandemic. Overall, the benchtop platforms perform well for identification of pathogens from a representative clinical sample. However, unlike identification, characterization of pathogens is likely to require higher titers, multiple libraries and/or multiple sequencing runs. Background The introduction of benchtop sequencers has made adoption of whole genome sequencing possible for a broader community of researchers than ever before. Concurrently, metagenomic sequencing (MGS) is rapidly emerging as a tool for interrogating complex samples that defy conventional analyses. In addition, next-generation sequencers are increasingly being used in clinical or related settings, for instance to track outbreaks. However, information regarding the analytical sensitivity or limit of detection (LoD) of benchtop sequencers is currently lacking. Furthermore, the specificity of sequence information at or near the LoD is unknown. Results In the present study, we assess the ability of three next-generation sequencing platforms to identify a pathogen (viral or bacterial) present in low titers in a clinically relevant sample (blood). Our results indicate that the Roche-454 Titanium platform is capable of detecting Dengue virus at titers as low as 1X10 2.5 pfu/mL, corresponding to an estimated 5.4X10 4 genome copies/ml maximum. The increased throughput of the benchtop sequencers, the Ion Torrent PGM and Illumina MiSeq platforms, enabled detection of viral genomes at concentrations as low as 1X10 4 genome copies/mL. Platform-specific biases were evident in sequence read distributions as well as viral genome coverage. For bacterial samples, only the MiSeq platform was able to provide sequencing reads that could be unambiguously classified as originating from Bacillus anthracis. Conclusion The analytical sensitivity of all three platforms approaches that of standard qPCR assays. Although all platforms were able to detect pathogens at the levels tested, there were several noteworthy differences. The Roche-454 Titanium platform produced consistently longer reads, even when compared with the latest chemistry updates for the PGM platform. The MiSeq platform produced consistently greater depth and breadth of coverage, while the Ion Torrent was unequaled for speed of sequencing. None of the platforms were able to verify a single nucleotide polymorphism responsible for antiviral resistance in an Influenza A strain isolated from the 2009 H1N1 pandemic. Overall, the benchtop platforms perform well for identification of pathogens from a representative clinical sample. However, unlike identification, characterization of pathogens is likely to require higher titers, multiple libraries and/or multiple sequencing runs. Background Metagenomic sequencing involves high-throughput sequencing of complex samples comprised of nucleic acid from multiple organisms. Although the technique is relatively new, published examples of metagenomic sequencing to detect and/or characterize a causal agent in diseases of animals and humans are already too numerous to be summarized here. In some cases, as in the case of a novel filovirus that caused an outbreak in Uganda in 2007, metagenomic sequencing was used to follow up when results from traditional diagnostic assays indicated a novel agent. For instance, the Ugandan filovirus had produced a positive result in a broadly cross-reactive IgM capture assay followed by mixed results in RT-PCR assays for known filoviruses, and so metagenomic sequencing was employed to characterize the genome of the novel virus [ 1 ]. Another example whereby more traditional diagnostic assays and metagenomic sequencing both played a role in detection is the identification of human metapneumovirus causing fatal infection of wild Rwandan gorillas [ 2 ]. The opposing scenario, in which traditional diagnostic assays completely fail and metagenomic sequencing plays more than just a supporting role, includes a recent report of astrovirus encephalitis in an immune-compromised teenage boy [ 3 ]. Metagenomic sequencing has many advantages over more traditional methods of pathogen detection, such as PCR, ELISA, cytopathic effect, etc. These advantages include relative speed, the ability to detect non-culturable organisms, and, perhaps most importantly, the fact that metagenomic sequencing requires little to no a priori knowledge of the organism(s). However, despite these advantages, there are some significant challenges involved. These difficulties may include computationally intensive analyses; in some cases the lack of appropriate reference genomes for comparison; difficulties in sample preparation, including matrix effects and biases introduced from extraction; as well as the need for sufficient depth and breadth of coverage to detect pathogens at potentially very low levels in a given sample. It is also important to note that there are currently no standards in terms of what constitutes "identification" of a pathogen in a sample. A number of organism-specific reads will be necessary to make a true positive call, but how many reads or signatures is necessary to invoke confidence is not standardized and may vary as per sample type/complexity or may be organism-dependent. It is likely that increased depth of coverage would increase confidence in a species or strain call, but again, there is no agreed upon standard regarding depth of coverage for metagenome samples, and in many cases it would be reasonable to expect that a region of a pathogen genome detected within a metagenome sample may be only present within the reads at 1X coverage. Just as there is currently no agreed upon standard to indicate what breadth or depth of coverage would be required to make an "identification," there is currently a paucity of knowledge regarding the actual limits of detection (LoD) for each sequencing platform and protocol. Detection of even one or several pathogen-specific reads in a clinical sample that are not found in control samples is likely to be interpreted as a positive result. However, in the absence of LoD data, it is difficult to conclude with any confidence that a pathogen is not present simply because no pathogen-specific reads are detected. Recently, Moore et al. conducted a LoD study whereby viral RNA in serial dilutions was spiked into a colorectal biopsy sample and sequenced using the Illumina platform. In this study, although the proportion of viral reads detected was less than expected, potentially due to issues quantitating the RNA, virus-specific reads were detected from samples spiked with less than picogram quantities [ 4 ]. In another study, Cheval et al. spiked plasma and cerebrospinal fluid with known concentrations of eleven different viruses to assess the level of detection by Roche-454 pyrosequencing as compared to Illumina, and found that the higher output (number of reads) produced by the Illumina platform resulted in better detection of the viruses per run. The authors report detection of viruses by their 454 sequencing at titers of 1X10 3 pfu/ml and higher, and by Illumina of viral genomes present at 1X10 2.4 genome copies/ml and higher [ 5 ]. Other studies have suggested that the LoD by 454 sequencing for an RNA virus with a genome size of ~10-11 kb lies between 1X10 2 and 1X10 3 pfu/mL [ 6 ]. Likewise, in a recent study using artificially constructed marine metagenomes, Pochon et al. determined that the 454 GS Junior was capable of detecting DNA from species present at levels greater than 0.64% of the whole. It should be noted, however, that this study used amplicons from a single gene to determine sensitivity [ 7 ]. Finally, variation in extraction techniques and/or sample preparation kits have been shown to affect the composition of sequence reads [ 8 , 9 ]. Thus, any reported LoD is likely to be very protocol specific, as modifications introduced at each step of the procedure can have drastic effects on overall sequence read output and quality. In a previous study, we assessed the ability of the Roche-454 platform to detect low-level pathogens present in a complex host background, insect vectors [ 6 ]. Results indicated that pathogen reads present in proportions of 0.1% were sufficient to cover 90% of a viral genome at a depth of >4X. The same study postulated the number of sequence reads required to identify a pathogen present in a sample in which 1 out of 100 pooled mosquitoes is infected. The projected number of reads necessary was expected to vary depending on titer, with high titer infections requiring hundreds of thousands of reads, while low titer infections would require tens of millions. In the current study we use the latest sequencing technologies with their increased throughput and lower cost to confirm these projections as well as establish metrics for pathogen identification from clinical samples. Results Limits of detection of Roche-454 pyrosequencing Our previous study suggested that as few as 0.1% of the total reads were sufficient to unambiguously identify a viral pathogen present in a complex background [ 6 ]. However, our experimental model was reflective of a disseminated infection and high viral load [ 6 , 10 - 12 ]. Therefore, we wished to determine if the Roche-454 platform was capable of detecting viruses at reduced titers. To determine the limits of detection (LoD), mock samples were created by spiking known amounts of Dengue virus Type 1 (DENV-1) into 1 mL aliquots of whole human blood. Total RNA was extracted from each sample and used for construction of cDNA libraries. Sequencing of samples representing each titer was conducted in replicate fashion on the Roche-454 platform and DENV-1 was detected in a range of titers from 1X10 3 to 1X10 5 pfu/ml. Since no DENV-1 specific reads were detected at the titer of 1X10 2 , it was determined that the effective LoD for 454 sequencing using this protocol must lie between 1X10 2 and 1X10 3 , and so several samples were prepared within this range (1X10 2.3 , 1X10 2.5 , and 1X10 2.7 ). Sequencing was initiated with the 1X10 2.5 sample and progressed down to 1X10 2.3 , from which no reads were detected. Overall, the decrease in the number of viral reads approximated the reduction in titer, especially at lower titers. At the lowest titer detectable, 1X10 2.5 , the breadth of coverage of the reference genome was 31% with an average depth of coverage of 0.35X upon read mapping (Figure 1 ). In a parallel approach, de novo assembly of reads yielded a single contig of 427 bp in length. By comparison, assemblies of samples at 1X10 5 pfu/mL and 1X10 4 pfu/mL yielded one contig of 9445 bp and two contigs of 4640 and 3566 bp, respectively. In both cases the largest contig in the de novo assembly belonged to DENV-1. It is striking that, in the case of the low titer samples, de novo assembly of the reads provides less information than would a taxonomic classification of all reads. In the case of the 1X10 2.5 pfu/mL sample, although 11/604,130 resulting reads map to the reference, covering 31% of the viral genome, de novo assembly of the reads produces a single contig comprised of 2 reads that covers only 4% of the viral genome (Table 1 ). This apparent difference between fraction of the viral genome covered by the de novo assembly and the reference mapping is due to insufficient breadth of coverage for the reads to overlap and be used in de novo assembly. Figure 1 Read mapping against DENV-1 genome at 1X10 2.5 pfu/ml. Reads resulting from 1 454 Titanium sequencing run of a cDNA library made from blood spiked with DENV-1 at titer of 1X10 2.5 pfu/ml were aligned to the reference genome NC_001477.1 using CLC Genomics Workbench version 6.0.4 at default parameters. Table 1 Roche-454 read statistics for DENV-1 at multiple titrations Dengue-1 pfu/mL Estimated max. genome copies/mL Reads mapped per 100,000 Fraction of reference covered Average C t Contig(s) by de novo assembly? Fraction of reference covered 1X10 2 1X10 4 – 1.7X10 4 0 0 NA N 0 1X10 2.3 2X10 4 – 3.4X10 4 0 0 27.65 N 0 1X10 2.5 3.2X10 4 – 5.4X10 4 0.34 0.05 ND N 0 1X10 2.5 3.2X10 4 – 5.4X10 4 1.82 0.31 27.03 Y 0.04 1X10 3 1X10 5 – 1.7X10 5 3.4 0.4 25.66 Y 0.20 1X10 3 1X10 5 – 1.7X10 5 4.7 0.72 ND ND ND 1X10 4 1X10 6 – 1.7X10 6 34.2 0.92 21.79 Y 0.76 1X10 5 1X10 7 – 1.7X10 7 190.9 0.99 18.05 Y 0.99 1X10 6 1X10 8 – 1.7X10 8 227.1 0.99 14.56 Y 0.99 Comparison of Roche-454, Illumina MiSeq and Ion Torrent PGM for detecting viral pathogens Next, we examined how the performance of benchtop sequencers would compare with that of the Roche-454. To that end, mock samples containing Influenza virus H1N1 were prepared in a similar fashion to the DENV-1 samples. For this experiment, the DENV-1 results were used as a guide as to what range of titers to investigate. The effective LoD for the 454 platform was found to be 1X10 2.5 , which we estimate to correlate with an upper limit of 31,600-53,720 genome copies/ml, based on work by Houng et al. and Wang et al., who found one DENV pfu to correspond to 100 and 170 genome copies by qPCR, respectively [ 13 , 14 ]. Therefore, an initial titer of 1X10 4 - genome copies/mL was chosen for these experiments. At 1X10 4 genome copies/mL the Roche-454 platform was capable of detecting the Influenza virus in only one out of two replicates, with only a single hit covering 3% of the flu genome (Table 2 ). At the same concentration, two biological replicates sequenced on the PGM platform scored 11 and 21 hits, covering 7% and 12% of the genome, respectively. By contrast, two biological replicates sequenced on the MiSeq platform scored 308 and 811 hits, covering 79% and 95% of the genome, respectively. Interestingly, even when the number of Influenza-specific reads is expressed as reads per 100,000, the MiSeq and PGM platforms both outperform the Roche-454, indicating that there may be some platform-specific factors that affect LoD in addition to total output per run. Table 2 Cross-platform comparison of read mapping to Influenza A H1N1 Platform #/Replicates Reads mapped (low) Reads mapped per 100,000 a Reads mapped (med) Reads mapped in pairs a Fraction of reference covered a PGM 1 3 11/20/ND .61/.80/ND 5/16/ND NA .07/.15/ND PGM 2 3 21/19/18 .87/1.3/.86 14/11/10 NA .12/.09/.14 Roche 2 1/0 .17/0 0/0 NA .03/- MiSeq 1 3 308/322/323 .82/.90/1.0 269/311/318 66/72/82 .79/.95/.80 MiSeq 2 3 811/942/912 2.2/2.4/2.2 741/934/914 296/216/248 .95/.95/.97 a: Reads mapped at low stringency. Examination of the read mappings revealed that the mapped reads were not evenly distributed along the viral genome for any of the three sequencing platforms. Figure 2 A depicts representative read mappings by platform for segment 5 of Influenza A H1N1. For the PGM, the regions of highest coverage were found most often toward the 5′ end of segment 5. Specifically, 14/19 mapped reads aligned within the first quarter of the genome segment. The single mapped read from the Roche-454 was 448 bases in length. By contrast, the longest mapped read for the PGM was 179 bases. As expected, the mapped reads from the MiSeq platform were all 150 bases in length. Since the PGM offers scalability in throughput, we sought to compare the sensitivity of the PGM using a 314 chip as well. The throughput of the 314 chip is similar to that of one-half of a 454 picotitre plate. Although not shown in Table 2 , two independent experiments using a 314 chip yielded a single, 84 base read mapped to Segment 5 (Additional file 1 : Figure S1). This single read mapped within the first quarter of Segment 5. These data strongly suggest that increased throughput is required for detection of pathogens present at low levels. Figure 2 Read mapping against representative segments of Influenza genome. In A) , reads resulting from a representative run of 454 Titanium sequencing, Ion Torrent PGM sequencing, and Illumina MiSeq sequencing were mapped to the reference Influenza A H1NI segment 5 [NC_002019.1], using CLC Genomics Workbench version 6.0.4 at default parameters, whereas in B) , they were mapped against segment 8 [NC_002020.1]. In B) , no Roche-454 reads mapped to the reference. In both A) and B) , coordinates of reference genome segment are displayed along the top and G/C content is graphed below reference in pink. A closer examination of the read mappings for both the PGM and MiSeq platforms indicated that some mapped reads were spurious. In order to eliminate these matches, stringency of read mapping parameters was increased. At the medium stringency setting, whereby the length of the read required to match was increased from 50% to 70%, the number of mapped reads dropped by roughly half for the PGM and by roughly 10% for the MiSeq. Furthermore, the number of MiSeq reads that mapped in pairs was roughly 25% (66 out of 269; Table 2 ). This is most likely due to polymorphisms in the sequenced strain that are not present in the NCBI reference genome. As expected, no reads mapped at the highest level of stringency tested, as the requirement for 100% identity would preclude any matches. For Influenza virus segment 8, the MiSeq yielded 14 mapped reads versus the single mapped read from the PGM. It is noteworthy that although the MiSeq reads covered more of the segment overall than the single PGM read did, both platforms produced a single read in an area of high G/C content. Given that the MiSeq consistently produced the highest depth of coverage, we next examined whether the increased coverage of the MiSeq was sufficient to determine the presence of a single nucleotide polymorphism (SNP) that conveys resistance to a widely used antiviral drug, oseltamavir (Ose). Previous studies have established that Influenza A strains from the 2009 epidemic had mutated to an Ose-resistant phenotype [ 15 ]. Thus, we examined the reads resulting from MiSeq sequencing of a 2009 epidemic strain for the presence of the mutation H275Y, and in neither of two biological replicates were any of the known SNPs detected. In fact, there was zero coverage of that specific region of the genome. Given that our experimental model, human blood, represents a relatively low-complexity sample (as compared to clinical samples from other body sites such as oropharyngeal swabs or fecal samples), spiked with Influenza A genomes at titers higher than would be expected to be encountered in a clinical sample, these data suggest that pathogen enrichment may be a necessary step for determination of drug resistance by metagenome sequencing. de novo assembly and BLASTn analysis of Influenza A reads In many cases the identity of a pathogen may not be known or the pathogen's genome may be very divergent at the nucleic acid level from that of known family members, necessitating analysis by means other than mapping reads to a reference genome. Therefore, de novo assemblies of the Influenza A metagenomes from the PGM and MiSeq data were performed, followed by BLASTn analysis. Assembly of the PGM reads using MIRA produced an average of 28,000 contigs, yet not a single contig was classified as Influenza A (Table 3 ). In fact, no contig was classified as a viral taxon. This result is surprising given the observed sequence depth in several regions of the Influenza A genome (see Figure 2 A). We also attempted assembly of the PGM reads using Velvet with similar assembly parameters. Again, none of the resultant contigs were able to be assigned to Influenza A. Upon closer inspection, it was observed that the identified reads were either substrings of each other and therefore absorbed during the merge step of assembly, or they had a stretch of mismatches long enough (≥ 10 bases) to cause the assembly algorithm to mark these reads as not overlapping. Table 3 De novo assembly and BLASTn statistics of Influenza A sequence reads Platform Assembler # total contigs # specific contigs PGM 1 MIRA 26,171 0 PGM 2 MIRA 30,719 0 MiSeq 1 Velvet 841,384 18 MiSeq 2 Velvet 768,912 11 By comparison, assembly of the MiSeq reads for two biological replicates using Velvet generated well over 750,000 contigs for both replicates, with 18 and 11 flu-specific contigs, respectively. Taxonomic classification of the contigs by MEGAN properly assigned the correct strain (Influenza A virus (A/swine/Iowa/15/1930(H1N1)) in 5 out of 18 and 0 out of 11 contigs. Importantly, only the strain level classifications were mis-assigned. In each case, the correct genus and species calls were made. Therefore, each of the contigs identified by MEGAN as Influenza but not correctly identified at strain level was individually subjected to BLASTn analysis and manual curation of results. This inspection revealed that the proper strain was indeed present in the hit tables (6 and 7 contigs, respectively), but that for each of these contigs, there were several strains hit with the identical E-value, including A/swine/Iowa/15/1930(H1N1), and therefore, a definitive strain call could not be made in those cases. Thus, for one of two replicates, a majority of the influenza-specific contigs resulting from de novo assembly of the MiSeq reads were either properly classified to the strain level, or conserved amongst a subset of strains including A/swine/Iowa/15/1930(H1N1). It was interesting to note that, in each case, contigs corresponding to Segments 4 (HA) and 6 (NA) were not assigned to the proper strain. These results demonstrate that the increased throughput of the MiSeq enables species level identification of reads present in proportions as low as .0008% of the total. Comparison of Influenza A sequencing replicates on the PGM and MiSeq We were interested in further examining the observed differences in the number of mapped reads between biological replicates and between genome segments. In addition, we wished to determine the reproducibility of our results. The biological replicates (derived from independent spikings and library preparations) sequenced on the MiSeq platform exhibited considerable variability (Table 2 ), and we were interested in determining at what stage this variability would arise- at library construction or sequencing. Thus, we performed replicate sequencing runs for each of the PGM and MiSeq libraries. The total number of mapped reads for each technical replicate was quite consistent (Table 2 ). This observation held true for both platforms. Moreover, each individual library yielded a number of mapped reads that was less than one standard deviation from the average. Thus, the inter-run variability seems to be limited and results of sequencing a given library preparation are highly reproducible. However, we did note several tendencies regarding the number of mapped reads per segment (Additional file 2 : Table S1 and Additional file 3 : Table S2). Firstly, segment 5 tended to be overrepresented in each PGM replicate. Specifically, for comparison, we expressed the proportion of mapped reads for a given genome segment as a function of that particular segment's length, as follows: y = # mapped reads per given segment / # total reads mapped / segment length × 1000 This value for segment 5 is more than twice that for any other segment by PGM sequencing (Additional file 3 : Table S2). This is an interesting observation given that segment 5, at 1565 bases in length, is the fifth largest segment in the genome. In every case for the PGM, segment 5 yielded the most mapped reads. This observation is consistent with our previous result with the 314 chip in which the only segment mapped was segment 5 (Additional file 1 : Figure S1). Secondly, segment 3 tended to be underrepresented, in both the PGM and MiSeq data. This trend was especially marked in the PGM libraries as compared to the MiSeq libraries (.03 and .05 respectively). These findings suggest that some sequences are inherently favored for sequencing by the PGM as well as the MiSeq. Thirdly, regardless of the number of reads mapped, the coordinates of the mapped reads in relation to the genome were widely variable (Figure 3 ). This effect was more pronounced in the PGM libraries. This suggests that the library diversity is greater than that sampled by a single sequencing run. Figure 3 Read mapping against segment 5 of Influenza A of replicate sequencing runs. Reads resulting from a replicate run of Ion Torrent PGM sequencing (top) and Illumina MiSeq sequencing (bottom) were mapped to the reference Influenza a H1NI segment 5, [NCBI accession: NC_002019 ] using CLC Genomics Workbench version 6.0 at default parameters. Coordinates of reference genome segment are displayed along the top and G/C content is graphed below reference in pink. Comparison of Roche-454, Illumina MiSeq and Ion Torrent PGM for detecting bacterial pathogens Although there have been performance comparisons of the benchtop platforms for sequencing bacterial isolates [ 16 , 17 ] to our knowledge, there has been no published report of comparisons of sensitivity and specificity in terms of speciation. Therefore, we aimed to establish a LoD for each sequencer using samples spiked with a known biothreat agent, Bacillus anthracis. Using the Sterne strain as a surrogate of anthrax infection, the quantity of bacterial cells present in a dilution series was established. Using three complementary quantitative techniques, mock samples with B. anthracis Sterne present at roughly 30,000 cells/mL of human blood were prepared. All three platforms delivered sequence reads matching the chromosome of B. anthracis Sterne by reference mapping (Table 4 ). As expected, given the increased throughput, both PGM and MiSeq yielded significantly higher numbers of reads as well as genome coverage. However, at low stringency, large numbers of reads were falsely counted as mapped. This trend was especially prominent in the PGM data. Most of the spurious reads were short (average: 40 bases) and low-complexity (i.e. rich in nucleotide repeats). This tendency was confirmed in the replicate samples. The same pattern was observed in the Roche-454 sequence reads as well. In general, the 454 reads were clustered near the 5′-end of the linear reference sequence, while the PGM and MiSeq reads spanned a much larger breadth. However, the 454 reads were uniformly longer and more information-rich. The MiSeq replicates demonstrated a similar tendency as that of the PGM; in general, the lower stringency mapping contained matched reads of short length (avg. 66 bases). For comparison, in replicate two for both platforms, 80.77% of mapped MiSeq reads mapped as non-perfect matches versus 98.83% of mapped PGM reads. The average read length of a non-perfectly matched MiSeq read was 53.22 bases, as compared to 39.99 bases for the PGM. Interestingly, the mean number of mismatches per read tended to be higher in the MiSeq data than that of the PGM (mean of 26 for MiSeq vs. 15 mismatches for PGM). Due to the slightly longer length of mapped MiSeq reads, there could conceivably be more opportunity for mismatched bases per read. Therefore, we compared the mismatches within MiSeq reads and PGM reads by expressing mean number of mismatches per read as a proportion of mean read length of non-perfectly matched reads. When expressed in this manner, the proportion of mismatches per length of non-perfectly matched reads is 0.5 for the MiSeq versus 0.4 for the PGM. Additionally, the number of indel related mismatches was found to be proportionally higher in the MiSeq reads than in the PGM reads. Conversely, the PGM reads had a higher number of A → T and G → C transversions than the MiSeq data. This is inconsistent with the known error models for both of these platforms. Table 4 Cross-platform Comparison of Read Mapping to B. anthracis Sterne Platform #Replicates Reads mapped (low) Reads mapped (med) Reads mapped (high) Fraction of reference covered a Detected by qPCR? PGM 2 29,534/15,676 7,689/4,286 247/178 .12/.09 Y Roche-454 2 384/376 249/240 65/56 .01/.01 Y MiSeq 2 10,415/41,242 3,024/9,985 1,633/7,930 .07/.19 Y a: Fraction of reference covered using high stringency mapping parameters as defined in Methods. Although de novo assemblies and taxonomic classification were attempted for both PGM and 454 reads, no dataset produced a single contig that matched Bacillus spp. This is most likely due to the lack of coverage of the bacterial genome, resulting in a lack of sufficient overlap between reads necessary to form a contig. By comparison, de novo assembly of one MiSeq run yielded 6,140,965 total contigs. Taxonomic classification of the assembled sequence reads rendered 1200 contigs that were assigned to the Bacillus cereus superfamily. More importantly, 70 contigs were properly classified as Bacillus anthracis Sterne. These results echo those of the viral samples in that assembled sequence reads from the MiSeq were able to identify the spiked-in microorganism, in this case to the strain level. Detection of genetic engineering We were interested in whether sequencing reads near the LoD could detect an instance of genetic engineering. Therefore, the strain of B. anthracis spiked into human blood for these experiments was a 34F2 strain of B. anthracis containing a stably integrated genomic copy of red fluorescent protein (RFP). Metagenomic sequence data was compared with a computer simulation of the likelihood of detecting this genetic manipulation. For this purpose, we developed a Perl-based algorithm to generate a distribution of the number of sequence reads necessary to detect a gene of 1000 bps inserted at five locations into a prokaryotic genome of 5 Mb in size. At 1000 organism-specific reads, the probability of detecting the inserted gene is roughly one-third (Figure 4 ). In fact, the minimum number of organism-specific hits necessary to ensure detection of the inserted gene is 10,000. Given that 5 of the 6 metagenomes sequenced here contained far fewer than 1000 bacillus-specific reads (Table 4 ), one would expect few, if any, sequence reads to match the RFP reference sequence. Indeed, read mapping to the available NCBI reference (EF606900.1) yielded zero matches to the RFP gene present in the 34F2 chromosome. These data suggest that increased titres, targeted enrichment, or both will be necessary for detection of gene-insertion events. Figure 4 Mathematical modeling of the likelihood of detecting a genetic modification in B. anthracis. The expected number of hits to an inserted gene of size 1 kb, at 5 copies, was simulated as a function of the number of organism-specific reads collected from the metagenomic sample. The relative size of each rectangle indicates the proportion of samplings for which a specific number of hits is expected. Limits of detection of Roche-454 pyrosequencing Our previous study suggested that as few as 0.1% of the total reads were sufficient to unambiguously identify a viral pathogen present in a complex background [ 6 ]. However, our experimental model was reflective of a disseminated infection and high viral load [ 6 , 10 - 12 ]. Therefore, we wished to determine if the Roche-454 platform was capable of detecting viruses at reduced titers. To determine the limits of detection (LoD), mock samples were created by spiking known amounts of Dengue virus Type 1 (DENV-1) into 1 mL aliquots of whole human blood. Total RNA was extracted from each sample and used for construction of cDNA libraries. Sequencing of samples representing each titer was conducted in replicate fashion on the Roche-454 platform and DENV-1 was detected in a range of titers from 1X10 3 to 1X10 5 pfu/ml. Since no DENV-1 specific reads were detected at the titer of 1X10 2 , it was determined that the effective LoD for 454 sequencing using this protocol must lie between 1X10 2 and 1X10 3 , and so several samples were prepared within this range (1X10 2.3 , 1X10 2.5 , and 1X10 2.7 ). Sequencing was initiated with the 1X10 2.5 sample and progressed down to 1X10 2.3 , from which no reads were detected. Overall, the decrease in the number of viral reads approximated the reduction in titer, especially at lower titers. At the lowest titer detectable, 1X10 2.5 , the breadth of coverage of the reference genome was 31% with an average depth of coverage of 0.35X upon read mapping (Figure 1 ). In a parallel approach, de novo assembly of reads yielded a single contig of 427 bp in length. By comparison, assemblies of samples at 1X10 5 pfu/mL and 1X10 4 pfu/mL yielded one contig of 9445 bp and two contigs of 4640 and 3566 bp, respectively. In both cases the largest contig in the de novo assembly belonged to DENV-1. It is striking that, in the case of the low titer samples, de novo assembly of the reads provides less information than would a taxonomic classification of all reads. In the case of the 1X10 2.5 pfu/mL sample, although 11/604,130 resulting reads map to the reference, covering 31% of the viral genome, de novo assembly of the reads produces a single contig comprised of 2 reads that covers only 4% of the viral genome (Table 1 ). This apparent difference between fraction of the viral genome covered by the de novo assembly and the reference mapping is due to insufficient breadth of coverage for the reads to overlap and be used in de novo assembly. Figure 1 Read mapping against DENV-1 genome at 1X10 2.5 pfu/ml. Reads resulting from 1 454 Titanium sequencing run of a cDNA library made from blood spiked with DENV-1 at titer of 1X10 2.5 pfu/ml were aligned to the reference genome NC_001477.1 using CLC Genomics Workbench version 6.0.4 at default parameters. Table 1 Roche-454 read statistics for DENV-1 at multiple titrations Dengue-1 pfu/mL Estimated max. genome copies/mL Reads mapped per 100,000 Fraction of reference covered Average C t Contig(s) by de novo assembly? Fraction of reference covered 1X10 2 1X10 4 – 1.7X10 4 0 0 NA N 0 1X10 2.3 2X10 4 – 3.4X10 4 0 0 27.65 N 0 1X10 2.5 3.2X10 4 – 5.4X10 4 0.34 0.05 ND N 0 1X10 2.5 3.2X10 4 – 5.4X10 4 1.82 0.31 27.03 Y 0.04 1X10 3 1X10 5 – 1.7X10 5 3.4 0.4 25.66 Y 0.20 1X10 3 1X10 5 – 1.7X10 5 4.7 0.72 ND ND ND 1X10 4 1X10 6 – 1.7X10 6 34.2 0.92 21.79 Y 0.76 1X10 5 1X10 7 – 1.7X10 7 190.9 0.99 18.05 Y 0.99 1X10 6 1X10 8 – 1.7X10 8 227.1 0.99 14.56 Y 0.99 Comparison of Roche-454, Illumina MiSeq and Ion Torrent PGM for detecting viral pathogens Next, we examined how the performance of benchtop sequencers would compare with that of the Roche-454. To that end, mock samples containing Influenza virus H1N1 were prepared in a similar fashion to the DENV-1 samples. For this experiment, the DENV-1 results were used as a guide as to what range of titers to investigate. The effective LoD for the 454 platform was found to be 1X10 2.5 , which we estimate to correlate with an upper limit of 31,600-53,720 genome copies/ml, based on work by Houng et al. and Wang et al., who found one DENV pfu to correspond to 100 and 170 genome copies by qPCR, respectively [ 13 , 14 ]. Therefore, an initial titer of 1X10 4 - genome copies/mL was chosen for these experiments. At 1X10 4 genome copies/mL the Roche-454 platform was capable of detecting the Influenza virus in only one out of two replicates, with only a single hit covering 3% of the flu genome (Table 2 ). At the same concentration, two biological replicates sequenced on the PGM platform scored 11 and 21 hits, covering 7% and 12% of the genome, respectively. By contrast, two biological replicates sequenced on the MiSeq platform scored 308 and 811 hits, covering 79% and 95% of the genome, respectively. Interestingly, even when the number of Influenza-specific reads is expressed as reads per 100,000, the MiSeq and PGM platforms both outperform the Roche-454, indicating that there may be some platform-specific factors that affect LoD in addition to total output per run. Table 2 Cross-platform comparison of read mapping to Influenza A H1N1 Platform #/Replicates Reads mapped (low) Reads mapped per 100,000 a Reads mapped (med) Reads mapped in pairs a Fraction of reference covered a PGM 1 3 11/20/ND .61/.80/ND 5/16/ND NA .07/.15/ND PGM 2 3 21/19/18 .87/1.3/.86 14/11/10 NA .12/.09/.14 Roche 2 1/0 .17/0 0/0 NA .03/- MiSeq 1 3 308/322/323 .82/.90/1.0 269/311/318 66/72/82 .79/.95/.80 MiSeq 2 3 811/942/912 2.2/2.4/2.2 741/934/914 296/216/248 .95/.95/.97 a: Reads mapped at low stringency. Examination of the read mappings revealed that the mapped reads were not evenly distributed along the viral genome for any of the three sequencing platforms. Figure 2 A depicts representative read mappings by platform for segment 5 of Influenza A H1N1. For the PGM, the regions of highest coverage were found most often toward the 5′ end of segment 5. Specifically, 14/19 mapped reads aligned within the first quarter of the genome segment. The single mapped read from the Roche-454 was 448 bases in length. By contrast, the longest mapped read for the PGM was 179 bases. As expected, the mapped reads from the MiSeq platform were all 150 bases in length. Since the PGM offers scalability in throughput, we sought to compare the sensitivity of the PGM using a 314 chip as well. The throughput of the 314 chip is similar to that of one-half of a 454 picotitre plate. Although not shown in Table 2 , two independent experiments using a 314 chip yielded a single, 84 base read mapped to Segment 5 (Additional file 1 : Figure S1). This single read mapped within the first quarter of Segment 5. These data strongly suggest that increased throughput is required for detection of pathogens present at low levels. Figure 2 Read mapping against representative segments of Influenza genome. In A) , reads resulting from a representative run of 454 Titanium sequencing, Ion Torrent PGM sequencing, and Illumina MiSeq sequencing were mapped to the reference Influenza A H1NI segment 5 [NC_002019.1], using CLC Genomics Workbench version 6.0.4 at default parameters, whereas in B) , they were mapped against segment 8 [NC_002020.1]. In B) , no Roche-454 reads mapped to the reference. In both A) and B) , coordinates of reference genome segment are displayed along the top and G/C content is graphed below reference in pink. A closer examination of the read mappings for both the PGM and MiSeq platforms indicated that some mapped reads were spurious. In order to eliminate these matches, stringency of read mapping parameters was increased. At the medium stringency setting, whereby the length of the read required to match was increased from 50% to 70%, the number of mapped reads dropped by roughly half for the PGM and by roughly 10% for the MiSeq. Furthermore, the number of MiSeq reads that mapped in pairs was roughly 25% (66 out of 269; Table 2 ). This is most likely due to polymorphisms in the sequenced strain that are not present in the NCBI reference genome. As expected, no reads mapped at the highest level of stringency tested, as the requirement for 100% identity would preclude any matches. For Influenza virus segment 8, the MiSeq yielded 14 mapped reads versus the single mapped read from the PGM. It is noteworthy that although the MiSeq reads covered more of the segment overall than the single PGM read did, both platforms produced a single read in an area of high G/C content. Given that the MiSeq consistently produced the highest depth of coverage, we next examined whether the increased coverage of the MiSeq was sufficient to determine the presence of a single nucleotide polymorphism (SNP) that conveys resistance to a widely used antiviral drug, oseltamavir (Ose). Previous studies have established that Influenza A strains from the 2009 epidemic had mutated to an Ose-resistant phenotype [ 15 ]. Thus, we examined the reads resulting from MiSeq sequencing of a 2009 epidemic strain for the presence of the mutation H275Y, and in neither of two biological replicates were any of the known SNPs detected. In fact, there was zero coverage of that specific region of the genome. Given that our experimental model, human blood, represents a relatively low-complexity sample (as compared to clinical samples from other body sites such as oropharyngeal swabs or fecal samples), spiked with Influenza A genomes at titers higher than would be expected to be encountered in a clinical sample, these data suggest that pathogen enrichment may be a necessary step for determination of drug resistance by metagenome sequencing. de novo assembly and BLASTn analysis of Influenza A reads In many cases the identity of a pathogen may not be known or the pathogen's genome may be very divergent at the nucleic acid level from that of known family members, necessitating analysis by means other than mapping reads to a reference genome. Therefore, de novo assemblies of the Influenza A metagenomes from the PGM and MiSeq data were performed, followed by BLASTn analysis. Assembly of the PGM reads using MIRA produced an average of 28,000 contigs, yet not a single contig was classified as Influenza A (Table 3 ). In fact, no contig was classified as a viral taxon. This result is surprising given the observed sequence depth in several regions of the Influenza A genome (see Figure 2 A). We also attempted assembly of the PGM reads using Velvet with similar assembly parameters. Again, none of the resultant contigs were able to be assigned to Influenza A. Upon closer inspection, it was observed that the identified reads were either substrings of each other and therefore absorbed during the merge step of assembly, or they had a stretch of mismatches long enough (≥ 10 bases) to cause the assembly algorithm to mark these reads as not overlapping. Table 3 De novo assembly and BLASTn statistics of Influenza A sequence reads Platform Assembler # total contigs # specific contigs PGM 1 MIRA 26,171 0 PGM 2 MIRA 30,719 0 MiSeq 1 Velvet 841,384 18 MiSeq 2 Velvet 768,912 11 By comparison, assembly of the MiSeq reads for two biological replicates using Velvet generated well over 750,000 contigs for both replicates, with 18 and 11 flu-specific contigs, respectively. Taxonomic classification of the contigs by MEGAN properly assigned the correct strain (Influenza A virus (A/swine/Iowa/15/1930(H1N1)) in 5 out of 18 and 0 out of 11 contigs. Importantly, only the strain level classifications were mis-assigned. In each case, the correct genus and species calls were made. Therefore, each of the contigs identified by MEGAN as Influenza but not correctly identified at strain level was individually subjected to BLASTn analysis and manual curation of results. This inspection revealed that the proper strain was indeed present in the hit tables (6 and 7 contigs, respectively), but that for each of these contigs, there were several strains hit with the identical E-value, including A/swine/Iowa/15/1930(H1N1), and therefore, a definitive strain call could not be made in those cases. Thus, for one of two replicates, a majority of the influenza-specific contigs resulting from de novo assembly of the MiSeq reads were either properly classified to the strain level, or conserved amongst a subset of strains including A/swine/Iowa/15/1930(H1N1). It was interesting to note that, in each case, contigs corresponding to Segments 4 (HA) and 6 (NA) were not assigned to the proper strain. These results demonstrate that the increased throughput of the MiSeq enables species level identification of reads present in proportions as low as .0008% of the total. Comparison of Influenza A sequencing replicates on the PGM and MiSeq We were interested in further examining the observed differences in the number of mapped reads between biological replicates and between genome segments. In addition, we wished to determine the reproducibility of our results. The biological replicates (derived from independent spikings and library preparations) sequenced on the MiSeq platform exhibited considerable variability (Table 2 ), and we were interested in determining at what stage this variability would arise- at library construction or sequencing. Thus, we performed replicate sequencing runs for each of the PGM and MiSeq libraries. The total number of mapped reads for each technical replicate was quite consistent (Table 2 ). This observation held true for both platforms. Moreover, each individual library yielded a number of mapped reads that was less than one standard deviation from the average. Thus, the inter-run variability seems to be limited and results of sequencing a given library preparation are highly reproducible. However, we did note several tendencies regarding the number of mapped reads per segment (Additional file 2 : Table S1 and Additional file 3 : Table S2). Firstly, segment 5 tended to be overrepresented in each PGM replicate. Specifically, for comparison, we expressed the proportion of mapped reads for a given genome segment as a function of that particular segment's length, as follows: y = # mapped reads per given segment / # total reads mapped / segment length × 1000 This value for segment 5 is more than twice that for any other segment by PGM sequencing (Additional file 3 : Table S2). This is an interesting observation given that segment 5, at 1565 bases in length, is the fifth largest segment in the genome. In every case for the PGM, segment 5 yielded the most mapped reads. This observation is consistent with our previous result with the 314 chip in which the only segment mapped was segment 5 (Additional file 1 : Figure S1). Secondly, segment 3 tended to be underrepresented, in both the PGM and MiSeq data. This trend was especially marked in the PGM libraries as compared to the MiSeq libraries (.03 and .05 respectively). These findings suggest that some sequences are inherently favored for sequencing by the PGM as well as the MiSeq. Thirdly, regardless of the number of reads mapped, the coordinates of the mapped reads in relation to the genome were widely variable (Figure 3 ). This effect was more pronounced in the PGM libraries. This suggests that the library diversity is greater than that sampled by a single sequencing run. Figure 3 Read mapping against segment 5 of Influenza A of replicate sequencing runs. Reads resulting from a replicate run of Ion Torrent PGM sequencing (top) and Illumina MiSeq sequencing (bottom) were mapped to the reference Influenza a H1NI segment 5, [NCBI accession: NC_002019 ] using CLC Genomics Workbench version 6.0 at default parameters. Coordinates of reference genome segment are displayed along the top and G/C content is graphed below reference in pink. Comparison of Roche-454, Illumina MiSeq and Ion Torrent PGM for detecting bacterial pathogens Although there have been performance comparisons of the benchtop platforms for sequencing bacterial isolates [ 16 , 17 ] to our knowledge, there has been no published report of comparisons of sensitivity and specificity in terms of speciation. Therefore, we aimed to establish a LoD for each sequencer using samples spiked with a known biothreat agent, Bacillus anthracis. Using the Sterne strain as a surrogate of anthrax infection, the quantity of bacterial cells present in a dilution series was established. Using three complementary quantitative techniques, mock samples with B. anthracis Sterne present at roughly 30,000 cells/mL of human blood were prepared. All three platforms delivered sequence reads matching the chromosome of B. anthracis Sterne by reference mapping (Table 4 ). As expected, given the increased throughput, both PGM and MiSeq yielded significantly higher numbers of reads as well as genome coverage. However, at low stringency, large numbers of reads were falsely counted as mapped. This trend was especially prominent in the PGM data. Most of the spurious reads were short (average: 40 bases) and low-complexity (i.e. rich in nucleotide repeats). This tendency was confirmed in the replicate samples. The same pattern was observed in the Roche-454 sequence reads as well. In general, the 454 reads were clustered near the 5′-end of the linear reference sequence, while the PGM and MiSeq reads spanned a much larger breadth. However, the 454 reads were uniformly longer and more information-rich. The MiSeq replicates demonstrated a similar tendency as that of the PGM; in general, the lower stringency mapping contained matched reads of short length (avg. 66 bases). For comparison, in replicate two for both platforms, 80.77% of mapped MiSeq reads mapped as non-perfect matches versus 98.83% of mapped PGM reads. The average read length of a non-perfectly matched MiSeq read was 53.22 bases, as compared to 39.99 bases for the PGM. Interestingly, the mean number of mismatches per read tended to be higher in the MiSeq data than that of the PGM (mean of 26 for MiSeq vs. 15 mismatches for PGM). Due to the slightly longer length of mapped MiSeq reads, there could conceivably be more opportunity for mismatched bases per read. Therefore, we compared the mismatches within MiSeq reads and PGM reads by expressing mean number of mismatches per read as a proportion of mean read length of non-perfectly matched reads. When expressed in this manner, the proportion of mismatches per length of non-perfectly matched reads is 0.5 for the MiSeq versus 0.4 for the PGM. Additionally, the number of indel related mismatches was found to be proportionally higher in the MiSeq reads than in the PGM reads. Conversely, the PGM reads had a higher number of A → T and G → C transversions than the MiSeq data. This is inconsistent with the known error models for both of these platforms. Table 4 Cross-platform Comparison of Read Mapping to B. anthracis Sterne Platform #Replicates Reads mapped (low) Reads mapped (med) Reads mapped (high) Fraction of reference covered a Detected by qPCR? PGM 2 29,534/15,676 7,689/4,286 247/178 .12/.09 Y Roche-454 2 384/376 249/240 65/56 .01/.01 Y MiSeq 2 10,415/41,242 3,024/9,985 1,633/7,930 .07/.19 Y a: Fraction of reference covered using high stringency mapping parameters as defined in Methods. Although de novo assemblies and taxonomic classification were attempted for both PGM and 454 reads, no dataset produced a single contig that matched Bacillus spp. This is most likely due to the lack of coverage of the bacterial genome, resulting in a lack of sufficient overlap between reads necessary to form a contig. By comparison, de novo assembly of one MiSeq run yielded 6,140,965 total contigs. Taxonomic classification of the assembled sequence reads rendered 1200 contigs that were assigned to the Bacillus cereus superfamily. More importantly, 70 contigs were properly classified as Bacillus anthracis Sterne. These results echo those of the viral samples in that assembled sequence reads from the MiSeq were able to identify the spiked-in microorganism, in this case to the strain level. Detection of genetic engineering We were interested in whether sequencing reads near the LoD could detect an instance of genetic engineering. Therefore, the strain of B. anthracis spiked into human blood for these experiments was a 34F2 strain of B. anthracis containing a stably integrated genomic copy of red fluorescent protein (RFP). Metagenomic sequence data was compared with a computer simulation of the likelihood of detecting this genetic manipulation. For this purpose, we developed a Perl-based algorithm to generate a distribution of the number of sequence reads necessary to detect a gene of 1000 bps inserted at five locations into a prokaryotic genome of 5 Mb in size. At 1000 organism-specific reads, the probability of detecting the inserted gene is roughly one-third (Figure 4 ). In fact, the minimum number of organism-specific hits necessary to ensure detection of the inserted gene is 10,000. Given that 5 of the 6 metagenomes sequenced here contained far fewer than 1000 bacillus-specific reads (Table 4 ), one would expect few, if any, sequence reads to match the RFP reference sequence. Indeed, read mapping to the available NCBI reference (EF606900.1) yielded zero matches to the RFP gene present in the 34F2 chromosome. These data suggest that increased titres, targeted enrichment, or both will be necessary for detection of gene-insertion events. Figure 4 Mathematical modeling of the likelihood of detecting a genetic modification in B. anthracis. The expected number of hits to an inserted gene of size 1 kb, at 5 copies, was simulated as a function of the number of organism-specific reads collected from the metagenomic sample. The relative size of each rectangle indicates the proportion of samplings for which a specific number of hits is expected. Discussion Historically, identification of causal agents of disease has relied heavily on one's ability to culture the organism in the laboratory and/or the use of organism-specific antibodies or sequence-based probes. However, these methods can be very limiting. For instance, some microorganisms are refractory to laboratory culture. In some cases, even microbiological assays for diseases that are manifested by cultivable organisms, such as endocarditis caused by staphylococci and streptococci, have a high false negative rate [ 18 ]. Serological assays are typically limited to identifying known or closely related organisms and antigenic drift and shift can result in false negatives. Even highly sensitive PCR-based assays must be continually updated due to signature degradation [ 19 ]. Additionally, for divergent viruses such as HIV-1, many PCR assays are unable to discriminate between closely related strains. This necessitates design of multiple probe sets, which is often a laborious task [ 20 ]. Additionally, the sensitivity of these assays is often low due to high numbers of false negatives [ 21 ]. Thus, there is a need for assays that are more robust and less pathogen specific. Prior to the widespread adoption of high-throughput sequencing (HTS), high-density oligonucleotide microarrays were used to determine the presence of microorganisms. Syndrome-specific panels showed success in diagnosis of infectious disease [ 22 ]. However, sequence features sufficiently different from the array probe will not hybridize, resulting in false negatives. By comparison, HTS represents a relatively unbiased approach to detection of causal agents of infectious disease. However, for metagenomic sequencing to be utilized in a routine clinical context would require some basic questions answered in terms of sensitivity and reproducibility. In this study, we compared three HTS platforms for their ability to detect pathogens in human blood. As compared to the traditional Roche-454 sequencer, the benchtop sequencers IonTorrent PGM and MiSeq were better able to detect a pathogen in human blood, in part by virtue of their increased throughput. Our reported LoD for viral samples on the Roche-454 of 1X10 2.5 pfu/mL is similar to but slightly lower than a previously reported value of 1X10 3 pfu/mL [ 5 ]. This may be a reflection of an increased number of reads in our study, differences in sequence library preparation or improvements in sequencing chemistry. Notably, this value is near the LoD for a validated qPCR assay [ 23 ]. Our reported LoD of 1X10 2.5 pfu/mL corresponds to an upper limit of 31,600 – 53,720 genome copies/mL. Indeed, at that titer, the sequence reads were of sufficient number and length to unequivocally discriminate between Dengue virus subtypes. However, a strain-level designation was not possible. The inability to make a strain-level call could conceivably have potential clinical consequences. For instance, a recent phylogenetic study of circulating strains of Dengue Virus 2 indicated that a single substitution on the prM 39 was responsible for fatal cases of Dengue Hemorrhagic Fever [ 24 ]. Using a different virus at a comparable titer, the Roche-454 platform was able to definitively identify the pathogen in only one of two replicates. Both the PGM and MiSeq surpassed the Roche-454 in terms of number of mapped reads as well as reproducibility. However, although the Roche-454 sequence data produced only a single hit, the length and quality of the hit were adequate to make a correct strain call (data not shown). The apparent difference in the ability of each platform to detect a given pathogen is a function of the total output per platform. When the values are expressed in a normalized manner (pathogen reads per 100,000), it becomes apparent that the analytical sensitivity of the two benchtop platforms is roughly the same and surpasses that of Roche-454 (Table 2 ). Neither technology seems to have a significant advantage regarding detection or characterization of a pathogen if the number of reads is held constant. When cost per sequencing run or per megabase is factored in, the benchtop sequencers are a more economical option. In this study, we also examined metagenomic sequencing for its ability to detect a genetically modified pathogen in a clinical sample. A BSL-2 strain of B. anthracis containing an inserted RFP gene was used as a surrogate for a genetically modified organism (GMO), and spiked into human blood at relatively high concentration. In this case, although reads were identified as likely originating from B. anthracis , in none of the samples was evidence of the inserted gene detected. This likely means that, for the time being, even with their substantial output, the benchtop sequencers are not suitable for detection of GMO from complex samples or characterization of threat agents from complex samples. Our results indicate, however, that although characterization of a given pathogen from a clinical sample by metagenome sequencing on a benchtop sequencer may not be possible without some pathogen-specific enrichment, identification of species and even strains is possible. The results of this study suggest that the benchtop sequencers perform well at the task of identifying a putative pathogen present at low titers. Each of the three platforms tested provided a number of reads that was sufficient to unambiguously identify the pathogen to species level in the case of virus and to the genus level in the case of bacteria. The data from each platform was very reproducible for technical replicates within library preparations. Indeed, the data from each platform was remarkably consistent in terms of quantity and quality (Table 2 ; Additional file 2 : Table S1). Additionally, there was little variation in the number of reads mapped or in taxonomic classification of contigs. These results suggest that library protocols and sequencing chemistries are robust and uniform enough to make a dependable identification of a given pathogen. Our results are in agreement with a recently published study in which gDNA from Bacillus anthracis was serially titrated into background DNA collected from air filters and soil. The results of this study demonstrated that, even with whole genome amplification prior to sequencing, it is difficult to assign a proper species classification to sequence reads from B. anthracis [ 25 ]. We did note several platform-specific variations in our data. For instance, in the case of the Influenza data from the PGM platform, segment 5 routinely exhibited coverage bias as compared to larger segments. There could be several possible explanations for this observation. The G/C-content of the Influenza A genome displays wide intra-segment and inter-segment variations (Additional file 4 : Figure S2). A closer examination of the mapped reads seems to show a slight correlation with areas of low G/C-content, but follow on experiments would be required to conclusively elucidate the impact of G/C-content in this context. A number of studies have noted that NGS data have distinct biases in areas of high G/C-content [ 17 , 26 , 27 ]. Additionally, template amplification via emulsion PCR is a potential source of reduced library diversity [ 28 ]. Moreover, inefficiencies during reverse transcription due to RNA secondary structure may be responsible for the observed coverage bias. Given that the initial step in PGM library construction involves random fragmentation with RNase III at 37°C, it is possible that some RNA strands did not completely unfold. This may be especially true for those segments with higher ΔG, such as segments 1, 2 and 5 (Additional file 4 : Figure S2). Although the MiSeq replicates also showed a slight skew towards segment 5, the preference was not as extreme as that for the PGM. This may be due in part to the method for RNA library construction on the MiSeq platform. The initial step involves a chemical fragmentation step at high temperature. Thus, the diversity of the libraries may be different from the start. On the other hand, the PGM replicates demonstrated some variation in mapped reads (Figure 3 ) suggesting that increased throughput might produce greater diversity. Another platform-specific characteristic observed in this study was the proportion of mismatches within a non-perfectly mapped read. This statistic was slightly higher for the MiSeq platform than for the PGM platform. Ideally, given the correct reference sequence, for a given platform this statistic would approach zero as sequencing error decreased. However, as in this case, this statistic may also be affected by nucleotide differences between the sequenced strain and the closest reference sequence available from NCBI, precluding us from making any conclusions as to sequencing error rates from these data. It is possible that, to some extent, the longer length of MiSeq reads allows for more opportunity for mapping of non-perfect matches, and this may contribute to a decreased LoD for MiSeq when mapping to a closely related but not identical reference sequence. However, the extent to which error rate versus strain level differences and read length affect this statistic as well as the LoD cannot be ascertained in the absence of a true reference sequence for the strain in question. A number of recent reports have attempted to define the limitations of metagenomic sequencing data. One study made use of simulated data sets to compare assemblies from three sequencing technologies (Sanger, pyrosequencing and Illumina). Unsurprisingly, the study concluded that assembly quality decreased rapidly with increasing sample complexity. For low complexity samples (10 genomes) the assemblies were comparable in quality and inclusiveness, while Illumina data produced superior assemblies in a higher complexity sample (100 genomes) [ 29 ]. These results mirror those presented here in that no one sequencing chemistry clearly surpassed another in terms of identification of a microorganism present in a low-complexity sample. It should be noted that the Illumina data in the study by Mende et al. were from the HiSeq platform and were extensively trimmed to provide high quality reads as input [ 28 ]. A separate report estimated that genome coverage of 20X was required for proper taxonomic classification of species present in a given metagenomic community [ 30 ]. This study is in agreement with our inability to make a correct species-level determination from our Bacillus anthracis samples. Additionally, our results complement an important conclusion from the previous report - that the efficiency of gene detection is most likely overestimated. There has been much effort to understand and improve metagenomic data from complex samples comprised mostly of bacterial species. There are fewer published studies examining the effects of different sequence technologies on viral metagenomics. A recent research paper attempted a comparison of Roche-454 and Illumina data for estimation of diversity in viral quasispecies, in this case HIV. The authors of that study noted that the increased throughout and lower error rate of the Illumina platform enabled improved reconstruction of viral haplotypes. However, due to the longer read lengths, the Roche-454 was superior when long range reconstruction was necessary [ 31 ]. Our results indicate that library diversity and overall throughput are the two key metrics in determining how a researcher or clinician may use metagenomic data from benchtop sequencers. For instance, a recent survey of Human papillomavirus DNA present in human skin tumors demonstrated that the increased throughput of the PGM enabled identification of seven additional viral subtypes as compared to data from Roche-454 sequencing of the same samples [ 32 ]. This result is similar to our observation that PGM data provided steady, reproducible identification of Influenza A virus in comparison to sequence data produced by the Roche-454. Overall, the MiSeq proved superior to both the Ion Torrent PGM and Roche-454 for both detection as well as classification of the pathogen present in our mock samples. Although there is no published report of this, it remains formally possible to identify a previously unknown agent from a single novel microbial read present in a complex metagenomic sample. Indeed, identification of novel agents has been reported with as few as 14 reads out of over 100,000 [ 33 ]. Whereas identification of an agent may require detection of only one or more reads, characterization , the crucial next step, is absolutely dependent on complete (100%) or nearly complete representation of the agent's entire genome at adequate depth of coverage, especially in the case of RNA viruses or other microorganisms likely to exhibit functionally relevant minority populations or quasispecies, or genetically modified organisms. In this case, it is necessary for follow-on experiments to more fully characterize the genome of the microorganism, such as Sanger sequencing using primers based on the novel fragment(s). It would be optimal if some of the original sample were available for such experiments. However, in many cases, the original sample may be precious or limited in terms of volume. This challenge can be more pronounced when identifying viral agents as opposed to bacterial agents. Viral genomes are orders of magnitude smaller (~1X10 4 -1X10 5 bps) than those of an average bacterial agent (~3-5X10 6 bp). Thus, the overall amount of viral nucleic acid may be in the picogram range, increasing the likelihood of two technical obstacles: 1) viral nucleic acid is outcompeted during amplification by other nucleic acids in the matrix, such as host ribosomal RNA if the matrix is tissue or, 2) if the overall amount of nucleic acids in the metagenome sample itself is low, then library preparation of the sample may fail as successive losses of genetic material occur in each step. Targeted amplification of organism-specific regions prior to sequencing has shown some promise. For instance, an assay in which multiplex PCR preceded sequencing was able to fully differentiate Bacillus anthracis, Yersinia pestis and Francisella tularensis [ 34 ]. Conclusions In this study, we sought to determine empirical limits of detection for metagenomic sequencing of clinical samples using three different sequencing platforms. We found that the analytical sensitivity of all three platforms approaches that of standard qPCR assays. Although all platforms were able to detect pathogens at the levels tested, there were several noteworthy differences. The Roche-454 Titanium platform produced consistently longer reads, even when compared with the latest chemistry updates for the PGM platform. The MiSeq platform produced consistently greater depth and breadth of coverage, while the Ion Torrent was unequaled for speed of sequencing. None of the platforms were able to verify a single nucleotide polymorphism responsible for antiviral resistance in an Influenza A strain isolated from the 2009 H1N1 pandemic. Additionally none of the platforms tested was able to detect evidence of genetic engineering in a bacterial biowarfare agent that was spiked into a clinical-type sample. Overall, the benchtop platforms perform well for identification of pathogens from a representative clinical sample. However, our results indicate that, unlike identification, characterization of pathogens is likely to require higher titers, multiples libraries and/or multiple sequencing runs. Methods Cells and viruses Bacillus anthracis [34F2 (NCBI taxon ID #526966)] used in this study was routinely cultured in Brain Heart Infusion (BHI) broth (KD Medical; Columbia, MD). Cell counts were performed using the track dilution method [ 35 ] on BHI agar plates. One mL aliquots of plaque-purified Dengue virus Type 1 and Type 2 were maintained at -80° until spiking blood. Influenza A virus stocks were purchased from ATCC (Manassas, VA): VR-1683(NCBI taxon ID # 380342) and VR-1736 (NCBI taxon ID # 710659). Sample preparation (i) Viral samples Frozen aliquots of plaque-purified Dengue virus Type 1, Dengue virus Type 2, or Swine Influenza A were ten-fold serially diluted in sterile, 1X PBS. One mL aliquots of sodium citrate-treated whole human blood (BioReclamation, Liverpool, NY) were spiked with 100 μL of diluted virus at various titers. After thorough mixing with a micropipette, Trizol LS (Life Technologies, Grand Island, NY) was added to the sample at a ratio of 3:1. Samples were processed for total RNA according to manufacturers' protocol. (ii) Bacterial samples Log-phase cultures of B. anthracis strain 34F2^RFP were ten-fold serially diluted in 1X sterile PBS. Two mL of sodium citrate-treated whole blood were spiked with 200 μL of diluted culture at various titers. After thorough mixing with a micropipette, genomic DNA (gDNA) was extracted with a BioStic® Bacteremia DNA Isolation kit (Mobio; Carlsbad, CA) according to manufacturers' protocol. Nucleic acid quality checks RNA integrity and purity were assayed using an RNA 6000 Pico chip on the Agilent Bioanalyzer™. RNA mass was determined by fluorescent detection using Qubit® Broad Range RNA kit (Life Technologies). All RNA samples were stored at -80°C until use. DNA integrity and purity were assessed by agarose gel electrophoresis on 0.8% E-Gels® (Life Technologies). DNA mass was determined using Qubit® Broad Range DNA kit (Life Technologies). All DNA samples were stored at -20°C until use. High-throughput sequencing (i) Roche-454 pyrosequencing For RNA samples, libraries were constructed by following the cDNA Rapid protocol (Roche Diagnostics, Mannheim Germany) using an input of 200 ng of total RNA. For DNA samples, libraries were constructed following the Rapid DNA protocol (Roche) using an input of 500 ng of gDNA. All emPCR reactions were performed using GS FLX Titanium Lib-L-LV kits. Template-to-bead ratios were optimized via titration. All sequencing runs were performed using a two-region gasket for each pico-titre plate. Each run was 200 cycles. (ii) Ion Torrent PGM® sequencing For RNA samples, libraries were constructed by following the Whole Transcriptome Library protocol using an input of 500 ng total RNA. For DNA samples, libraries were constructed using the IonXpress™ Plus gDNA Fragment Library protocol using an input of 500 ng gDNA. Quality control of all libraries was performed on the Agilent Bioanalyzer using a High Sensitivity chip. Library templates were clonally amplified using the Ion One Touch 2™, following the manufacturers' protocol. Template dilutions were calculated by extrapolation from a qPCR standard curve using the Ion Library Quantitation Kit (Life Technologies). Recovered template-positive Ion sphere particles (ISPs) were subjected to enrichment according to template corresponding protocol. Ion Sphere quality control was performed on enriched and un-enriched template ISPs using the Ion Sphere™ Quality Control Kit (Life Technologies). Samples containing an optimum number of template ISPs and satisfactory enrichment were subjected to the standard Ion PGM™ 200 Sequencing v2 protocol. (iii) Illumina MiSeq sequencing Illumina TruSeq cDNA libraries were prepared from 400 ng total RNA, omitting the polyA selection step. Each library was subjected to a full MiSeq run using a 300 cycle kit, paired end sequencing. A quality control tool for high throughput sequence, FASTQC, a java stand-alone program, was downloaded from Babraham Bioinformatics Institute: http://www.bioinformatics.babraham.ac.uk/projects/fastqc/ and each fastq file was checked for quality. Quantitative PCR (i) RNA samples Detection and quantification of Dengue virus and Influenza A virus in samples was performed with SYBR® Green-based real time RT-PCR. Equal masses of total RNA from each sample were analyzed in duplicate using Express One SYBR® GreenER™ (Life Technologies) following manufacturer's instructions. Specificity of primer pairs (Additional file 5 : Table S3) for each virus was checked using NCBI Primer BLAST against the nr database. Standard curves for Dengue virus were constructed using ten-fold serial dilutions of viral RNA extracted from a stock aliquot and are expressed in pfu/mL. Standard curves for Influenza A virus were constructed using a 250 base RNA oligo (Bio-Synthesis Inc., Lewisville, TX) representing the region of the gene to be amplified. Influenza A values are expressed in genome copies/mL. (ii) DNA samples Detection and quantification of Bacillus anthracis in samples was performed using SYBR® Green-based real time PCR. Equal masses of gDNA were analyzed in duplicate using Power SYBR® Green (Applied Biosystems, Grand Island, NY). Specificity of primer pairs (Additional file 5 : Table S3) for each gene was checked using NCBI Primer BLAST against the nr database. Standard curves for each gene were constructed using ten-fold serial dilutions of a plasmid construct containing the complete coding sequence of each gene. Values are expressed in copies/mL. Bioinformatics (i) Reference mapping Sequence data mapping was performed using CLC Genomics Workbench v6.0.4 (CLC Inc, Aarhus, Denmark). CLC Reference Mapper was run with default settings (Insertion cost = 3, Deletion cost = 3, Mismatch cost = 2, Length fraction = 0.5 and similarity fraction = 0.8). CLC default settings were arbitrarily assigned as low stringency settings. Medium stringency settings were arbitrarily defined as Insertion cost = 3, Deletion cost = 3, Mismatch cost = 2, Length fraction = 0.7 and similarity fraction = 0.8. High stringency settings were defined as Insertion cost = 3, Deletion cost = 3, Mismatch cost = 3, Length fraction = 1.0 and similarity fraction = 1.0. GS Reference Mapper software (v2.0.01.14; Roche/454) was used to produce reference-guided assemblies of each of the Roche datasets with respect to the DENV-1 genome (GenBank; DVU88536). (ii) de novo assembly and BLASTn analysis of contigs MIRA V3.4.0.1 production version was used for the assembly of PGM data. The SFF file was converted to fastq using the sff_extract.py utility. The job parameter was modified with the keywords "denovo,genome,accurate,iontor" to indicate a de novo assembly for the corresponding platform. In addition, the modifier "-GE:not" was set to 12 processors for parallel computing. The assembly parameters for MIRA for the PGM data were, minimum overlap = 17 bps, minimum contig size =150 bps, minimum neighbor quality needed for tagging = 20, minimum read length =80. For the Velvet assembly the k-mer size was 33. The assembly of the MiSeq data was conducted using velvet version 1.2.1. The velvet assembly was conducted using a k-mer of size 31 and "short" modifier. For the downstream analysis, BLASTN and annotation, only those contigs with size greater than 150 bps were considered. BLASTN was conducted with the NCBI-BLAST++ version 2.2.28 against the NT BLAST database from 04/22/2013. The taxonomical annotation was obtained with an in-house script and the results were visualized using MEGAN version 4.70.4. (iii) Mathematical modeling Given a particular genome (G) of size (N), we were interested in a particular gene (H) of size (M) within G. If a sequencing experiment is conducted and a particular read from the experiment matches the sequence of H, this is labeled a hit. In order to examine the range of read counts that will result in a range from one hit with low probability to one hit with near certainty, we decided to recreate the parameters of a sequencing experiment in silico . The simulation was written in PERL, where N was set to 5.23 Mbps, M was set to 1000 bps, and five copies of H were used in the simulation. The simulation assumes that all parts of G are equally likely to be sequenced. A thousand iterations of the simulation produced the expected number of hits for read counts of 1 k, 2 k, 3 k, 4 k, 5 k, 8 k, 10 k and 15 k. Cells and viruses Bacillus anthracis [34F2 (NCBI taxon ID #526966)] used in this study was routinely cultured in Brain Heart Infusion (BHI) broth (KD Medical; Columbia, MD). Cell counts were performed using the track dilution method [ 35 ] on BHI agar plates. One mL aliquots of plaque-purified Dengue virus Type 1 and Type 2 were maintained at -80° until spiking blood. Influenza A virus stocks were purchased from ATCC (Manassas, VA): VR-1683(NCBI taxon ID # 380342) and VR-1736 (NCBI taxon ID # 710659). Sample preparation (i) Viral samples Frozen aliquots of plaque-purified Dengue virus Type 1, Dengue virus Type 2, or Swine Influenza A were ten-fold serially diluted in sterile, 1X PBS. One mL aliquots of sodium citrate-treated whole human blood (BioReclamation, Liverpool, NY) were spiked with 100 μL of diluted virus at various titers. After thorough mixing with a micropipette, Trizol LS (Life Technologies, Grand Island, NY) was added to the sample at a ratio of 3:1. Samples were processed for total RNA according to manufacturers' protocol. (ii) Bacterial samples Log-phase cultures of B. anthracis strain 34F2^RFP were ten-fold serially diluted in 1X sterile PBS. Two mL of sodium citrate-treated whole blood were spiked with 200 μL of diluted culture at various titers. After thorough mixing with a micropipette, genomic DNA (gDNA) was extracted with a BioStic® Bacteremia DNA Isolation kit (Mobio; Carlsbad, CA) according to manufacturers' protocol. (i) Viral samples Frozen aliquots of plaque-purified Dengue virus Type 1, Dengue virus Type 2, or Swine Influenza A were ten-fold serially diluted in sterile, 1X PBS. One mL aliquots of sodium citrate-treated whole human blood (BioReclamation, Liverpool, NY) were spiked with 100 μL of diluted virus at various titers. After thorough mixing with a micropipette, Trizol LS (Life Technologies, Grand Island, NY) was added to the sample at a ratio of 3:1. Samples were processed for total RNA according to manufacturers' protocol. (ii) Bacterial samples Log-phase cultures of B. anthracis strain 34F2^RFP were ten-fold serially diluted in 1X sterile PBS. Two mL of sodium citrate-treated whole blood were spiked with 200 μL of diluted culture at various titers. After thorough mixing with a micropipette, genomic DNA (gDNA) was extracted with a BioStic® Bacteremia DNA Isolation kit (Mobio; Carlsbad, CA) according to manufacturers' protocol. Nucleic acid quality checks RNA integrity and purity were assayed using an RNA 6000 Pico chip on the Agilent Bioanalyzer™. RNA mass was determined by fluorescent detection using Qubit® Broad Range RNA kit (Life Technologies). All RNA samples were stored at -80°C until use. DNA integrity and purity were assessed by agarose gel electrophoresis on 0.8% E-Gels® (Life Technologies). DNA mass was determined using Qubit® Broad Range DNA kit (Life Technologies). All DNA samples were stored at -20°C until use. High-throughput sequencing (i) Roche-454 pyrosequencing For RNA samples, libraries were constructed by following the cDNA Rapid protocol (Roche Diagnostics, Mannheim Germany) using an input of 200 ng of total RNA. For DNA samples, libraries were constructed following the Rapid DNA protocol (Roche) using an input of 500 ng of gDNA. All emPCR reactions were performed using GS FLX Titanium Lib-L-LV kits. Template-to-bead ratios were optimized via titration. All sequencing runs were performed using a two-region gasket for each pico-titre plate. Each run was 200 cycles. (ii) Ion Torrent PGM® sequencing For RNA samples, libraries were constructed by following the Whole Transcriptome Library protocol using an input of 500 ng total RNA. For DNA samples, libraries were constructed using the IonXpress™ Plus gDNA Fragment Library protocol using an input of 500 ng gDNA. Quality control of all libraries was performed on the Agilent Bioanalyzer using a High Sensitivity chip. Library templates were clonally amplified using the Ion One Touch 2™, following the manufacturers' protocol. Template dilutions were calculated by extrapolation from a qPCR standard curve using the Ion Library Quantitation Kit (Life Technologies). Recovered template-positive Ion sphere particles (ISPs) were subjected to enrichment according to template corresponding protocol. Ion Sphere quality control was performed on enriched and un-enriched template ISPs using the Ion Sphere™ Quality Control Kit (Life Technologies). Samples containing an optimum number of template ISPs and satisfactory enrichment were subjected to the standard Ion PGM™ 200 Sequencing v2 protocol. (iii) Illumina MiSeq sequencing Illumina TruSeq cDNA libraries were prepared from 400 ng total RNA, omitting the polyA selection step. Each library was subjected to a full MiSeq run using a 300 cycle kit, paired end sequencing. A quality control tool for high throughput sequence, FASTQC, a java stand-alone program, was downloaded from Babraham Bioinformatics Institute: http://www.bioinformatics.babraham.ac.uk/projects/fastqc/ and each fastq file was checked for quality. (i) Roche-454 pyrosequencing For RNA samples, libraries were constructed by following the cDNA Rapid protocol (Roche Diagnostics, Mannheim Germany) using an input of 200 ng of total RNA. For DNA samples, libraries were constructed following the Rapid DNA protocol (Roche) using an input of 500 ng of gDNA. All emPCR reactions were performed using GS FLX Titanium Lib-L-LV kits. Template-to-bead ratios were optimized via titration. All sequencing runs were performed using a two-region gasket for each pico-titre plate. Each run was 200 cycles. (ii) Ion Torrent PGM® sequencing For RNA samples, libraries were constructed by following the Whole Transcriptome Library protocol using an input of 500 ng total RNA. For DNA samples, libraries were constructed using the IonXpress™ Plus gDNA Fragment Library protocol using an input of 500 ng gDNA. Quality control of all libraries was performed on the Agilent Bioanalyzer using a High Sensitivity chip. Library templates were clonally amplified using the Ion One Touch 2™, following the manufacturers' protocol. Template dilutions were calculated by extrapolation from a qPCR standard curve using the Ion Library Quantitation Kit (Life Technologies). Recovered template-positive Ion sphere particles (ISPs) were subjected to enrichment according to template corresponding protocol. Ion Sphere quality control was performed on enriched and un-enriched template ISPs using the Ion Sphere™ Quality Control Kit (Life Technologies). Samples containing an optimum number of template ISPs and satisfactory enrichment were subjected to the standard Ion PGM™ 200 Sequencing v2 protocol. (iii) Illumina MiSeq sequencing Illumina TruSeq cDNA libraries were prepared from 400 ng total RNA, omitting the polyA selection step. Each library was subjected to a full MiSeq run using a 300 cycle kit, paired end sequencing. A quality control tool for high throughput sequence, FASTQC, a java stand-alone program, was downloaded from Babraham Bioinformatics Institute: http://www.bioinformatics.babraham.ac.uk/projects/fastqc/ and each fastq file was checked for quality. Quantitative PCR (i) RNA samples Detection and quantification of Dengue virus and Influenza A virus in samples was performed with SYBR® Green-based real time RT-PCR. Equal masses of total RNA from each sample were analyzed in duplicate using Express One SYBR® GreenER™ (Life Technologies) following manufacturer's instructions. Specificity of primer pairs (Additional file 5 : Table S3) for each virus was checked using NCBI Primer BLAST against the nr database. Standard curves for Dengue virus were constructed using ten-fold serial dilutions of viral RNA extracted from a stock aliquot and are expressed in pfu/mL. Standard curves for Influenza A virus were constructed using a 250 base RNA oligo (Bio-Synthesis Inc., Lewisville, TX) representing the region of the gene to be amplified. Influenza A values are expressed in genome copies/mL. (ii) DNA samples Detection and quantification of Bacillus anthracis in samples was performed using SYBR® Green-based real time PCR. Equal masses of gDNA were analyzed in duplicate using Power SYBR® Green (Applied Biosystems, Grand Island, NY). Specificity of primer pairs (Additional file 5 : Table S3) for each gene was checked using NCBI Primer BLAST against the nr database. Standard curves for each gene were constructed using ten-fold serial dilutions of a plasmid construct containing the complete coding sequence of each gene. Values are expressed in copies/mL. (i) RNA samples Detection and quantification of Dengue virus and Influenza A virus in samples was performed with SYBR® Green-based real time RT-PCR. Equal masses of total RNA from each sample were analyzed in duplicate using Express One SYBR® GreenER™ (Life Technologies) following manufacturer's instructions. Specificity of primer pairs (Additional file 5 : Table S3) for each virus was checked using NCBI Primer BLAST against the nr database. Standard curves for Dengue virus were constructed using ten-fold serial dilutions of viral RNA extracted from a stock aliquot and are expressed in pfu/mL. Standard curves for Influenza A virus were constructed using a 250 base RNA oligo (Bio-Synthesis Inc., Lewisville, TX) representing the region of the gene to be amplified. Influenza A values are expressed in genome copies/mL. (ii) DNA samples Detection and quantification of Bacillus anthracis in samples was performed using SYBR® Green-based real time PCR. Equal masses of gDNA were analyzed in duplicate using Power SYBR® Green (Applied Biosystems, Grand Island, NY). Specificity of primer pairs (Additional file 5 : Table S3) for each gene was checked using NCBI Primer BLAST against the nr database. Standard curves for each gene were constructed using ten-fold serial dilutions of a plasmid construct containing the complete coding sequence of each gene. Values are expressed in copies/mL. Bioinformatics (i) Reference mapping Sequence data mapping was performed using CLC Genomics Workbench v6.0.4 (CLC Inc, Aarhus, Denmark). CLC Reference Mapper was run with default settings (Insertion cost = 3, Deletion cost = 3, Mismatch cost = 2, Length fraction = 0.5 and similarity fraction = 0.8). CLC default settings were arbitrarily assigned as low stringency settings. Medium stringency settings were arbitrarily defined as Insertion cost = 3, Deletion cost = 3, Mismatch cost = 2, Length fraction = 0.7 and similarity fraction = 0.8. High stringency settings were defined as Insertion cost = 3, Deletion cost = 3, Mismatch cost = 3, Length fraction = 1.0 and similarity fraction = 1.0. GS Reference Mapper software (v2.0.01.14; Roche/454) was used to produce reference-guided assemblies of each of the Roche datasets with respect to the DENV-1 genome (GenBank; DVU88536). (ii) de novo assembly and BLASTn analysis of contigs MIRA V3.4.0.1 production version was used for the assembly of PGM data. The SFF file was converted to fastq using the sff_extract.py utility. The job parameter was modified with the keywords "denovo,genome,accurate,iontor" to indicate a de novo assembly for the corresponding platform. In addition, the modifier "-GE:not" was set to 12 processors for parallel computing. The assembly parameters for MIRA for the PGM data were, minimum overlap = 17 bps, minimum contig size =150 bps, minimum neighbor quality needed for tagging = 20, minimum read length =80. For the Velvet assembly the k-mer size was 33. The assembly of the MiSeq data was conducted using velvet version 1.2.1. The velvet assembly was conducted using a k-mer of size 31 and "short" modifier. For the downstream analysis, BLASTN and annotation, only those contigs with size greater than 150 bps were considered. BLASTN was conducted with the NCBI-BLAST++ version 2.2.28 against the NT BLAST database from 04/22/2013. The taxonomical annotation was obtained with an in-house script and the results were visualized using MEGAN version 4.70.4. (iii) Mathematical modeling Given a particular genome (G) of size (N), we were interested in a particular gene (H) of size (M) within G. If a sequencing experiment is conducted and a particular read from the experiment matches the sequence of H, this is labeled a hit. In order to examine the range of read counts that will result in a range from one hit with low probability to one hit with near certainty, we decided to recreate the parameters of a sequencing experiment in silico . The simulation was written in PERL, where N was set to 5.23 Mbps, M was set to 1000 bps, and five copies of H were used in the simulation. The simulation assumes that all parts of G are equally likely to be sequenced. A thousand iterations of the simulation produced the expected number of hits for read counts of 1 k, 2 k, 3 k, 4 k, 5 k, 8 k, 10 k and 15 k. (i) Reference mapping Sequence data mapping was performed using CLC Genomics Workbench v6.0.4 (CLC Inc, Aarhus, Denmark). CLC Reference Mapper was run with default settings (Insertion cost = 3, Deletion cost = 3, Mismatch cost = 2, Length fraction = 0.5 and similarity fraction = 0.8). CLC default settings were arbitrarily assigned as low stringency settings. Medium stringency settings were arbitrarily defined as Insertion cost = 3, Deletion cost = 3, Mismatch cost = 2, Length fraction = 0.7 and similarity fraction = 0.8. High stringency settings were defined as Insertion cost = 3, Deletion cost = 3, Mismatch cost = 3, Length fraction = 1.0 and similarity fraction = 1.0. GS Reference Mapper software (v2.0.01.14; Roche/454) was used to produce reference-guided assemblies of each of the Roche datasets with respect to the DENV-1 genome (GenBank; DVU88536). (ii) de novo assembly and BLASTn analysis of contigs MIRA V3.4.0.1 production version was used for the assembly of PGM data. The SFF file was converted to fastq using the sff_extract.py utility. The job parameter was modified with the keywords "denovo,genome,accurate,iontor" to indicate a de novo assembly for the corresponding platform. In addition, the modifier "-GE:not" was set to 12 processors for parallel computing. The assembly parameters for MIRA for the PGM data were, minimum overlap = 17 bps, minimum contig size =150 bps, minimum neighbor quality needed for tagging = 20, minimum read length =80. For the Velvet assembly the k-mer size was 33. The assembly of the MiSeq data was conducted using velvet version 1.2.1. The velvet assembly was conducted using a k-mer of size 31 and "short" modifier. For the downstream analysis, BLASTN and annotation, only those contigs with size greater than 150 bps were considered. BLASTN was conducted with the NCBI-BLAST++ version 2.2.28 against the NT BLAST database from 04/22/2013. The taxonomical annotation was obtained with an in-house script and the results were visualized using MEGAN version 4.70.4. (iii) Mathematical modeling Given a particular genome (G) of size (N), we were interested in a particular gene (H) of size (M) within G. If a sequencing experiment is conducted and a particular read from the experiment matches the sequence of H, this is labeled a hit. In order to examine the range of read counts that will result in a range from one hit with low probability to one hit with near certainty, we decided to recreate the parameters of a sequencing experiment in silico . The simulation was written in PERL, where N was set to 5.23 Mbps, M was set to 1000 bps, and five copies of H were used in the simulation. The simulation assumes that all parts of G are equally likely to be sequenced. A thousand iterations of the simulation produced the expected number of hits for read counts of 1 k, 2 k, 3 k, 4 k, 5 k, 8 k, 10 k and 15 k. Ethics No research regarding human or animal subjects was conducted at NMRC or USUHS as part of this work. Human blood was obtained commercially as a laboratory reagent (and in the absence of all identifiers) from BioreclamationIVT ( http://www.bioreclamation.com ) and therefore did not require institutional review by an ethics committee at NMRC. Bioreclamation collects blood from healthy, consented, paid human donors. Competing interests The authors declare no competing financial or non-financial interests. Authors' contributions KGF and KABL conceived of and designed the study. KGF and SLS prepared the samples. KGF, CLR and TL performed the sequencing. KGF, JEHG and KABL analyzed the data. JEHG performed the de novo assemblies and taxonomic classification of the contigs. KGF and KABL wrote the manuscript with input and edits from VM, AJM and JEHG. All authors read and approved the final manuscript. Supplementary Material Additional file 1: Figure S1 Read mapping against segment 5 of Influenza genome. Reads resulting from Ion Torrent 314 chip were mapped to the reference Influenza a H1NI segment 5 [NCBI accession: NC_002019], using CLC Genomics Workbench version 6.0 at default parameters. Coordinates of reference genome segment are displayed along the top and G/C content is graphed below reference in pink. Click here for file Additional file 2: Table S1 Mapped reads by Influenza A segment for MiSeq and PGM replicates a . a: Statistics for one of two independent libraries at low stringency parameters. Click here for file Additional file 3: Table S2 Proportion of mapped reads as a function of Influenza A genome segment size for MiSeq and PGM replicates a . a: Statistics for one of two independent libraries at low stringency parameters. Click here for file Additional file 4: Figure S2 Predicted free energy of individual Influenza genome segments. Using CLC Genomics Workbench version 6.0, Gibb's free energy was predicted for each genome segment of the NCBI reference strain Influenza A virus (A/Puerto Rico/8/34(H1N1)). Click here for file Additional file 5: Table S3 Sequences of primers used in this study. Click here for file Acknowledgements The authors wish to thank Dr. Raviprakesh Kanakette for providing the Dengue virus stocks and Dr. Shanmuga Sozhamannan and Stephanie Servetas for providing B. anthracis 34F2^RFP. The authors also wish to express our gratitude to Tom Slezak for critical reading of the manuscript. These studies were supported by Defense Threat Reduction Agency project CBM.DIAGB.03.10.NM to KABL. VM is a military service member and this work was prepared as part of his official duties. Title 17 U.S.C. §105 provides that 'Copyright protection under this title is not available for any work of the United States Government.' Title 17 U.S.C. §101 defines a U.S. Government work as a work prepared by a military service member or employee of the U.S. Government as part of that person's official duties. The opinions or assertions contained herein are the private ones of the author(s) and are not to be construed as official or reflecting the views of either the Department of the Navy or the Department of Defense.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7157734/

Rapid and Sensitive Multiplex Assay for the Detection of B. anthracis Spores from Environmental Samples

Prompt and accurate detection of Bacillus anthracis spores is crucial in the event of intentional spore dissemination in order to reduce the number of expected casualties. Specific identification of these spores from environmental samples is both challenging and time-consuming. This is due to the high homology with other Bacillus species as well as the complex composition of environmental samples, which further impedes assay sensitivity. Previously, we showed that a short incubation of B.anthracis spores in a defined growth medium results in rapid germination, bacterial growth, and secretion of toxins, including protective antigen. In this work, we tested whether coupling the incubation process to a newly developed immune-assay will enable the detection of secreted toxins as markers for the presence of spores in environmental samples. The new immune assay is a flow cytometry-based multiplex that simultaneously detects a protective antigen, lethal factor, and edema factor. Our combined assay detects 1 × 10 3 –1 × 10 4 /mL spores after a 2 h incubation followed by the ~80 min immune-multiplex detection. Extending the incubation step to 5 h increased assay sensitivity to 1 × 10 2 /mL spore. The protocol was validated in various environmental samples using attenuated or fully virulent B. anthracis spores. There was no substantial influence of contaminants derived from real environmental samples on the performance of the assay compared to clean samples, which allow the unequivocal detection of 3 × 10 3 /mL and 3 × 10 2 /mL spores following 2 and 5 hour's incubation, respectively. Overall, we propose this method as a rapid, sensitive, and specific procedure for the identification of B. anthracis spores in environmental samples. 1. Introduction Bacillus anthracis (B. anthracis) , a gram-positive member of the Bacillus cereus (B. cereus) group, is the causal agent of anthrax, a serious and often fatal disease of animals and humans [ 1 , 2 ]. This pathogen is classified as a tier 1 bio-threat agent due to its highly pathogenic nature [ 3 ]. Anthrax spores are extremely stable in the environment as well as easy to disperse, making these agents extremely suitable for use as a bio-weapon. In humans, three types of anthrax disease have been recorded according to the route of infection: Cutaneous inoculation via a cut or abrasion, gastrointestinal infection resulting from the consumption of contaminated meat, and the inhalational form of the disease. Inhalational anthrax is a rare disease naturally, historically associated with industrial exposure to spores [ 2 , 4 ]. This form is also the most likely outcome of a deliberate release of aerosolized spores [ 1 , 5 ]. The virulence of B. anthracis relies on the presence of two virulence plasmids: pXO2, encoding the capsule operon, and pXO1, encoding the tripartite toxin genes. These include the protective antigen (PA), lethal factor (LF), and edema factor (EF) proteins. PA is a pore-forming protein, necessary for the translocation of the catalytic LF (metalloprotease specific for MAP kinases) and EF (calmodulin-dependent adenylate cyclase) into target mammalian cells. PA is non-toxic by itself, while LF and EF are inactive outside the host cell and depend on PA activity to function [ 1 ]. In the event of B. anthracis spores being used as a bio-weapon (such as a terror attack), a method enabling direct detection of spores in environmental samples can prove invaluable for accurate detection and risk assessment. Detection of B. anthracis in clinical or environmental samples is usually performed by applying one or more of the following methodologies; classical microbiological methods are still used as initial indications for the presence of the bacterium. These methods include gram stain, hemolysis and motility tests, capsule production, and gamma phage susceptibility [ 6 ]. Nucleic acid-based techniques, from PCR to whole genome sequencing and transcriptomics [ 7 , 8 , 9 ] are also applied. Immunoassays are widely applied for the detection of the bacterium, usually by using antibodies specific to the PA or capsule [ 10 ]. In addition, specific enzyme activity tests, usually of the LF [ 6 ], have also been developed. Though these assays are considered highly sensitive in homogeneous samples, their specificity and sensitivity are significantly hampered in complex environmental samples such as soil samples [ 11 ]. This can be related to the existence of ~2 × 10 9 background bacteria/g of soil, usually of closely related species, which are found in the top 1 m ground layer [ 12 ]. In addition, the chemical and physical composition of environmental samples is complex, unknown, and highly variable, possibly containing substances that may impede the assays usually applied for spore detection [ 10 ]. The most direct way to overcome these problems is to isolate B. anthracis by plating the sample on selective agar media. This solution is time consuming and usually results in enrichment of the desired bacteria rather than absolute purification. Alternatively, spores could be purified from the samples using affinity-based methods [ 13 , 14 , 15 , 16 , 17 , 18 , 19 ]. The detection of B. anthracis spores from environmental samples in various immune affinity-based assays has been shown to be effective for the detection of 1 × 10 5 –1 × 10 6 spores, in assays taking several hours to complete [ 11 ]. PCR based assays using specific primers and probes for the virulence plasmids were able to identify 1 × 10 5 spores/g soil using multiplex-PCR and 1 × 10 3 –1 × 10 4 spores/g soil using real-time PCR with assays that are somewhat faster, taking a few hours to complete [ 7 , 8 , 20 ]. The drawback of both PCR and antibody-based tests is the lack of ability to determine the viability of the spores. Currently, rapid viability PCR (RV-PCR) that combines qPCR with bacterial growth is the only diagnostic assay validated by the Environmental Protection Agency (EPA) for the detection of live spores. This test detects approximately 10 live spores in the presence of 1 × 10 6 autoclaved spores within 17 h [ 21 , 22 ]. In this study, we examine the possibility of detecting the spores directly from the environmental sample by a rapid homogenous growth based test. Previously, it was shown that PA could serve as a surrogate marker for the estimation of the concentration of B. anthracis bacteria in animal models [ 23 , 24 ]. Moreover, we demonstrated that a short incubation of B. anthracis spores in Dulbecco's Modified Eagle's medium (DMEM) growth media supplemented with 10% serum in the presence of CO 2 results in the rapid production and secretion of virulence factors during spore germination and initial bacterial growth, mainly PA [ 25 ]. Herein, we have applied these findings in developing a multiplexed flow cytometry (FCM) based assay to simultaneously detect PA, LF, and EF as surrogates for the presence of B. anthracis spores in environmental samples. The new assay includes two steps: An initial toxin biosynthesis step of 2–5 h, followed by a ~80 min immune-multiplex detection step of the secreted virulence factors. We demonstrate that this 4–7 h test is specific, sensitive, and can indirectly detect viable B. anthracis spores with a significant dynamic range, in environmental samples. 2. Results The proposed B. anthracis spore detection method is a two-step test where the environmental sample is cultured followed by immune-based detection of the tripartite toxin components in the growth medium. In order to achieve this goal we optimized the growth conditions to maximize toxin secretion, and developed a specific, sensitive multiplex test for the detection of the secreted toxins. 2.1. Development of a Magnetic Bead Immunoassay Based Flow Cytometry Method for the Detection of Anthrax Virulence Factors The basic principle of the assay developed in this work is depicted in Figure 1 A. In this study, three magnetic beads with different distinct fluorescence intensity profiles were used. These were coupled separately to specific monoclonal antibodies against EF, LF, or PA. In the presence of the relevant antigen, the Interaction of the antibody-coupled beads with the antigen results in a complex formation that can then be captured magnetically for washing out irrelevant proteins and particles throughout the assay. Consequently, labeling occurs by specific polyclonal IgG antibodies targeted against the antigen of interest (i.e., PA, LF, and EF). The resulting sandwich-complex is then detected through the addition of a secondary fluorescent-labeled antibody. FCM analysis allows the identification of the labeling-fluorophore in parallel to the beads' fluorescence. Figure 1 B shows a representative flow cytometry analysis of the three individual assays for the detection of EF, LF, and PA proteins separately. In order to characterize the assay sensitivity, dose-response assays were performed for the three individual antigens separately ( Figure 2 ). The mean fluorescence intensity (MFI) for the resulting individual Alexa fluor 488 stained complexes was proportional to the concentration of the antigen in the sample. The results were calculated as the ratios between the signal (S) measured for each concentration compared to the signal measured with the assay run against antigen-free PBS (noise, N). This is termed the S/N ratio. S/N ration of 2 and above is considered a positive signal. Results demonstrate a wide dynamic range with a detection threshold of 0.03 ng/mL and 0.3–1 ng/mL for PA and LF, respectively. For EF, the assay proved less sensitive, and a positive signal was obtained from a concentration of 30–100 ng/mL. The magnetic beads chosen for this assay are MegaPlex-C beads. These beads are unique as, in addition to being magnetic, they are available in a wide range of particular and distinct fluorescence intensities. These intensity level differences (hence "color-codes") allow the separation of different bead sub-populations by FCM analysis using the 633 nm laser. This feature enabled the incorporation of the three individual assays into a single, multiplexed assay. Provided that sensitivity and specificity are not compromised, this format allows rapid and simultaneous detection of PA, LF, and EF virulence proteins in the same sample. To examine the FCM ability to distinguish between the different beads in a multiplex assay, we tested whether mixed beads can be separately gated by their color-code. Antigens (PA, LF, and EF) were added to one mixture ("positive control") while the other was kept as a negative control (without antigens). Analysis of these samples showed that each individual bead population could be identified and separately gated ( Figure 3 A). Once we successfully demonstrated the ability to differentially identify discrete bead populations in multiplex format, we set out to estimate the effect of the multiplex format on the detection sensitivity of the assay. In order to assess assay sensitivity and specificity, multiplexed dose-response assays were performed with purified antigens. No significant changes in assay sensitivity ( Figure 3 B) for PA and LF were observed compared to the individual assays ( Figure 2 ). A slight decrease in sensitivity was observed for EF detection. Importantly, no cross-reactivity was observed for LF and EF proteins. A minimal amount of cross-reactivity at high PA concentrations (1 µg/mL) with EF and LF coupled beads was measured. Since our purified PA antigen contained traces of bacterial proteins [ 26 ], we assumed that these impurities were the cause of the cross-reactivity. To ascertain assay specificity, a supernatant derived after a 24 h incubation of delta EF/LF Vollum strain [ 27 ] was tested ( Supplementary Materials Figure S1 ). As hypothesized, no cross-reactivity of PA was observed with EF and LF coupled beads. 2.2. Applying the New Multiplex Assay for B. anthracis Spore Detection Thus far, we demonstrated that our new multiplex immune-assay can simultaneously detect three secreted virulence factors of B. anthracis . This assay is rapid (approximately 80 min), sensitive, and specific. Previous research indicates that PA can serve as a surrogate marker to estimate the concentration of B. anthracis [ 23 ]. In addition, we have also demonstrated that a short incubation in DMEM (supplemented with 10% serum in the presence of CO 2 ) results in germination and bacterial growth coupled with the early secretion of virulence factors [ 25 , 28 ]. Since this toxin production process is performed under batch conditions (microtiter dish), the concentration of these factors in the growth media increases over time. Therefore, potentially we can increase the toxin concentration in the growth medium by prolonging the incubation phase. Coupling this incubation process with sensitive and specific detection assays may enable spore detection in environmental samples containing low spore concentrations. Therefore, the proposed test combines sample incubation under increased toxin secretion conditions with the multiplex FCM based assay. In order to ascertain the optimal growth conditions enabling unequivocal identification of virulence factors in relation to the initial spore concentration, B. anthracis ATCC14185 spores (5 × 10 5 CFU/mL) were suspended in medium containing 10% or 50% fetal bovine serum (FBS) and incubated at 37 °C for 2 h in the presence of 5% or 10% CO 2 . Following incubation, the presence of virulence factors in the growth medium was assessed by the multiplexed assay. Results ( Figure 4 ) show detectable levels of PA following 2 h incubation while no significant detection of LF or EF was observed. Though positive detection of PA (S/N > 2) was obtained in all samples, the highest signal was reached in samples incubated in 50% FBS growth medium in 10% CO 2 . These were thus defined as the optimal conditions for increased toxins secretion. These results validate our hypothesis that PA is secreted in detectable quantities at the early stages of bacterial growth and is, therefore, suitable as a surrogate marker for spore detection. To evaluate the effect of extended incubation time on toxin secretion and assay sensitivity, different doses of ATCC14185 spores (1 × 10 3 –1 × 10 7 CFU/mL) were incubated for 2–5 h under optimal conditions. Following the incubation process, toxin presence in the growth medium was determined by the multiplexed immunoassay. Our results ( Figure 5 ) show that prolonged incubation time significantly improves the assay's sensitivity. Similarly to the previous results, PA was the first to be detected in all samples. Positive detection of PA was possible within 2 and 3 h of incubation in 1 × 10 5 and 1 × 10 3 –1 × 10 4 spores/mL, respectively. Positive detection of LF was obtained after 3 h incubation for 1 × 10 5 and 5 h for 1 × 10 4 spores/mL. EF was detected only at high spore concentrations (above 1 × 10 6 CFU/mL), after 4–5 h incubation. These results emphasize the benefit of using a dynamic process, in which samples can be collected continually along 5 h of incubation, thus increasing assay sensitivity. Since this method is supposed to detect spores in environmental samples, we tested our protocol on samples taken from asphalt-paved roads, sidewalks, and soil samples, which were spiked with known spore doses. First, we tested the effect of various environmental samples on the assay noise level in our multiplex assay. To this end, 10 environmental samples were incubated in a growth medium at 37 °C 10% CO 2 for 5 h. Following incubation, the growth medium was tested for the presence of EF/LF/PA by the multiplex assay. No positive result was obtained in any of the negative samples ( Figure 6 A). Next, we assessed the effect of environmental contaminants on B. anthracis toxins secretion during the incubation process. Environmental samples (n = 10) were spiked with 5 × 10 3 CFU/mL B. anthracis ATCC14185 spores, processed (Material and Methods), resuspended in growth medium and incubated at 37 °C 10% CO 2 for 5 h. All environmental samples ( Figure 6 ) were positive for PA (S/N range 4–25). However, the spiked environmental samples reached a lower signal (~2 fold decrease) relative to the same spore concentration spiked into PBS. This reduction may result from decreased spore dose in the final sample due to the spiking/processing protocol used, by signal inhibition resulting from the environmental sample constituents, or a combination of both. To validate these results, 40 additional environmental samples, collected in four separate locations in Israel, were spiked with 1 × 10 4 CFU/mL spores and processed. Following 5 h incubation, PA was detected in all samples, while LF was detected only in 15/40 samples (S/N range 2–3, Figure 6 B). In agreement with the above, PA levels significantly varied between samples, but all were clearly positive (S/N range 18–102, Figure 6 B), indicating that the method is indeed suitable for B. anthracis spore detection in an environmental sample. 2.3. Assay Validation Using Fully Virulent B. anthracis Vollum Strain Spores Having performed all of our experiments on B. anthracis ATCC14185 spores, we validated our results using a fully virulent Vollum strain. Vollum spores (1 × 10 2 –1 × 10 7 CFU/mL) were suspended in growth medium and incubated for 2 and 5 h. Results in Figure 7 A show positive detection of PA and LF at concentrations of 1 × 10 3 CFU/mL and 1 × 10 5 CFU/mL spores, respectively, following a 2 h incubation process. Prolonging incubation process to 5 h enables the detection of PA at concentrations of 1 × 10 2 CFU/mL and LF at concentrations of 1 × 10 3 CFU/mL. A positive signal for EF was observed only from 1 × 10 4 CFU/mL ( Figure 7 B). Positive PA detection was achieved in 14/14 different soil samples spiked with 1 × 10 3 CFU Vollum spores/g after 2 h incubation (S/N range 3.5–5.5, Figure 7 C). Prolonging incubation to 5 h ( Figure 7 D) resulted in detection of PA in soil samples spiked with 1 × 10 2 CFU spores (S/N range 17–35). LF was observed (S/N range 2.5–4) in all 1 × 10 3 CFU spores samples following 5 h. The assay proved more sensitive to the presence of Vollum spores compared to ATCC14185 spores. This could result from differences in toxin secretion between the two strains. The ATCC14185 strain is a vaccine strain, used to produce PA-containing drug substance and was mutated to facilitate this application [ 29 ]. These mutations may explain this difference. Overall, the proposed method combining biosynthesis and identification of toxins can serve as a rapid and sensitive assay for the indirect detection of viable B. anthracis spores in environmental samples. 2.1. Development of a Magnetic Bead Immunoassay Based Flow Cytometry Method for the Detection of Anthrax Virulence Factors The basic principle of the assay developed in this work is depicted in Figure 1 A. In this study, three magnetic beads with different distinct fluorescence intensity profiles were used. These were coupled separately to specific monoclonal antibodies against EF, LF, or PA. In the presence of the relevant antigen, the Interaction of the antibody-coupled beads with the antigen results in a complex formation that can then be captured magnetically for washing out irrelevant proteins and particles throughout the assay. Consequently, labeling occurs by specific polyclonal IgG antibodies targeted against the antigen of interest (i.e., PA, LF, and EF). The resulting sandwich-complex is then detected through the addition of a secondary fluorescent-labeled antibody. FCM analysis allows the identification of the labeling-fluorophore in parallel to the beads' fluorescence. Figure 1 B shows a representative flow cytometry analysis of the three individual assays for the detection of EF, LF, and PA proteins separately. In order to characterize the assay sensitivity, dose-response assays were performed for the three individual antigens separately ( Figure 2 ). The mean fluorescence intensity (MFI) for the resulting individual Alexa fluor 488 stained complexes was proportional to the concentration of the antigen in the sample. The results were calculated as the ratios between the signal (S) measured for each concentration compared to the signal measured with the assay run against antigen-free PBS (noise, N). This is termed the S/N ratio. S/N ration of 2 and above is considered a positive signal. Results demonstrate a wide dynamic range with a detection threshold of 0.03 ng/mL and 0.3–1 ng/mL for PA and LF, respectively. For EF, the assay proved less sensitive, and a positive signal was obtained from a concentration of 30–100 ng/mL. The magnetic beads chosen for this assay are MegaPlex-C beads. These beads are unique as, in addition to being magnetic, they are available in a wide range of particular and distinct fluorescence intensities. These intensity level differences (hence "color-codes") allow the separation of different bead sub-populations by FCM analysis using the 633 nm laser. This feature enabled the incorporation of the three individual assays into a single, multiplexed assay. Provided that sensitivity and specificity are not compromised, this format allows rapid and simultaneous detection of PA, LF, and EF virulence proteins in the same sample. To examine the FCM ability to distinguish between the different beads in a multiplex assay, we tested whether mixed beads can be separately gated by their color-code. Antigens (PA, LF, and EF) were added to one mixture ("positive control") while the other was kept as a negative control (without antigens). Analysis of these samples showed that each individual bead population could be identified and separately gated ( Figure 3 A). Once we successfully demonstrated the ability to differentially identify discrete bead populations in multiplex format, we set out to estimate the effect of the multiplex format on the detection sensitivity of the assay. In order to assess assay sensitivity and specificity, multiplexed dose-response assays were performed with purified antigens. No significant changes in assay sensitivity ( Figure 3 B) for PA and LF were observed compared to the individual assays ( Figure 2 ). A slight decrease in sensitivity was observed for EF detection. Importantly, no cross-reactivity was observed for LF and EF proteins. A minimal amount of cross-reactivity at high PA concentrations (1 µg/mL) with EF and LF coupled beads was measured. Since our purified PA antigen contained traces of bacterial proteins [ 26 ], we assumed that these impurities were the cause of the cross-reactivity. To ascertain assay specificity, a supernatant derived after a 24 h incubation of delta EF/LF Vollum strain [ 27 ] was tested ( Supplementary Materials Figure S1 ). As hypothesized, no cross-reactivity of PA was observed with EF and LF coupled beads. 2.2. Applying the New Multiplex Assay for B. anthracis Spore Detection Thus far, we demonstrated that our new multiplex immune-assay can simultaneously detect three secreted virulence factors of B. anthracis . This assay is rapid (approximately 80 min), sensitive, and specific. Previous research indicates that PA can serve as a surrogate marker to estimate the concentration of B. anthracis [ 23 ]. In addition, we have also demonstrated that a short incubation in DMEM (supplemented with 10% serum in the presence of CO 2 ) results in germination and bacterial growth coupled with the early secretion of virulence factors [ 25 , 28 ]. Since this toxin production process is performed under batch conditions (microtiter dish), the concentration of these factors in the growth media increases over time. Therefore, potentially we can increase the toxin concentration in the growth medium by prolonging the incubation phase. Coupling this incubation process with sensitive and specific detection assays may enable spore detection in environmental samples containing low spore concentrations. Therefore, the proposed test combines sample incubation under increased toxin secretion conditions with the multiplex FCM based assay. In order to ascertain the optimal growth conditions enabling unequivocal identification of virulence factors in relation to the initial spore concentration, B. anthracis ATCC14185 spores (5 × 10 5 CFU/mL) were suspended in medium containing 10% or 50% fetal bovine serum (FBS) and incubated at 37 °C for 2 h in the presence of 5% or 10% CO 2 . Following incubation, the presence of virulence factors in the growth medium was assessed by the multiplexed assay. Results ( Figure 4 ) show detectable levels of PA following 2 h incubation while no significant detection of LF or EF was observed. Though positive detection of PA (S/N > 2) was obtained in all samples, the highest signal was reached in samples incubated in 50% FBS growth medium in 10% CO 2 . These were thus defined as the optimal conditions for increased toxins secretion. These results validate our hypothesis that PA is secreted in detectable quantities at the early stages of bacterial growth and is, therefore, suitable as a surrogate marker for spore detection. To evaluate the effect of extended incubation time on toxin secretion and assay sensitivity, different doses of ATCC14185 spores (1 × 10 3 –1 × 10 7 CFU/mL) were incubated for 2–5 h under optimal conditions. Following the incubation process, toxin presence in the growth medium was determined by the multiplexed immunoassay. Our results ( Figure 5 ) show that prolonged incubation time significantly improves the assay's sensitivity. Similarly to the previous results, PA was the first to be detected in all samples. Positive detection of PA was possible within 2 and 3 h of incubation in 1 × 10 5 and 1 × 10 3 –1 × 10 4 spores/mL, respectively. Positive detection of LF was obtained after 3 h incubation for 1 × 10 5 and 5 h for 1 × 10 4 spores/mL. EF was detected only at high spore concentrations (above 1 × 10 6 CFU/mL), after 4–5 h incubation. These results emphasize the benefit of using a dynamic process, in which samples can be collected continually along 5 h of incubation, thus increasing assay sensitivity. Since this method is supposed to detect spores in environmental samples, we tested our protocol on samples taken from asphalt-paved roads, sidewalks, and soil samples, which were spiked with known spore doses. First, we tested the effect of various environmental samples on the assay noise level in our multiplex assay. To this end, 10 environmental samples were incubated in a growth medium at 37 °C 10% CO 2 for 5 h. Following incubation, the growth medium was tested for the presence of EF/LF/PA by the multiplex assay. No positive result was obtained in any of the negative samples ( Figure 6 A). Next, we assessed the effect of environmental contaminants on B. anthracis toxins secretion during the incubation process. Environmental samples (n = 10) were spiked with 5 × 10 3 CFU/mL B. anthracis ATCC14185 spores, processed (Material and Methods), resuspended in growth medium and incubated at 37 °C 10% CO 2 for 5 h. All environmental samples ( Figure 6 ) were positive for PA (S/N range 4–25). However, the spiked environmental samples reached a lower signal (~2 fold decrease) relative to the same spore concentration spiked into PBS. This reduction may result from decreased spore dose in the final sample due to the spiking/processing protocol used, by signal inhibition resulting from the environmental sample constituents, or a combination of both. To validate these results, 40 additional environmental samples, collected in four separate locations in Israel, were spiked with 1 × 10 4 CFU/mL spores and processed. Following 5 h incubation, PA was detected in all samples, while LF was detected only in 15/40 samples (S/N range 2–3, Figure 6 B). In agreement with the above, PA levels significantly varied between samples, but all were clearly positive (S/N range 18–102, Figure 6 B), indicating that the method is indeed suitable for B. anthracis spore detection in an environmental sample. 2.3. Assay Validation Using Fully Virulent B. anthracis Vollum Strain Spores Having performed all of our experiments on B. anthracis ATCC14185 spores, we validated our results using a fully virulent Vollum strain. Vollum spores (1 × 10 2 –1 × 10 7 CFU/mL) were suspended in growth medium and incubated for 2 and 5 h. Results in Figure 7 A show positive detection of PA and LF at concentrations of 1 × 10 3 CFU/mL and 1 × 10 5 CFU/mL spores, respectively, following a 2 h incubation process. Prolonging incubation process to 5 h enables the detection of PA at concentrations of 1 × 10 2 CFU/mL and LF at concentrations of 1 × 10 3 CFU/mL. A positive signal for EF was observed only from 1 × 10 4 CFU/mL ( Figure 7 B). Positive PA detection was achieved in 14/14 different soil samples spiked with 1 × 10 3 CFU Vollum spores/g after 2 h incubation (S/N range 3.5–5.5, Figure 7 C). Prolonging incubation to 5 h ( Figure 7 D) resulted in detection of PA in soil samples spiked with 1 × 10 2 CFU spores (S/N range 17–35). LF was observed (S/N range 2.5–4) in all 1 × 10 3 CFU spores samples following 5 h. The assay proved more sensitive to the presence of Vollum spores compared to ATCC14185 spores. This could result from differences in toxin secretion between the two strains. The ATCC14185 strain is a vaccine strain, used to produce PA-containing drug substance and was mutated to facilitate this application [ 29 ]. These mutations may explain this difference. Overall, the proposed method combining biosynthesis and identification of toxins can serve as a rapid and sensitive assay for the indirect detection of viable B. anthracis spores in environmental samples. 3. Discussion B. anthracis , the causative agent of anthrax, has very close relatives in the B. cereus group [ 30 ], with which it shares a great deal of morphological, biochemical, and genetic similarities [ 31 ]. The background presence of these soil bacteria complicates the specific identification of B. anthracis in environmental samples. Among the few features distinguishing B. anthracis within the B. cereus group (excluding biovar anthracis [ 32 ]), is the presence of two virulence plasmids: pXO1, encoding the tripartite toxin genes and pXO2, encoding the capsule operon. The tripartite toxin is comprised of PA, LF, and EF [ 2 ]. B. anthracis is one of the most notorious infectious agents, and thus remains classified as a tier 1 bio-threat affecting both livestock and humans [ 3 ]. Its ability to produces highly resistant spores, the fact that it is endemic (and, therefore, available for isolation and culture) in most of the world, and that the spores are easily dispersed, all make this bacterium highly suitable for weaponization [ 2 ]. In the scenario of a deliberate release of anthrax spores, diagnostic efforts will focus on two types of samples—clinical samples from suspected patients and environmental samples from suspected contaminated areas. In this work, we focused on the development of a method that enables sensitive detection of spores in environmental samples. Such an assay will be useful both in initial detection and risk assessment as well in subsequent decontamination efforts. We demonstrated that the optimal conditions for the incubation process were determined to be in supplemented DMEM containing 50% serum (hence growth medium) under 10% CO 2 . Furthermore, this incubation protocol, combined with the immunoassay, enables the detection of 1 × 10 3 –1 × 10 4 /mL spore concentration following a short, 2-hour incubation. Extending the incubation step to 5 hours increased assay sensitivity to allow the detection of 1 × 10 2 /mL spore. The extended protocol was validated in 50 various environmental samples spiked in 5 × 10 3 –1 × 10 4 CFU/mL B. anthracis 14,185 spores. All spiked samples were found unequivocally positive. In addition, 2 h incubation of 14 soil samples spiked with the fully virulent B. anthracis BA3500 spores resulted in the detection of 3 × 10 3 /mL spores. Extending the incubation to 5 h resulted in the detection of 3 × 10 2 /mL spores. Our assay has not been tested for more prolonged incubations, but it stands to reason that prolonging incubation duration will improve sensitivity. Dealing with environmental samples has proven to be challenging. Existing assays have been shown to suffer from limited assay specificity [ 33 ] and/or sensitivity [ 34 ]. Assays usually include specific selection and isolation of B. anthracis colonies, a process that may take days. Additional identification assays have low sensitivities for environmental samples, suffering a 3–4 orders of magnitude decrease in sensitivity compared to pure samples [ 10 ]. This was demonstrated using specific phage [ 35 ] or other commercial assays [ 36 ]. The detection of B. anthracis spores from environmental samples in various immune affinity-based assays has been shown to be effective for the detection of 1 × 10 5 –1 × 10 6 spores [ 11 , 36 ]. In addition, using specific reporter phage enables the detection of 1 × 10 5 CFU in assays, taking several hours to completion [ 35 ]. Regarding the specificity of these assay, caution is required, as antibodies against B. anthracis can also cross-react with other B. cereus -group strains [ 37 ]. PCR-based assays allow for the identification of ~1 × 10 5 spores/g soil using multiplex-PCR and 1 × 10 3 –1 × 10 4 spores/g soil using real-time PCR [ 7 , 8 , 20 ]. A highly sensitive PCR-based assay, detecting ~10 live spores in the presence of 1 × 10 6 autoclaved spores, requires ~17 h to perform [ 21 , 22 ]. Our results demonstrate that in this assay, the presence of an environmental contaminant did not prove detrimental to the germination process, bacterial growth, virulence factor secretion, or the ability to achieve positive detection. The newly assay is comprised of two steps: The incubation period, as described above, and the newly multiplexed magnetic beads immunoassay. The latter is comprised of three individual incubation steps for antigen capture and the labeling prior to FCM analysis, totaling an 80 min assay. The multiplexed nature of the assay allows for simultaneous detection of PA, LF, and EF in a small sample volume, which allows dynamic and continuous measurement of the secreted virulent factors during the incubation process. This assay applies Luminex particles. The application of such particle for FCM was previously described by Vignali [ 38 ] as a sensitive, simple, and reliable multiplexing method for the quantitation of cytokine, while specific use of Luminex particles for the detection of B. anthracis was previously described by Mechaly et al. [ 39 ]. This assay described the identification of soluble capsule and PA as specific markers in homogenous blood culture using the xMAP technology by Luminex. In contrast to the classical microbiological methods, the short incubation process in our new assay enables the secretion of the B. anthracis virulent factors without a substantial influence on the biological and chemical composition of the environmental sample. Another positive effect of the incubation step is the proliferation of the bacteria. This proliferation should prove beneficial to additional diagnostic measures, such as classical immunological assays (i.e., immuno-fluorescent assay) and nucleic acid-based detection. We, therefore, propose that the newly developed methodology described in this work allows the detection of spore concentrations that are 3–4 orders of magnitude lower than published immuno-affinity assays [ 11 , 36 ], being comparable to sensitive PCR-based assays [ 7 , 8 , 20 ], while being performed in under 3.5 h for the total assay. This time frame could potentially be shortened even further by modifying the multiplex assay step to be performed as a one-step assay (all in one) by using direct labeled secondary antibodies or a biotin-streptavidin complex for signal enhancement. We plan to pursue this possibility in the near future. Such a sensitive, specific, and relatively rapid test may prove invaluable not only in the identification of the initial release of the threat agent, but also in the extensive forensic investigation following such an event and even play a part in monitoring the effectiveness of decontamination efforts. Outside the realm of bio-terrorism, this assay may prove valuable in epidemiological studies following natural outbreaks involving livestock and pastures, allowing the analysis of soil samples pertaining to such events. 4. Materials and Methods 4.1. B. anthracis Strains The strains used in this study were B. anthracis Vollum (ATCC 14578) (pXO1 + pXO2 + ) and Vollum Δ lef Δ cya [ 27 ]. ATCC 14185 (pXO1 + pXO2 - ), which is a non-encapsulated attenuated strain that produces the pXO1 encoded virulence factors, including PA, LE, and EF, was also used. All strains were from the Israel Institute for Biological Research collection. 4.2. Antibodies and Proteins The monoclonal antibodies utilized in this study were mouse anti-PA (m55) [ 40 ], mouse anti-EF (EMN2/1), and mouse anti-LF (LMN 1/5) [ 27 ]. These antibodies were purified from a mouse ascetic fluid using protein G chromatography (GE Healthcare, Uppsala, Sweden) according to the manufacturer's instructions and dialyzed against phosphate-buffered saline (PBS; PH7.4; Biological Industries, Beth Haemek, Israel). Anti EF, LF, and PA polyclonal antibodies were obtained from hyperimmune guinea pigs immunized with purified EF, LF, or PA, respectively. Polyclonal antibodies were used as a serum. Donkey anti guinea pig IgG (H+L) conjugated to Alexa fluor 488 was purchased from Jackson immune research laboratories (cat #706-545-148). PA was purified, as described previously [ 41 ]. EF and LF were isolated by genetically inserting a His-tag sequence at the carboxy terminus of each protein in the background of the ATCC 14185 strain. Separate strains were produced for the purification of EF and LF. The proteins were purified on a nickel-column from the supernatant of bacterial cultures, as described in Reuveny et al. 2001 [ 41 ]. 4.3. Preparation of Coupled Magnetic Beads In this study, we used Luminex MagPlex-C microspheres, 2.5 × 10 6 /mL, regions 12, 15, and 20 (Luminex, Austin, TX; catalog numbers MC10012, MC10015 and MC10020). Specific antibodies were coupled to beads (5 µg of each per 1 × 10 6 beads) using the Luminex antibody coupling kit (catalog number 40-50016) according to the manufacturer's instruction. 4.4. Magnetic Bead Based Detection Assays The optimization of the magnetic bead-based assay for each of the individual antigens tested was done separately. Assays were performed in a final volume of 100 μl in 96-well Greiner black microplates (Greiner, PP F-Bottom chimney well, 655209). The assays were comprised of three steps; 1. Coupled beads (5 × 10 4 ) were mixed with 100 μl of either purified antigen or samples following the incubation process. The mixtures were then agitated in the dark at room temperature for 30 min in an orbital micro shaker (Dynatech, England). The plates were then washed by a microplate washer (Tecan, Hydrospeed, 30054550) equipped with a smart-2MBS 96 well magnetic plate. 2. Captured beads were re-suspended in 100 μl of guinea pig polyclonal IgG antibodies (diluted 1:1000 in PBS 1% BSA 0.05% tween) targeted against the specific antigen of interest and incubated for 30 min. 3. Following an additional magnetic wash, anti-Guinea Pig IgG:A488 (1:800) was added to the captured beads for an additional 10 min incubation. Beads were magnetically washed and resuspended in 200 μl PBS 1% BSA for FCM analysis using a BDTM Fortesa flow cytometer. Multiplexed assays were carried out as described for individual assays using 2 prepared assay mixtures containing the antibody-coupled beads and the anti-antigen polyclonal antibodies. 4.5. Environmental Sample Collection, Preparation, and Toxin Biosynthesis Process Environmental samples were collected in 4 distinct locations. Samples included asphalt paved roads, sidewalks, and soil samples. As for the asphalt paved roads and sidewalks, samples were collected from 20 cm 2 areas using two swabs soaked in 1× PBS, followed by one dry swab. At this stage, the swabs were inoculated with B. anthracis spores. The swabs were then vortexed in 4 mL PBS for 1 min, and 1 mL from the suspension was transferred to a clean 1.5 mL tube. Soil samples were comprised of 1 gram soil per sample. Samples were suspended in 1 mL PBS containing B. anthracis spores. All samples were vortexed, allowed 5 min incubation at room temperature (RT), prior to supernatants being transferred to new tubes, and centrifuged at 14,000 g for 5 min. Supernatants were discarded, and samples were resuspended in 400 µl growth medium containing DMEM + 50% Fetal Bovine Serum (FBS). The suspended samples were then plated on 48 well plates for 2–5 h incubation at 37 °C under 10% CO 2 . 4.6. Signal Analysis The mean fluorescence intensity (MFI) for the resulting individual Alexa fluor 488 stained complexes was proportional to the concentration of the relevant antigen in the sample. The results were calculated as the ratios between the signal (S) measured for each concentration compared to the signal measured with the assay run against antigen-free PBS (noise, N). This was termed the S/N ratio. This calculation enabled the normalization of multiple experiments and the determination of a universal threshold for positive samples, in accordance with the International Council for Harmonization of Technical Requirements for Pharmaceuticals for Human Use (ICH) guidelines for validation of analytical procedures [ 42 ]. 4.1. B. anthracis Strains The strains used in this study were B. anthracis Vollum (ATCC 14578) (pXO1 + pXO2 + ) and Vollum Δ lef Δ cya [ 27 ]. ATCC 14185 (pXO1 + pXO2 - ), which is a non-encapsulated attenuated strain that produces the pXO1 encoded virulence factors, including PA, LE, and EF, was also used. All strains were from the Israel Institute for Biological Research collection. 4.2. Antibodies and Proteins The monoclonal antibodies utilized in this study were mouse anti-PA (m55) [ 40 ], mouse anti-EF (EMN2/1), and mouse anti-LF (LMN 1/5) [ 27 ]. These antibodies were purified from a mouse ascetic fluid using protein G chromatography (GE Healthcare, Uppsala, Sweden) according to the manufacturer's instructions and dialyzed against phosphate-buffered saline (PBS; PH7.4; Biological Industries, Beth Haemek, Israel). Anti EF, LF, and PA polyclonal antibodies were obtained from hyperimmune guinea pigs immunized with purified EF, LF, or PA, respectively. Polyclonal antibodies were used as a serum. Donkey anti guinea pig IgG (H+L) conjugated to Alexa fluor 488 was purchased from Jackson immune research laboratories (cat #706-545-148). PA was purified, as described previously [ 41 ]. EF and LF were isolated by genetically inserting a His-tag sequence at the carboxy terminus of each protein in the background of the ATCC 14185 strain. Separate strains were produced for the purification of EF and LF. The proteins were purified on a nickel-column from the supernatant of bacterial cultures, as described in Reuveny et al. 2001 [ 41 ]. 4.3. Preparation of Coupled Magnetic Beads In this study, we used Luminex MagPlex-C microspheres, 2.5 × 10 6 /mL, regions 12, 15, and 20 (Luminex, Austin, TX; catalog numbers MC10012, MC10015 and MC10020). Specific antibodies were coupled to beads (5 µg of each per 1 × 10 6 beads) using the Luminex antibody coupling kit (catalog number 40-50016) according to the manufacturer's instruction. 4.4. Magnetic Bead Based Detection Assays The optimization of the magnetic bead-based assay for each of the individual antigens tested was done separately. Assays were performed in a final volume of 100 μl in 96-well Greiner black microplates (Greiner, PP F-Bottom chimney well, 655209). The assays were comprised of three steps; 1. Coupled beads (5 × 10 4 ) were mixed with 100 μl of either purified antigen or samples following the incubation process. The mixtures were then agitated in the dark at room temperature for 30 min in an orbital micro shaker (Dynatech, England). The plates were then washed by a microplate washer (Tecan, Hydrospeed, 30054550) equipped with a smart-2MBS 96 well magnetic plate. 2. Captured beads were re-suspended in 100 μl of guinea pig polyclonal IgG antibodies (diluted 1:1000 in PBS 1% BSA 0.05% tween) targeted against the specific antigen of interest and incubated for 30 min. 3. Following an additional magnetic wash, anti-Guinea Pig IgG:A488 (1:800) was added to the captured beads for an additional 10 min incubation. Beads were magnetically washed and resuspended in 200 μl PBS 1% BSA for FCM analysis using a BDTM Fortesa flow cytometer. Multiplexed assays were carried out as described for individual assays using 2 prepared assay mixtures containing the antibody-coupled beads and the anti-antigen polyclonal antibodies. 4.5. Environmental Sample Collection, Preparation, and Toxin Biosynthesis Process Environmental samples were collected in 4 distinct locations. Samples included asphalt paved roads, sidewalks, and soil samples. As for the asphalt paved roads and sidewalks, samples were collected from 20 cm 2 areas using two swabs soaked in 1× PBS, followed by one dry swab. At this stage, the swabs were inoculated with B. anthracis spores. The swabs were then vortexed in 4 mL PBS for 1 min, and 1 mL from the suspension was transferred to a clean 1.5 mL tube. Soil samples were comprised of 1 gram soil per sample. Samples were suspended in 1 mL PBS containing B. anthracis spores. All samples were vortexed, allowed 5 min incubation at room temperature (RT), prior to supernatants being transferred to new tubes, and centrifuged at 14,000 g for 5 min. Supernatants were discarded, and samples were resuspended in 400 µl growth medium containing DMEM + 50% Fetal Bovine Serum (FBS). The suspended samples were then plated on 48 well plates for 2–5 h incubation at 37 °C under 10% CO 2 . 4.6. Signal Analysis The mean fluorescence intensity (MFI) for the resulting individual Alexa fluor 488 stained complexes was proportional to the concentration of the relevant antigen in the sample. The results were calculated as the ratios between the signal (S) measured for each concentration compared to the signal measured with the assay run against antigen-free PBS (noise, N). This was termed the S/N ratio. This calculation enabled the normalization of multiple experiments and the determination of a universal threshold for positive samples, in accordance with the International Council for Harmonization of Technical Requirements for Pharmaceuticals for Human Use (ICH) guidelines for validation of analytical procedures [ 42 ].

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9146367/

Green Revolution to Gene Revolution: Technological Advances in Agriculture to Feed the World

Technological applications in agriculture have evolved substantially to increase crop yields and quality to meet global food demand. Conventional techniques, such as seed saving, selective breeding, and mutation breeding (variation breeding), have dramatically increased crop production, especially during the 'Green Revolution' in the 1990s. However, newer issues, such as limited arable lands, climate change, and ever-increasing food demand, pose challenges to agricultural production and threaten food security. In the following 'Gene Revolution' era, rapid innovations in the biotechnology field provide alternative strategies to further improve crop yield, quality, and resilience towards biotic and abiotic stresses. These innovations include the introduction of DNA recombinant technology and applications of genome editing techniques, such as transcription activator-like effector (TALEN), zinc-finger nucleases (ZFN), and clustered regularly interspaced short palindromic repeats/CRISPR associated (CRISPR/Cas) systems. However, the acceptance and future of these modern tools rely on the regulatory frameworks governing their development and production in various countries. Herein, we examine the evolution of technological applications in agriculture, focusing on the motivations for their introduction, technical challenges, possible benefits and concerns, and regulatory frameworks governing genetically engineered product development and production. 1. Introduction Agriculture plays a critical role in transforming economies and promoting food security [ 1 ]. Apart from providing livelihoods for 570 million farmers worldwide [ 2 ], the current global agricultural output feeds 7.5 billion people, with three major crops, rice ( Oryza sativa ), corn ( Zea mays ), and wheat ( Triticum aestivum ), providing over 60% of our food energy intake [ 3 ]. To keep up with the food demand, technological applications in agriculture have evolved substantially to maximize growth yield. Early innovation, such as the conventional selective breeding technique, involves crossing desired parental plants and selecting offspring with relevant combined characteristics [ 4 ]. This technique resulted in high-yielding plant varieties in wheat and rice during the 'Green Revolution' era in the 1950s ( Figure 1 ). Indeed, selective breeding techniques and other improvements in agricultural practices (e.g., improved irrigation systems, chemical fertilizers, and heavy machinery) have tremendously increased global food production over the past few decades [ 5 , 6 , 7 ]. However, as food demand is expected to increase by 60% [ 8 ] to feed 9.7 billion people in 2050 [ 9 ], a novel strategy is needed to promote food security [ 1 ]. Conventional breeding techniques have their disadvantages, such as (1) uncontrolled genetic mutations, (2) the need for the parental plants to be closely related to produce offspring, and (3) the laborious and time-consuming procedure of crossing and backcrossing hybrids to obtain the elite lines [ 5 , 10 ]. Unlike conventional methods, modern biotechnology tools provide a more specific and controlled way of altering plant DNA or proteins. This may result in high-yielding and nutritious crops and/or those which are more resilient to extreme weather, arid conditions, and diseases [ 11 ]. As biotechnology tools overcome the limitations associated with conventional techniques, they are considered promising tools to further improve global agronomic output and promote food security. However, concerns regarding biotechnology applications have been raised by critics. They include biosafety issues, ethical issues, and the long-term impact on human health and the environment. In this review, we describe technological progress in crop improvement, focusing on the shift from the use of conventional breeding techniques during the 'Green Revolution' era towards the more complex molecular techniques in the so-called 'Gene Revolution' era ( Figure 1 ). The motivation behind this shift is examined. We then identify the distinction between several modern biotechnology tools and highlight their definitional ambiguities in the current regulatory framework. We also discuss the current regulatory frameworks governing genetically modified (GM) food production in certain countries and suggest how the newer gene-editing tools should fit into these regulatory frameworks. The current benefits and concerns associated with biotechnology-derived foods are also listed and discussed. Ultimately, the review addresses how modern biotechnology tools complement traditional methods by offering novel ways to improve crops and food production. 2. Mutation Breeding to Increase Plant Varieties Scientists have long acknowledged the role of naturally occurring mutations in plants that may produce traits that thrive in the changing environment. However, since natural mutations occur by chance, the possibility of superior traits emerging naturally happens very slowly [ 12 ]. Thus, 'mutation breeding' was developed to induce random mutations to mimic and expedite this process. Pioneering studies on radiation-induced mutations using X-rays on Drosophila [ 13 ], and plants, such as barley [ 14 ] and corn [ 15 ], laid the foundation for gene function and repair studies. These studies paved the way for mutational breeding applications in agriculture [ 16 ]. Later, the use of highly energetic gamma-rays further accelerated the mutational breeding process, in which the number of mutations occurring per cell was affected by different levels of radiation imposed on the plants, including the distances of the plant to the radiation source or the exposure time [ 17 ]. Chemical-induced mutagenesis of plants was achieved using alkylating agents, such as ethylmethanesulfonate (EMS), methylnitrosourea (MNU), and ethylnitrosourea (ENU). Among these alkylating agents, EMS has been widely used to introduce random point mutations in plants to generate variants of agronomic importance [ 18 ]. In the recent decade, fast neutron radiation-induced mutagenesis has been increasingly used due to its efficiency in rapidly generating a library of knockout genes [ 19 ]. Its ability to create multiple deletions ranging from 1 bp to 18 Mb has been valuable in delineating several biosynthesis pathways in important crops, such as raffinose family oligosaccharide synthesis [ 20 ] and seed protein synthesis in the soybean ( Glycine max ) [ 21 ]. To date, there are about 3200 radiation-induced plant varieties produced from 210 plant species cultivated in over 70 different countries [ 16 ]. However, undirected mutagenesis techniques have their own disadvantages, including: (1) the 'randomness' of the mutation induction, which may cause off-target alterations across the plant genome; (2) limitations in producing desirable dominant alleles; and (3) the laborious work of screening out mutant plants with desirable traits from a huge population [ 22 ]. Therefore, a more direct and efficient way of DNA alteration in the plant genome is desirable to accelerate the production of desirable variants while reducing the risk of off-target mutations. 3. Breeding Strategies to Increase Crop Yield Plant hybridization is a widely used conventional technique to produce crops with better characteristics, such as better yield, improved color, and increased resilience to environmental stresses and diseases [ 23 ]. It involves crossing two different species or varieties of the same plant by transferring the pollen of a plant into the female part of a different plant resulting in 'hybrids'. This gene method is different from another selective breeding technique referred to as 'inbreeding' that involves two genetically similar or biologically related parents. The main goal of hybridization is to acquire the best possible characteristic(s), while inbreeding aims to maintain stable characteristic(s) over time [ 24 , 25 ]. Since most crops are hermaphroditic (having both male and female organs on the same flower), the hybridization technique always seeks to reduce or avoid the process of self-pollination during flowering [ 26 ]. One way to achieve this is to make the female line male-sterile via mechanical or genetic alteration [ 27 ]. One prominent example of the former is the detasseling (the process of removing the tassel) of maize which is a cost- and time-productive method due to the presence of the anthers at the top of the plant [ 28 ]. However, mechanical sterilization is not commercially applicable for other crops that consist of the male and female components coexisting in the same flower [ 27 ]. Therefore, a genetic means was used to avoid self-pollination using self-incompatibility (SI) or cytoplasmic male sterility (CMS) systems [ 29 ]. The SI system largely depends on the S-locus, where tightly linked genes form co-adapted haplotypic combinations which control pollen and pistil specificities. After pollination, the pollen will be rejected if the specificity is encoded by the same haplotype as that of the pistil [ 30 ]. In CMS, the production of functioning pollen is prohibited by virtue of mitochondrial dysfunction, which produces open reading frames (ORFs) that induce pollen disruption. Spontaneous fertility reversion is possible, although happening at a relatively low rate in major crops, such as rice and corn, and other crops, such as carrot, common bean, and mustard greens [ 31 ]. Both SI and CMS systems are still used today. However, they are limited to the sporophytic system, such as Brassica , and gametophytic systems in Papaveraceae, Solanaceae, Rosaceae, and Plantaginaceae [ 32 ]. The difficulties in developing and applying these systems, however, hinder their utilization for other crops. 4. The Rise of Recombinant DNA Technology and Genetically Modified (GM) Crops The next major technological landmark in agriculture involves recombinant DNA technology (also known as 'genetic engineering' or 'genetically modified organism [GMO] technology'). This technology was used to produce mostly herbicide- or pesticide-resistant GM crops in the early 1990s, marking the start of the 'Gene Revolution' era [ 33 ]. The inception of DNA recombination technology started earlier with the exploitation of Agrobacterium tumefaciens to transfer transgenes into host plants ( Table 1 ). This eventually opened a new realm in crop improvement [ 34 , 35 ]. Bevan et al. [ 36 ] demonstrated that a chimeric antibiotic-resistant gene could be transferred into tobacco ( Nicotiana tabacum ), and the transformed plant cells could be selected on antibiotic-supplemented growth media. The first virus-resistant transgenic plant expressing tobacco mosaic virus (TMV) coat protein was then generated, showing delayed symptom development when infected with TMV compared to non-transgenic lines [ 37 ]. The first transgenic insect-resistant plant was generated in the following year by expressing an insecticidal Bt2 protein from Bacillus thuringiensis (Bt) in tobacco [ 38 ]. Biotechnology has opened up a new research frontier using RNA interference (RNAi) with the discovery of the 'co-suppression' phenomenon that caused gene silencing in petunia ( Petunia hybrida ) [ 45 ]. Unexpectedly, an attempt to overexpress chalcone synthase (CHS), a key enzyme in anthocyanin biosynthesis, in petunia did not produce flower petals with darker pigment [ 45 ]. Instead, the integration of the chimeric petunia CHS transgene resulted in either white or partially white flower petals in 42% of the transgenic petunia. This indicated a possible 'co-suppression' phenomenon between the endogenous and the introduced CHS genes, which caused a block in the anthocyanin biosynthesis [ 45 ]. The 'co-suppression' mechanism was further elucidated in subsequent work by Lindbo et al. [ 59 ] on tobacco etch virus (TEV)-resistance in transgenic tobacco. In their experiment, the transgene mRNA was abundantly detected in the noninfected plants while barely detected in the transgenic tobacco that recovered from the TEV infection. This strongly suggested the co-suppression of the virus and that the transgene must operate at the RNA level since the TEV has an RNA genome. The simplicity of the RNAi mechanism makes it an effective tool for gene knockdown studies. This eventually led to the production of transgenic crops with commercial value. A prominent example is the commercialization of the 'Flavr Savr' tomato (cherry tomato; Lycopersicon esculentum ) in 1993, approved by the US Food and Drug Administration (FDA) ( Table 1 ). The Flavr Savr tomato was based on RNAi technology, where an antisense expression cassette of the endogenous polygalacturonase (PG) gene was integrated into the tomato. PG was found to dissolve cell-wall pectin in the fruit ripening process. By silencing the PG expression, the transgenic Flavr Savr tomato had a longer shelf-life than the non-transgenic tomato [ 60 ]. In 1995, the US Environmental Protection Agency (EPA) approved the first pesticide-producing food crops (Bt potato and Bt corn) and non-food crops (Bt cotton) [ 49 ]. In the following year, transgenic glyphosate-resistant soybean was the first herbicide-resistant crop to be marketed for the consumer market in the USA [ 50 ]. The adoption of GM crops has been swift in the USA. The cultivation area of GM crops in the USA has increased from 3 million hectares (ha) in 1996 [ 61 ] to 71.5 million ha in 2019 [ 62 ]. Similarly, the global adoption of GM crops has substantially increased since their first introduction in 1996. The global cultivation area has increased 112-fold from only 1.7 million ha in 1996 to 190.4 million ha in 2019 [ 62 ]. A total of 44.2%, 43.9%, 9.9%, and 1.5% of GM crops planted in the world can be found in North America, South America, Asia, and Africa, respectively. Only 0.32% and 0.05% of the GM crops were planted in Oceania and Europe, respectively [ 62 ]. The top producer of GM crops in 2019 was the USA (71.5 million ha), followed by Brazil (52.8 million ha) and Argentina (24.0 million ha). Up to 17 million farmers in 29 countries have planted 14 types of GM crops with various traits of improvement, including soybean, corn, cotton, canola ( Brassica napus ), alfalfa ( Medicago sativa ), sugar beet ( Beta vulgaris ), sugarcane ( Saccharum officinarum ), papaya ( Carica papaya ), safflower ( Carthamus tinctorius ), potato ( Solanum tuberosum ), eggplant ( Solanum melongena ), squash ( Cucurbita ), apple ( Malus pumila ), and pineapple ( Ananas comosus ) [ 62 ]. The first generation of GM crops contained only a single introduced trait (mono-trait) [ 63 ]. Adding new transgenes to an existing transgenic plant has been challenging due to difficulties in introgressing additional transgenic traits that segregate independently [ 64 ]. As recombinant DNA technology advances, multiple traits can now be integrated within the same GM plant, a process known as 'gene stacking' [ 64 ]. This innovation has become the preferred strategy in GM crop production, especially in the USA. In 2020, the cultivation of gene-stacked GM corn and GM cotton ( Gossypium ) covered 86% and 87% of their total cultivation areas, compared to 4% and 33% in 2000, respectively ( Figure 2 ). Insect tolerance, herbicide tolerance, and virus resistance are the three most introduced individual traits in GM crops ( Figure 3 ). In recent years, several newer approaches to facilitate gene stacking in crops have been reported, such as marker-assisted selection (MAS)-based gene pyramiding [ 65 ] and gene assembly in Agrobacterium by nucleic acid transfer using recombinase technology (GAANTRY) [ 66 ]. Newer genome editing tools that allow repeated gene integration within the same transgenic loci will further expedite the production of stacked-gene varieties. Climate change has further increased the need to produce improved crops with better resilience in the ever-changing environment. If the issue is not addressed, climate change may soon disrupt food availability, reduce access to food, and affect food quality and yield [ 67 , 68 , 69 , 70 ]. Therefore, continuous innovations require to be made to enable specific changes in the crop genome to speed up the process of identifying plants with desirable traits. 5. A New Era of Genome Editing Using Sequence-Specific Nuclease (SSN)-Based Tools The first generation of GM plants usually involves the random insertion of foreign genes throughout the host plant genome. Unlike GM technology, newer genome editing tools use sequence-specific nucleases (SSN) which allow alteration of pre-determined DNA sequences in the host plant genome by harnessing native DNA repair machinery [ 71 ]. Due to their specificity, SSN-based tools have been widely applied to introduce compositional changes or to produce novel variants with higher resilience against biotic and abiotic stresses [ 72 ]. To date, there are four major SSNs that have been used in genome editing, including engineered homing endonucleases or meganucleases, zinc finger nucleases (ZFNs), transcription activator-like effector nucleases (TALENs) and clustered regularly interspersed short palindromic repeat (CRISPR)/Cas9 systems [ 73 , 74 , 75 , 76 ]. SSN acts as a molecular 'scissor' to induce double-strand breaks (DSB) at or near the site of interest in the plant genome [ 77 ]. This triggers either the error-prone non-homologous end-joining (NHEJ) mechanisms that delete a single, or set of, DNA sequence(s) at the repair site or homologous-directed repair (HDR), which allows single base substitution or whole gene replacement [ 78 ] ( Figure 4 ). The resulting DNA modification can be classified as Type-1 (small insertion/deletion), Type-2 (substitution), or Type-3 (large insertion), depending on the repair mechanism pathway initiated [ 79 ]. In Type-1 modification, the cell's native repair machinery re-joins the DNA breakage via the NHEJ mechanism ( Figure 4 ). The mutated sequence can then be isolated by self- or back-crossing to remove all the transgenes expressing the SSN machinery from the resulting mutants. The simplicity of the CRISPR/Cas9 system for producing Type-1 modifications has made it the most used approach for crop improvement compared to the other SSNs [ 80 ]. In addition, recent loss-of-function (knockout) studies using SSN-based techniques have elucidated various gene functions during plant developmental stages. For example, transformed potatoes showed increased late blight resistance when susceptibility (S) genes, namely StDND1 , StCHL1 , and StDMR6-1 , were knocked out using CRISPR/Cas9-based editing [ 81 ]. Resistance against bacterial blight was also improved in rice by editing OsSWEET11 , OsSWEET13 , and OsSWEET14 genes via CRISP/Cas9 [ 82 ]. In addition, in rice, resistance against the seed-borne rice pathogen, Burkholderia glumae , was improved with CRISPR/Cas9-based mutation in Oryza sativa MITOGEN-ACTIVATED PROTEIN KINASE 5 ( OsMPK5 ) [ 83 ]. In another study, enhanced rice grain size and yield were observed by editing several genes ( Gn1a , DEP1 , GS3 , and IPA1 ) that are responsible for the plant architecture [ 84 ]. CRISPR/Cas-targeted mutation in tomato ( Solanum lycopersicum ), SlJAZ2 , improved resistance against tomato bacterial speck disease caused by Pseudomonas syringae pv. tomato DC3000 [ 85 ]. Meanwhile, CRISPR/Cas-editing of the SICLV3 promoter in tomatoes increased the fruit size and number of flower buds [ 86 ]. CRISPR/Cas9-based knockout of three pairs of FAD2 homoeologs in the oilseed crop Camelina ( Camelina sativa ), increased monounsaturated fatty acids (MUFAs) by 80% with a stunted bushy phenotype, while transformants containing two pairs of CsFAD2 homoeologs knocked out, and the other pair from the heterozygous wild-type, showed normal growth and seed MUFA levels increased by up to 60% [ 87 ]. Loss of seed shattering or seed dormancy was observed in canola by knocking out two ALCATRAZ genes [ 88 ]. Applications of other SSNs for crop improvement have also been reported, such as reduced browning in white button mushrooms ( Agaricus bisporus ) [ 89 ] and phytate in corn [ 55 ] using ZFN, and the creation of fragrant rice by disrupting the OsBADH2 gene using TALEN, which resulted in the synthesis of 2-acetyl-1-pyrroline (2AP), a key fragrance compound [ 90 ]. Knocking out strategic genes through Type-1 modification could also increase crop resistance to various diseases, such as rice blast disease and citrus canker disease, by mutating the ethylene response factor (ERF) gene, OsERF922 , in rice [ 91 ] and the canker susceptibility gene, CsLOB1 , in grapefruit ( Citrus × paradisi ) [ 92 ], respectively. Furthermore, unlike the classical mutagenesis technique, Type-1 modification allows simultaneous alteration of multiple alleles from the same loci. This is particularly useful for polyploid plants in which the phenotype of the recessive mutation is usually hidden and only shows up in their progenies, requiring laborious crossing events to regain the desired traits. Examples include the use of TALENs in wheat to introduce multi-allelic DNA changes in the Mildew Locus O (MLO) region to establish a mildew-resistance trait [ 93 ], and the use of CRISPR/Cas9 to alter four different alleles from the gene encoding granule-bound starch synthase (GBSS) in tetraploid potato [ 94 ]. Overall, SSN-based genome editing for Type-1 modification has a relatively high success rate, with the recovery rate of plants with knockout gene(s) ranging from 2% to 75%, with a median of 25% [ 95 , 96 ]. Type-2 and Type-3 SSN-based modifications utilize the less frequent, albeit high-fidelity, HDR repair mechanism with strand invasion of an oligonucleotide or a 'repair template' containing the desired mutations ( Figure 4 ). The repair template is homologous to the target site, and its length can vary from tens to hundreds of bases. Single or several base substitution is generally classified as Type-2 modification, whereas Type-3 modification involves larger DNA fragments or entire gene sequence introduction. Several herbicide-resistant GM crops have been generated by altering acetolactate synthase (ALS), a key enzyme targeted by many herbicides, with the variant degree of Type-2 modification, such as single point CRISPR/Cas9-mediated mutation in watermelon [ 97 ] and rice [ 98 ], double-point TALEN-mediated mutation in rice [ 99 ], and multiple points CRISPR/Cas9-mediated mutation in tomato [ 100 ]. Moreover, substituting thymine with adenine in position 317 of the NON-RIPENING ( NOR ) gene using CRISPR/Cas9-mediated mutation produced a longer shelf-life tomato than the wild-type [ 101 ]. In rice, TALEN-mediated depletion of the cytoplasmic male sterility-associated mitochondrial gene, ORF312 , indicated that ORF312 is a cytoplasmic male sterility (CMS)-causative gene [ 102 ]. Insertion in SSN-mediated Type-3 modification is precise, reducing the risk of genome disruption or positional effects. For example, Shi et al. [ 103 ] generated a highly productive and novel drought-tolerant corn variant by replacing the native promoter in the ARGOS8 gene with a GOS2 promoter using an 800 bp repair template. Herbicide-resistant rice and cassava had also been generated through CRISPR/Cas9 SSN-mediated Type-3 modification using a 476 bp repair template to introduce multiple point mutations in the rice ALS gene [ 104 ] and a 4096 bp template to replace the cassava ESPS gene promoter with a strong constitutive 2 × 35S promoter [ 105 ]. So far, the largest insertion mediated by CRISPR/Cas9 editing was 5.2 kb encoding carotenoid biosynthesis in rice [ 106 ]. SSN-mediated gene insertion can facilitate gene stacking in crop improvement through 'trait landing pads' that allow the assembling of multiple transgenes at the same loci [ 107 ]. This allows the stack of genes to be passed down to the progenies in a single cassette during plant breeding, allowing easier introgression into desired elite lines. The inefficiency of gene replacement or gene insertion through innate HDR has remained a bottleneck to fully exploiting SSN-based genome editing for crop improvement. Few reports have utilized the more dominant NHEJ repair mechanism to achieve Type-2 modification by inducing two DSBs flanking the targeted sequence with a pair of SSNs. A repair template is supplied with ligation overhangs that are compatible with the two induced DSBs [ 96 , 108 ]. Another strategy is to fuse Cas9 endonuclease with Agrobacterium VirD2 relaxase, which brings the repair template in proximity to the DSB to increase the rate and efficiency of HDR [ 109 ]. In recent years, the cytosine and adenine base editor, mediated by CRISPR/Cas9, has emerged as an efficient target base editing tool to produce Type-2 modification for crop improvement [ 110 , 111 ]. In this ingenious innovation, the CRISPR/Cas9 was fused to the cytidine deaminase enzyme, which enabled direct single base substitution (C → T or G → A) without inducing DSB, eliminating the need for a repair template [ 112 , 113 ]. Adoption of this latest innovation has been swift and successful for the rapid generation of GM crops with herbicide-resistant traits in rice [ 114 ], corn [ 115 ], and canola [ 116 ], and impaired amylose biosynthesis in potato [ 117 ]. Nonetheless, several challenges that impair widespread base editing applications remain to be resolved, which have been reviewed by Bharat et al. [ 118 ]. Current research on SSN has focused on increasing its efficiency while minimizing off-target mutation. Simultaneously, there are continuous efforts to further innovate the tool by developing a DNA-free delivery system to eliminate the integration of transgenes that encode SSN-based components into the plant host genome. This could be achieved by delivering preassembled ribonucleoprotein complexes (RNPs), composed of purified recombinant enzyme Cas9 and in vitro-transcribed or synthesized guide RNA (gRNA), into plant protoplasts. Such a strategy was successfully demonstrated by Park et al. [ 119 ] in cabbage ( Brassica oleracea ) with a 2% mutation frequency. However, this approach could be extremely challenging in crop species with inefficient protoplast isolation and regeneration [ 96 ]. Another delivery method that has been reported was bombarding CRISPR/Cas9 RNPs into plant cells. A 4.4% mutation rate has been reported in bread wheat using this method [ 120 ]. Regardless of the strategy employed to achieve DNA-free genome editing, the goal is to overcome the hurdle of additional regulatory approval, especially in countries that rely on process-based regulation [ 121 ] and seek to minimize public concerns [ 106 ]. 6. The Rapid Emergence of the CRISPR/Cas System for High Specificity Gene Editing 6.1. Variety of CRISPR Enzymes and Current Applications in Crops The discovery of CRISPR/Cas-based editing tools marked a major milestone for plant engineering [ 122 ]. First used in mammals, the CRISPR/Cas system involves pairing a gRNA and the CRISPR-associated (Cas) nuclease to recognize complementary nucleic acid sequences for cleavage. Successful targeting requires complementarity between the gRNA and the target site as well as a short sequence flanking region known as a protospacer-adjacent motif (PAM) [ 123 , 124 ]. Unlike ZFNs and TALENs, which require protein-DNA interaction and recoding of large DNA sequences (500–1500 bp) for each new target site, CRISPR-Cas9 is adaptable to many target sites. This is performed by simply changing and integrating the 20-bp protospacer of the guide RNA into the gRNA plasmid backbone while the Cas9 protein remains unaltered [ 124 ]. Due to its simplicity and efficiency, CRISPR systems have become the leading genome editing technology in various plant species, including model plants, food crops, industrial crops, and ornamental plants ( Table 2 ). In addition, potential CRISPR applications for medicinal plants, such as Salvia miltiorrhiza , Dendrobium officinale , Cannabis sativa , and Opium poppy , have been proposed. However, they are still in the early stages of development [ 125 ]. Most mutation alterations induced by the CRISPR/Cas system to study gene functions are characterized by a small number of nucleotide insertions or deletions at the target site. For example, CRISPR/Cas9-mediated, 1-bp to 2-bp deletions in Solanum lycopersicum SALT-RELATED MYB1-LIKE (SlSRM1-LIKE) caused abnormal tomato leaf development with several morphological changes, including thinner leaves, wrinkled edges, raised veins, disordered edge veins, and left and right asymmetry [ 144 ]. A single base pair CG deletion in the Hordeum vulgare (barley) MITOGEN-ACTIVATED PROTEIN KINASE6 ( HvMPK6 ) using CRISPR/Cas9 led to severely reduced grain germination and abnormal seedlings with a shootless phenotype [ 145 ]. In addition, in barley, CRISPR/Cas9-mediated deletions ranging from 1 bp to 25 bp were detected during the double knockout of Hordeum vulgare MICRORCHIDIA1 ( HvMORC1 ) and HvMORC6a . Both genes play important roles in plant immunity and genome stability [ 146 ]. In corn, CRISPR/Cas9-mediated knockout of Zea mays PHOSPHOLIPASE D3 ( ZmPLD3 ) resulted in either 1-bp insertion or a combination of 1-bp insertion and 5-bp deletion in two separate mutant lines, which triggered haploid induction [ 147 ]. In rice, relatively larger deletions of 71 and 33 bp (rather than 1-bp deletion) were necessary to effectively knockout Oryza sativa MICRORNA168a ( OsMIR168a ), which caused significant transcription profile changes. This indicates OsMIR168a 's major transcriptional regulatory role, possibly through its potential target genes, such as Oryza sativa ARGONAUTE1s ( OsAGO1s ) and OsAGO18 [ 148 ]. Potential phenotypic consequences of these transcription profile changes were investigated using KEGG enrichment analysis, suggesting OsMIR168a 's vital roles during plant growth and development as well as in plant–pathogen interaction [ 148 ]. Cas9 nuclease isolated from Streptococcus pyogenes (SpCas9) has been the most widely used in CRISPR/Cas systems [ 149 ]. However, SpCas9 has its disadvantages, such as its tendency to recognize DNA sequences with high similarity with the target site, causing off-target mutations [ 150 , 151 , 152 ]. Moreover, a limited number of DNA sequences can be altered with SpCas9 due to stringent NGG (N  =  A, T, C, or G) PAM requirements [ 153 ]. Additionally, the delivery of SpCas9 via a viral-based vector can be challenging due to its relatively large size, which exceeds the cargo size of the vector [ 154 ]. Several innovations have been made to overcome these challenges, such as substituting SpCas9 with a natural variant, Staphylococcus aureus Cas9 (SaCas9) which recognizes 5′-NNGRRT and has a shorter coding sequence [ 155 ]. Another improvement to the SpCas9 toolbox is the use of Cas9 nickase to increase binding specificity and reduce off-target DNA recognition [ 156 ]. Recently, PAM-less plant genome engineering has been established using an engineered SpRY Cas9 variant, which recognizes almost all PAM sites (NRN>NYN) [ 157 ]. Other SpCas9 variants have been developed, such as Cas9-NG [ 158 ], xCas9 [ 159 , 160 ], and iSpyMacCas9 [ 161 ], which have expanded the Cas–PAM compatibility, further improving the CRISPR/Cas toolbox for genome editing. Cas12 nucleases are another family of Cas proteins applied in CRISPR/Cas-based plant genome editing. The Cas12 protein family, especially the Cas12a effector (formerly known as Cpf1), is considered a major improvement to the CRISPR/Cas system due to several characteristics: (1) smaller size [ 162 ], (2) lack of need for trans-activating crRNA (tracrRNA), (3) ability to cleave DNA via its RuvC domain, and (4) having intrinsic RNAse activity that can process its own guide RNA array, allowing multigene editing from a single RNA transcript. These traits further enhance CRISPR/Cas-based editing efficiency and specificity [ 162 , 163 , 164 ]. Moreover, further expansion of the PAM recognition range is currently explored with Cas12a orthologs, such as LbCas12a, AsCas12a, and FnCas12a and engineered variants, such as LbCas12a-E795L [ 165 ], AsCas12a Ultra [ 166 ], and LbCas12a-RVR [ 167 ]. So far, LbCas12a is the most well-known for its high efficiency in various crops, including Arabidopsis ( Arabidopsis thaliana ) [ 168 , 169 , 170 ], rice [ 171 , 172 , 173 ], corn [ 168 , 174 ], benthi ( Nicotiana benthamiana ) [ 169 ], tomato [ 169 ], lettuce ( Lactuca sativa ) [ 175 ], cotton [ 176 , 177 ], and citrus [ 178 ]. Although Cas9 is still the most routinely used nuclease in genome editing, Cas12a popularity is rapidly gaining momentum. Cas12a application has been shown to increase both NHEJ- and HDR-mediated editing efficiency due to the 5′ extension (4 to 25 nucleotides) of the multiple CRISPR RNAs (crRNAs) as a single-guide RNA (sgRNA) [ 179 ]. In rice, the use of both FnCas12a and LbCas12a, together with crRNA and repairing template DNA, is capable of mediating both NHEJ- and HDR-based genome editing [ 180 ]. In addition, in rice, the use of Cas12a, multiple crRNA, and donor repair templates in an all-in-one expression vector resulted in efficient HDR-mediated biallelic gene targeting within one generation [ 181 ]. This is a notable improvement compared to the predominant NHEJ-mediated monoallelic alteration, which usually resulted in a low mosaic recombination frequency in rice [ 182 ] and Arabidopsis [ 183 ]. As a newly identified nuclease for genome editing, continuous improvements have been made to the Cas12a system to be efficiently used for base editing [ 184 , 185 ] and transcriptional regulation alteration [ 184 , 186 , 187 , 188 , 189 ]. Another effector protein, C2c2 (later named Cas13a), targets 28 nucleotides downstream of crRNA, producing single-stranded RNA degradation and inhibiting the transcription of the pre-targeted genes. Cas13a has been used to modify rice with a maximal knockdown of 78% [ 190 ]. Cas13 has also been shown to significantly inhibit TMV in tobacco leaves, showing its potential use for disease resistance in crops [ 191 ]. Furthermore, modification of Cas13 (i.e., dCas13) allows safer genetic disease treatment in mammalian cells as RNA editing is better for recovery and post-transcriptional regulation than DNA editing [ 192 , 193 , 194 ]. In general, Cas13 provides more accurate silencing than RNAi. However, its application in crops is still being investigated [ 194 ]. 6.2. CRISPR Reagents Delivery Systems Delivery of CRISPR reagents into the plant host has been a major challenge due to the complex genome structure, polyploidy, and possible genomic rearrangements in plants [ 195 ]. Recent advancements in CRISPR delivery involve several methods, including Agrobacterium-mediated gene transfer, biolistic delivery, and the use of protoplasts as efficient systems for gene-editing and regeneration in plants. One innovation in the Agrobacterium-mediated delivery system is to co-deliver developmental regulators (DR) with CRISPR components into the host plants [ 196 ]. The expression of DRs can be effectively used in plants that are hard to regenerate or have a long regeneration time [ 195 ]. Another interesting method involves using viral vectors as delivery vehicles for CRISPR/Cas components [ 197 ]. Adeno-associated viruses (AAVs) have been widely used for this purpose due to the high number of approved AAVs for human clinical trials and the fact that AAVs show fewer immunogenic effects than other viruses [ 198 ]. Other reasons for applying viral vectors include their ability to carry large DNA payloads and transduce a wide range of dividing and non-dividing cells [ 199 ]. Recently, an exciting discovery of a 'hypercompact' CRISPR-Casφ system (~70 kDa; half the size of Cas9 nuclease) by Pausch et al. [ 200 ] may allow gene editing by resolving cargo barriers of the positive-strand RNA virus (PSV). The PSV would carry a single guide RNA (sgRNA) with the Cas9 T-DNA construct. The sgRNA is expected to allow efficient germline invasion and collection of seeds with the heritable mutations, skipping the regeneration step in tissue culture [ 201 ]. Agrobacterium-mediated delivery and particle-based bombardment are two commonly used techniques to deliver CRISPR components into the plant host genome. The former involves the use of a binary vector, superbinary vector, and dual binary vector, and has recently progressed to the use of a ternary vector, further expediting the application of CRISPR-based plant genome editing [ 202 ]. On the other hand, particle-based bombardment involves using gold or tungsten particles coated with biomolecules that are physically inserted into the plant host genome via high-velocity bombardment using a gene gun or biolistic device [ 203 ]. Finally, protoplasts (plant cells without cell walls) offer a feasible DNA-free genome editing system. The protoplast system allows for the pre-evaluation of the gene-editing components before applying them to a full-scale transformation in the host plant. Despite being a popular tool for genome editing, the CRISPR technology faces some technical challenges in delivering the CRISPR/Cas system into crop genomes. These include the low transformation efficiency and recalcitrance to the regeneration of several commercially important crops [ 204 ]. Hence, developing efficient and tissue-culture-independent delivery methods is indispensable. 6.1. Variety of CRISPR Enzymes and Current Applications in Crops The discovery of CRISPR/Cas-based editing tools marked a major milestone for plant engineering [ 122 ]. First used in mammals, the CRISPR/Cas system involves pairing a gRNA and the CRISPR-associated (Cas) nuclease to recognize complementary nucleic acid sequences for cleavage. Successful targeting requires complementarity between the gRNA and the target site as well as a short sequence flanking region known as a protospacer-adjacent motif (PAM) [ 123 , 124 ]. Unlike ZFNs and TALENs, which require protein-DNA interaction and recoding of large DNA sequences (500–1500 bp) for each new target site, CRISPR-Cas9 is adaptable to many target sites. This is performed by simply changing and integrating the 20-bp protospacer of the guide RNA into the gRNA plasmid backbone while the Cas9 protein remains unaltered [ 124 ]. Due to its simplicity and efficiency, CRISPR systems have become the leading genome editing technology in various plant species, including model plants, food crops, industrial crops, and ornamental plants ( Table 2 ). In addition, potential CRISPR applications for medicinal plants, such as Salvia miltiorrhiza , Dendrobium officinale , Cannabis sativa , and Opium poppy , have been proposed. However, they are still in the early stages of development [ 125 ]. Most mutation alterations induced by the CRISPR/Cas system to study gene functions are characterized by a small number of nucleotide insertions or deletions at the target site. For example, CRISPR/Cas9-mediated, 1-bp to 2-bp deletions in Solanum lycopersicum SALT-RELATED MYB1-LIKE (SlSRM1-LIKE) caused abnormal tomato leaf development with several morphological changes, including thinner leaves, wrinkled edges, raised veins, disordered edge veins, and left and right asymmetry [ 144 ]. A single base pair CG deletion in the Hordeum vulgare (barley) MITOGEN-ACTIVATED PROTEIN KINASE6 ( HvMPK6 ) using CRISPR/Cas9 led to severely reduced grain germination and abnormal seedlings with a shootless phenotype [ 145 ]. In addition, in barley, CRISPR/Cas9-mediated deletions ranging from 1 bp to 25 bp were detected during the double knockout of Hordeum vulgare MICRORCHIDIA1 ( HvMORC1 ) and HvMORC6a . Both genes play important roles in plant immunity and genome stability [ 146 ]. In corn, CRISPR/Cas9-mediated knockout of Zea mays PHOSPHOLIPASE D3 ( ZmPLD3 ) resulted in either 1-bp insertion or a combination of 1-bp insertion and 5-bp deletion in two separate mutant lines, which triggered haploid induction [ 147 ]. In rice, relatively larger deletions of 71 and 33 bp (rather than 1-bp deletion) were necessary to effectively knockout Oryza sativa MICRORNA168a ( OsMIR168a ), which caused significant transcription profile changes. This indicates OsMIR168a 's major transcriptional regulatory role, possibly through its potential target genes, such as Oryza sativa ARGONAUTE1s ( OsAGO1s ) and OsAGO18 [ 148 ]. Potential phenotypic consequences of these transcription profile changes were investigated using KEGG enrichment analysis, suggesting OsMIR168a 's vital roles during plant growth and development as well as in plant–pathogen interaction [ 148 ]. Cas9 nuclease isolated from Streptococcus pyogenes (SpCas9) has been the most widely used in CRISPR/Cas systems [ 149 ]. However, SpCas9 has its disadvantages, such as its tendency to recognize DNA sequences with high similarity with the target site, causing off-target mutations [ 150 , 151 , 152 ]. Moreover, a limited number of DNA sequences can be altered with SpCas9 due to stringent NGG (N  =  A, T, C, or G) PAM requirements [ 153 ]. Additionally, the delivery of SpCas9 via a viral-based vector can be challenging due to its relatively large size, which exceeds the cargo size of the vector [ 154 ]. Several innovations have been made to overcome these challenges, such as substituting SpCas9 with a natural variant, Staphylococcus aureus Cas9 (SaCas9) which recognizes 5′-NNGRRT and has a shorter coding sequence [ 155 ]. Another improvement to the SpCas9 toolbox is the use of Cas9 nickase to increase binding specificity and reduce off-target DNA recognition [ 156 ]. Recently, PAM-less plant genome engineering has been established using an engineered SpRY Cas9 variant, which recognizes almost all PAM sites (NRN>NYN) [ 157 ]. Other SpCas9 variants have been developed, such as Cas9-NG [ 158 ], xCas9 [ 159 , 160 ], and iSpyMacCas9 [ 161 ], which have expanded the Cas–PAM compatibility, further improving the CRISPR/Cas toolbox for genome editing. Cas12 nucleases are another family of Cas proteins applied in CRISPR/Cas-based plant genome editing. The Cas12 protein family, especially the Cas12a effector (formerly known as Cpf1), is considered a major improvement to the CRISPR/Cas system due to several characteristics: (1) smaller size [ 162 ], (2) lack of need for trans-activating crRNA (tracrRNA), (3) ability to cleave DNA via its RuvC domain, and (4) having intrinsic RNAse activity that can process its own guide RNA array, allowing multigene editing from a single RNA transcript. These traits further enhance CRISPR/Cas-based editing efficiency and specificity [ 162 , 163 , 164 ]. Moreover, further expansion of the PAM recognition range is currently explored with Cas12a orthologs, such as LbCas12a, AsCas12a, and FnCas12a and engineered variants, such as LbCas12a-E795L [ 165 ], AsCas12a Ultra [ 166 ], and LbCas12a-RVR [ 167 ]. So far, LbCas12a is the most well-known for its high efficiency in various crops, including Arabidopsis ( Arabidopsis thaliana ) [ 168 , 169 , 170 ], rice [ 171 , 172 , 173 ], corn [ 168 , 174 ], benthi ( Nicotiana benthamiana ) [ 169 ], tomato [ 169 ], lettuce ( Lactuca sativa ) [ 175 ], cotton [ 176 , 177 ], and citrus [ 178 ]. Although Cas9 is still the most routinely used nuclease in genome editing, Cas12a popularity is rapidly gaining momentum. Cas12a application has been shown to increase both NHEJ- and HDR-mediated editing efficiency due to the 5′ extension (4 to 25 nucleotides) of the multiple CRISPR RNAs (crRNAs) as a single-guide RNA (sgRNA) [ 179 ]. In rice, the use of both FnCas12a and LbCas12a, together with crRNA and repairing template DNA, is capable of mediating both NHEJ- and HDR-based genome editing [ 180 ]. In addition, in rice, the use of Cas12a, multiple crRNA, and donor repair templates in an all-in-one expression vector resulted in efficient HDR-mediated biallelic gene targeting within one generation [ 181 ]. This is a notable improvement compared to the predominant NHEJ-mediated monoallelic alteration, which usually resulted in a low mosaic recombination frequency in rice [ 182 ] and Arabidopsis [ 183 ]. As a newly identified nuclease for genome editing, continuous improvements have been made to the Cas12a system to be efficiently used for base editing [ 184 , 185 ] and transcriptional regulation alteration [ 184 , 186 , 187 , 188 , 189 ]. Another effector protein, C2c2 (later named Cas13a), targets 28 nucleotides downstream of crRNA, producing single-stranded RNA degradation and inhibiting the transcription of the pre-targeted genes. Cas13a has been used to modify rice with a maximal knockdown of 78% [ 190 ]. Cas13 has also been shown to significantly inhibit TMV in tobacco leaves, showing its potential use for disease resistance in crops [ 191 ]. Furthermore, modification of Cas13 (i.e., dCas13) allows safer genetic disease treatment in mammalian cells as RNA editing is better for recovery and post-transcriptional regulation than DNA editing [ 192 , 193 , 194 ]. In general, Cas13 provides more accurate silencing than RNAi. However, its application in crops is still being investigated [ 194 ]. 6.2. CRISPR Reagents Delivery Systems Delivery of CRISPR reagents into the plant host has been a major challenge due to the complex genome structure, polyploidy, and possible genomic rearrangements in plants [ 195 ]. Recent advancements in CRISPR delivery involve several methods, including Agrobacterium-mediated gene transfer, biolistic delivery, and the use of protoplasts as efficient systems for gene-editing and regeneration in plants. One innovation in the Agrobacterium-mediated delivery system is to co-deliver developmental regulators (DR) with CRISPR components into the host plants [ 196 ]. The expression of DRs can be effectively used in plants that are hard to regenerate or have a long regeneration time [ 195 ]. Another interesting method involves using viral vectors as delivery vehicles for CRISPR/Cas components [ 197 ]. Adeno-associated viruses (AAVs) have been widely used for this purpose due to the high number of approved AAVs for human clinical trials and the fact that AAVs show fewer immunogenic effects than other viruses [ 198 ]. Other reasons for applying viral vectors include their ability to carry large DNA payloads and transduce a wide range of dividing and non-dividing cells [ 199 ]. Recently, an exciting discovery of a 'hypercompact' CRISPR-Casφ system (~70 kDa; half the size of Cas9 nuclease) by Pausch et al. [ 200 ] may allow gene editing by resolving cargo barriers of the positive-strand RNA virus (PSV). The PSV would carry a single guide RNA (sgRNA) with the Cas9 T-DNA construct. The sgRNA is expected to allow efficient germline invasion and collection of seeds with the heritable mutations, skipping the regeneration step in tissue culture [ 201 ]. Agrobacterium-mediated delivery and particle-based bombardment are two commonly used techniques to deliver CRISPR components into the plant host genome. The former involves the use of a binary vector, superbinary vector, and dual binary vector, and has recently progressed to the use of a ternary vector, further expediting the application of CRISPR-based plant genome editing [ 202 ]. On the other hand, particle-based bombardment involves using gold or tungsten particles coated with biomolecules that are physically inserted into the plant host genome via high-velocity bombardment using a gene gun or biolistic device [ 203 ]. Finally, protoplasts (plant cells without cell walls) offer a feasible DNA-free genome editing system. The protoplast system allows for the pre-evaluation of the gene-editing components before applying them to a full-scale transformation in the host plant. Despite being a popular tool for genome editing, the CRISPR technology faces some technical challenges in delivering the CRISPR/Cas system into crop genomes. These include the low transformation efficiency and recalcitrance to the regeneration of several commercially important crops [ 204 ]. Hence, developing efficient and tissue-culture-independent delivery methods is indispensable. 7. SSN-based Genome Editing: Good or Bad? SSN-based genome editing is critical for developing high-yielding, high-quality, and climate-resilient crops. However, despite significant contributions toward developing these crops, potential risks and ethical issues on the commercial release and consumption of genome-edited foods are still being debated. The main benefit of SSN-based genome editing technologies is that these technologies could eliminate the transgene and produce crops with no difference from those generated from conventional breeding. These approaches might increase the acceptability of genome-edited crops and reduce ethical concerns about them. Furthermore, many efforts have been made to improve the efficiency of the SSN-based systems by uncovering new proteins and/or improving existing proteins in these systems. Technical limitations of SSN-based genome editing technologies remain, however, creating significant concerns. These include the possibilities of low on-target editing efficiency and off-target and incomplete editing. For example, the presence of many identical target sites within a genome could affect the accuracy of the Cas9 endonuclease in targeting the correct location. Hence, improving the predictability of off-target editing is imperative to ensure strong public trust and the wider acceptability of genome-edited crops. Fortunately, significant efforts have been made to overcome these technical limitations, including improving delivery methods, methods for increasing the efficiency of DSB repair by HDR and precise gene regulation, and multiplexed and high-throughput genome editing approach. Another concern is whether the edited crops will be affected indefinitely, or if the edited genes will be transferred to future generations, potentially affecting the crops in unexpected ways. The Cas9 and the sgRNA in the CRISPR/Cas9-based system are generally expressed from transgenes integrated into the plant genome [ 205 ]. Therefore, they must be removed because the presence of CRISPR/Cas9 might create challenges for the differentiation of previously generated mutations from newly generated ones. Genetic migration of the edited gene sequence from the genome-edited species to a wild-type species may have environmental consequences, thus limiting their adoption in agriculture [ 206 ]. However, whether SSN-based genome editing technologies are beneficial for crop improvement requires cultivating these edited crops in field conditions. Most previously reported genome-edited crops have not yet reached the field because of biosafety and regulatory issues. In 2021, the first CRISPR/Cas9-edited crop entered the market [ 58 ]. Known as the Sicilian Rouge High GABA, these CRISPR-edited tomatoes contain high levels of γ-aminobutyric acid (GABA), which is claimed to lower blood pressure and promote relaxation [ 62 ]. The commercialization of the CRISPR-edited tomato draws comparison with the previous release of the first RNAi-based 'Flavr Savr' tomato in 1994 in terms of their development, regulations, and public reception. Interestingly, the CRISPR-edited tomatoes were not regulated as GM food as they are 'transgene-free', therefore escaping the definition of a GM crop [ 207 , 208 ]. However, since the release of SSN-edited crops largely depends on the regulatory approval process, it is worth examining the current regulatory landscape governing genome-edited agricultural crop production. 8. SSN-based Genome Editing: A Modern Technology within a Conventional Regulatory Framework Modern SSN-based genome editing allows more precise DNA alteration than conventional breeding techniques. Although off-target mutations in SSN-based genome editing have been reported, they involved few nucleotide changes within the same threshold of naturally-occurring mutations in plants [ 209 ]. On the other hand, mutation breeding techniques involving harmful rays or alkylating agents could introduce thousands of random DNA mutations across the entire plant genome [ 210 ]. The mutation breeding technique is considered a conventional method as it is supposed to mimic the natural process of mutations in plants and has been used to produce improved crop variants for almost a century [ 211 ]. To avoid unnecessary barriers imposed by stringent regulations, proponents argue that Type-1 and Type-2 SSN-based editing tools should be viewed as 'precision breeding' and properly be distinguished from the conventional DNA recombinant technology [ 212 ]. In the European Union (EU), a GMO is described in Directive 2001/18/EC as 'an organism, except for human beings, in which the genetic material has been altered in a way that does not occur naturally by mating and/or natural recombination'. Therefore, even though mutation of DNA sequences that result from a genome editing technique and mutation breeding technique involves similar fundamental processes (induction of DSB and the activation of DNA repair machinery), only products from the latter technique and protoplast fusion are excluded from the definition of GMO. This was established by the recent ruling of the European Court of Justice (ECJ), which has confirmed that plants developed by genome editing approaches are covered by existing biosafety legislation. This includes the Directive 2001/18/EC, which implies that, because GM technology was involved in a plant's development process, it should be subjected to a pre-market risk assessment according to the comprehensive general framework outlined in the Directive [ 80 , 213 ]. According to the ECJ, the risks of using new genome editing tools are possibly similar to those of GMOs resulting from classical mutagenesis techniques. They pointed out the lack of a long safety record due to the rapid introduction of novel products enabled by these new genome editing tools. In contrast, the US Department of Agriculture (USDA) has directed that plants that were developed using genome editing tools fall outside of its regulatory purview. This is because the plants do not contain any transgene from viruses or bacteria, which normally occur from the Agrobacterium -mediated transformation technique [ 214 ]. So far, five CRISPR-edited plants (soybean, camelina, corn, white button mushroom, and green bristlegrass) and several other TALEN-edited plants are already in development and have received USDA approval [ 214 ]. The ruling is significant because the unregulated genome-edited plants are no longer subjected to a lengthy and costly risk assessment evaluation. Variations in the regulatory frameworks among many countries are expected as the regulatory processes are based on several factors, including the definition of living modified organism (LMO), the process of making the LMO-based product, and/or the risk associated with the intended use of the LMO-based product [ 121 , 215 ]. The two main protocols that are referred to for LMO definitions and concepts are (1) the Cartagena Protocol on Biosafety, and (2) Risk Analysis Principles for Foods Derived from Biotechnology (Codex Alimentarius Commission). According to the Cartagena Protocol, any plants produced as a result of genetic engineering are considered LMO, which is defined as 'any living organism that processes a novel combination of genetic material obtained through the use of modern technology'. LMOs are also subjected to the risk analysis process, which should consider any hazardous, nutritional, or other safety concerns. In addition, the safety assessment should compare the food derived from modern biotechnology and its conventional counterpart to find any new or altered hazards, nutritional or other safety concerns [ 216 ]. Once the GMO is considered 'substantially equivalent' to its non-GMO counterpart, it is considered safe for food or feed consumption [ 217 ]. The LMO definition and the process involved in its development are the basis of many legal frameworks in the EU and other countries, such as Australia, New Zealand, Brazil, China, Japan, Saudi Arabia, and Thailand [ 218 ]. The EU probably has the strictest regulatory framework for conventional GM crop cultivation, which requires thorough environmental, human, and animal health safety assessments by the European Food Safety Authority (EFSA), as well as European Commission (EC) approval [ 207 , 219 ]. The insect-resistant corn, MON810 expressing Bt protein, Cry1Ab, was the only GM crop approved for cultivation in 1998 in the EU [ 220 ]. Despite positive recommendations from the EFSA, no other GM variants have been approved by the EC for cultivation, since most EU countries, including France and Germany, either partially or fully ban GMO cultivation [ 207 ]. Interestingly, despite its strong stance against GM crop cultivation, the EU is still one of the major importers of GM crops, including GM corn, cotton, soybean, canola, and sugar beet. The EU also depends on GM feed for their livestock industry, with an annual import of 30 million metric tons annually [ 221 ]. Pre-market assessments on transgene-stacked plants also vary between different countries. In the USA and Canada, a transgenic plant with stacked traits is not considered a new GMO event if the toxicological and allergenicity tests on the individual traits have been done before the stacking event [ 222 ]. However, in the EU and other countries, such as Argentina, Korea, and Japan, the same stacked gene product would still be considered a new and separate GMO event requiring an additional risk assessment, even if each trait had been pre-assessed before the stacking event [ 138 , 223 , 224 , 225 ]. As modern biotechnology tools become more complex, there have been renewed discussions on the current regulatory frameworks governing genetically engineered crop production. It has become crucial to acknowledge the different types of modification induced by SSN-based genome editing techniques within the LMO regulatory framework. Treating all types of SSN-mediated modifications within the same regulatory framework might hamper technological progress made in crop improvement. Among the three types of SSN-based modifications, Type-3 is the most similar to the conventional recombinant DNA technology which allows the replacement or addition of a whole gene, often foreign, in the plant host genome [ 226 ]. In contrast, Type-1 and Type-2 alter plant genomic sequences endogenously without the insertion of exogenous DNA, and, therefore, should not be regulated the same way as Type-3 modifications. Recently, the EFSA panels considered that the EU's Guidance for risk assessment of food and feed from genetically modified plants and the Guidance on the environmental risk assessment of genetically modified plants were sufficient but should not be applicable to all plants produced via Type-1 and Type-2 editing. In addition, the panel members did not find any new hazards linked to Type-1 and Type-2 modifications when compared with Type-3 modification and conventional breeding [ 227 ]. A combination of various biotechnological approaches will give rise to GM crops with various traits, each one with a different level of risk [ 228 ]. Risk assessments on SSN-based products should be conducted on a case-by-case basis with great transparency by considering the processes involved in their development [ 80 ]. Similarly, a transparent regulatory framework is vital in gaining public trust. However, most of the existing LMO regulatory frameworks do not provide any means of ensuring an acceptable level of transparency [ 229 ]. Strauss and Sax [ 228 ] suggested that a registry of all GM products entering the market should be created. So far, the USDA's Agricultural Marketing Service has published a list of GM crops and foods available throughout the world and for which the responsible agency must maintain a record (USDA Agricultural Marketing Service, 2021). Such registry should also include applications with differing regulatory statuses in various countries (e.g., SDN-1 in Argentina, Brazil, and the USA compared to the EU and New Zealand). Such transparency will help clarify the varying regulatory status of GM applications globally, improve international trade, and, hopefully, increase public understanding of GM products [ 229 ]. 9. Benefits of Agricultural Biotechnology 9.1. Improved Crop Yield and Efficient Land Use In the Green Revolution era, the widespread utilization of fertilizers and pesticides has significantly boosted agricultural production. Unfortunately, their usage as agricultural inputs has been reported to be a limiting factor towards the end of the Green Revolution era as the global yield has begun to plateau for some major cereal crops [ 230 ]. A prominent benefit of GM technology is its ability to increase crop yield within the same cultivation area with fewer inputs, mitigating the shrinking size of arable lands available for crop cultivation. According to the FAO, the arable land for food production per person will decrease from 0.24 ha in 2014 to only 0.18 ha in 2050. This does not include the additional usage of land to produce biofuel feedstock or the effects of urbanization or soil degradation [ 231 ]. Thus, there is a compelling need to produce higher agricultural yields by adopting GM technology to increase the food supply. The main impact of GM technology on improving crop yields has been through better weed control and reducing the damage caused by pests through the cultivation of herbicide-tolerant and insect-resistant crops. As a result, from 1996 to 2018, the application of GM technology has increased the global production of the main crops by producing an additional 498 million tons of corn, 278 million tons of soybean, 32.6 million tons of cotton lint, and 14.1 million tons of canola [ 232 ]. Without the cultivation of GM crops during this period, additional arable land of 12.3 million ha of soybeans, 8.1 million ha of corn, 3.1 million ha of cotton, and 0.7 million ha of canola would have been needed to cultivate the conventional crop equivalent [ 232 ]. This would have required the clearance of more areas from the tropical forests for cultivation and the use of more fertilizers, herbicides, and pesticides, as well as irrigation, to gain the same reported yield. The increase in crop yield is supported by a meta-analysis of 147 original studies from 1996 to 2014, which reported that GM technology had increased crop yields by 22%, with the yield gains larger for insect-resistant crops than for herbicide-tolerant crops [ 233 ]. Another interesting meta-analysis of data from 130 publications found that gene overexpression, or the ectopic expression of transporter genes or other gene types, in three major GM cereals (rice, wheat, and corn) had significantly increased the overall grain yield by 16.7% on average [ 234 ]. Studies on these GM crops have mainly focused on genes with probable essential roles in improving the nitrogen uptake efficiency of crops. One example is the overexpression of alanine aminotransferase ( AlaAT ) genes, which are responsible for the increase in nitrogen utilization efficiency (NUtE; the biomass or grain yield per unit of nitrogen uptake) in canola and rice [ 235 , 236 ]. Li et al. [ 234 ] further suggested that the increased yield in the GM crops might depend on the higher shoot biomass, nitrogen uptake efficiency (NUpE; plant roots capacity to acquire nitrogen from the soil), and partial factor productivity of nitrogen (PFPN; grain yield per unit of nitrogen applied in soil). In another meta-analysis study on peer-reviewed literature (from 1996 to 2016) on GM corn, the study found strong evidence of higher grain yield, ranging from 5.6 to 24.5%, higher than for the true non-GE or near-isogenic line [ 237 ]. The GM lines also contained lower concentrations of mycotoxins (~28.8%), fumonisin (~30.6%), and thricotecens (~36.5%) [ 237 ]. The evidence clearly shows the benefits of GM technology in improving crop yield and reducing the accumulation of harmful toxins in the grain. 9.2. Economic Benefits to Farmers and Consumers Through better management of weeds and pests and reduction in cost production, GM technology has significantly benefited farmers, with an additional gross income of USD 225.1 billion for the period 1996–2018 [ 232 ]. In 2018, most of the income benefits were earned by farmers in developing countries, where they received 53.7% of total income benefits, with an average of USD 4.41 received for each extra dollar invested in GM crop seeds [ 232 ]. This is consistent with previous studies showing that biotechnology applications in agriculture have brought economic benefits to numerous small-scale landholders in developing countries [ 238 , 239 ]. Moreover, GM technology not only benefits farmers and agribusinesses, but also consumers through lower costs of food supplies. It is conceivable that without the adoption of agricultural biotechnology that helped boost food supplies, commodity prices would have risen [ 240 ]. 9.3. Reduced Environmental Impacts of Agriculture The adoption of biotechnology in agriculture from 1996 to 2018 has lessened agriculture's environmental impact by reducing pesticide spraying by 775 million kg, representing a global reduction of 8.3% [ 241 ]. A meta-analysis demonstrated an overall reduction in chemical pesticide use by 37% from 1996 to 2014 due to biotechnology adoption [ 233 ]. The shift from conventional tillage (CT) to reduced tillage or no-tillage (RT/NT) farming systems in the cultivation of GM crops has resulted in a further reduction in fuel use by 12,799 million liters which have led to a significant reduction in global greenhouse gas (GHG) emissions of 34,172 million kg of carbon dioxide. Consequently, soil quality was enhanced by the retention of about 302,364 million kg of carbon dioxide [ 241 ]. 9.4. Increased Tolerance to Crop Diseases The global crop yield loss due to emerging and re-emerging pests and diseases is relatively high and was estimated to be 21.5%, 30%, 22.5%, 17.2%, and 21.4% for wheat, rice, corn, potato, and soybean, respectively [ 242 ]. The food-deficit regions of the Indo-Gangetic Plain and sub-Saharan Africa are reported to have suffered the highest crop losses [ 242 ]. Since the development of virus-resistant tobacco expressing the TMV coat protein [ 37 ], various biotechnological strategies have been applied to confer disease resistance in crops. These strategies include intervention in pathogen recognition/perception, pathogen effector binding, altering the expression of genes in plant defense signaling, targeting recessive resistance traits/susceptibility genes, interspecies transfer of dominant plant resistance genes, and utilization of antimicrobial peptides and RNAi [ 243 ]. One of the most successful stories of biotechnological application in crops to mitigate virus infection is the papaya ringspot virus (PRSV)-resistant papaya, which can resist PRSV infection through the expression of a coat protein from PRSV [ 244 ]. The development of the transgenic cultivar successfully averted the devastating loss of the papaya industry caused by PRSV in Hawaii [ 245 ]. 9.5. Nutrient Enhancement of Staple Crops Staple crops, such as rice, contain low levels of beneficial phytonutrients (nutraceuticals) and micronutrients, often below the recommended daily allowance [ 246 ]. In 2020, it was estimated that nearly 10% of the world's population (around 768 million people) were undernourished. More than half of all undernourished people live in Asia (418 million), while more than a third live in Africa (282 million) and a smaller proportion (60 million) in Latin America and the Caribbean [ 247 ]. Low- and middle-income countries rely more on starchy food staples, such as rice, banana ( Musa spp.), cassava ( Manihot esculenta ), and corn. However, the majority are deficient in beneficial phytonutrients and micronutrients. The adoption of biotechnology is believed to offer an effective and sustainable strategy to produce GM biofortified crops with specific nutrient-enriched content. This is particularly important in countries where the technology is urgently needed to help alleviate nutrient-deficiency-related illnesses [ 248 ]. There has been considerable progress in developing biofortified staple crops, predominantly via synthetic metabolic engineering [ 249 ]. The best-known example, and the first biofortified staple crop utilizing this method, is the β-carotene-enriched 'Golden Rice' [ 250 ]. The bioavailability of the β-carotene in rice, a precursor of provitamin A, could reduce vitamin A deficiency, which affects an estimated 190 million preschool-age children worldwide, of whom 91.5 million reside in Southeast Asia [ 251 ]. The Golden Rice was produced by introducing the entire β-carotene biosynthetic pathway through the multigene transformation of rice endosperm on two T-DNAs [ 51 ]. The first T-DNA carried the daffodil ( Narcissus pseudonarcissus ) phytoene synthase gene, NpPSY1 , and the bacterial ( Erwinia uredovora ) phytoene desaturase gene, EuCRT1 , controlled by an endosperm-specific glutelin (Gt1) promoter and constitutive cauliflower mosaic virus (CaMV) 35S promoter, respectively. The second T-DNA carried the daffodil lycopene β-cyclase, NpLYC-b gene under the control of a rice glutelin promoter and a selectable marker. While the β-carotene enhancement in rice was successful, the carotenoid concentration was only increased by 1.6 μg/g dry weight (DW) [ 51 ]. This prompted the production of 'Golden Rice 2′ (GR2), where the rice was engineered with corn ZmPSY and E. uredovora EuCRTI genes, both controlled by the native rice glutelin promoter [ 54 ]. The GR2 form contains a higher carotenoid accumulation of up to 23-fold (about 37 µg/g DW) than the original Golden Rice. Although GR2 is registered as safe in Australia, the USA, Canada, and New Zealand and possesses import approvals, the Philippines is the only country so far that has authorized the direct use of GR2 in food, feed, and processing. Since the production of GR2, increases in β-carotene levels through GM technology have been observed across an array of food crops, such as sorghum, corn, wheat, banana, and canola [ 252 ]. The successful generation of biofortified crops that involved the simultaneous expression of multigenes, with some generating multiple essential nutrients, was also reported. Enhancement of multiple vitamins, such as β-carotene, folate, and ascorbate, in the rice endosperm was achieved through the introduction of ZmPSY1 and Eu CRTI (carotenoid pathway), rice dehydroascorbate reductase, OsDHAR (ascorbate pathway), and folE (folate pathway) using an unlinked direct DNA transfer co-transformation strategy [ 253 ]. In a more recent study, multi-nutrient biofortified rice was developed by expressing Arabidopsis thaliana NICOTIANAMINE SYNTHASE1 ( AtNAS1 ), Phaseolus vulgaris FERRITIN ( PvFERRITIN ), bacterial CRT1, and ZmPSY in a single genetic locus that increased the levels of iron, zinc, and β-carotene content in the rice endosperm [ 254 ]. In an example that employed the multigene stacking strategy, the production of 'second generation' folate (Vitamin B 9 )-biofortified rice was achieved through simultaneous expression of four transgenes ( GTPCHI , ADCS , FPGS , and FBP ) [ 255 ]. Through this strategy, the folate content was increased significantly by 150-fold and has improved stability during post-harvest storage. Hence, this showed that the multigene stacking strategy is a highly efficient method for folate biofortification in rice, since the expression of a single transgene GTPCHI led only to about a 10-fold increase of folate concentration, whilst the co-expression of GTPCHI and ADCS resulted in a 100-fold folate enhancement [ 256 ]. 9.6. Production of Plant-Based Pharmaceuticals Global immunization coverage has declined from 86% in 2019 to 83% in 2020 due to the lack of access to immunization, aggravated by the straining of health systems due to the COVID-19 pandemic [ 257 ]. Children have been particularly affected, with the number of completely unvaccinated children increasing by 3.4 million in 2020 [ 257 ]. Vaccination coverage, specifically in underdeveloped countries, may be increased by developing a plant-based vaccine or 'edible vaccine'. This innovation offers an exciting alternative by delivering a vaccine that can be easily administered and stored without refrigerated conditions [ 258 ]. An edible vaccine is produced by integrating specific genes encoding the desired antigenic protein into the plant host genome. Once the plant-derived vaccine is consumed, the release of antigens will stimulate a specific autoimmune response via mucosal immunity. Various candidates for plant-derived vaccines using economically important crops are currently under development, such as potato [ 259 ] and banana [ 260 ] expressing hepatitis B vaccine, tomato expressing triple vaccines against shigellosis, anthrax, and cholera [ 261 ], spinach ( Spinacea oleraceae ) expressing HIV-1 vaccine [ 262 ] and carrot ( Daucus carota ) expressing Helicobacter pylori vaccine [ 263 ]. An identical technique has been applied in chloroplast transformation, allowing a much higher accumulation of antigen protein because of the multi-copy nature of the chloroplast genome compared to the single-copy nature of the nuclear genome [ 264 ]. 9.1. Improved Crop Yield and Efficient Land Use In the Green Revolution era, the widespread utilization of fertilizers and pesticides has significantly boosted agricultural production. Unfortunately, their usage as agricultural inputs has been reported to be a limiting factor towards the end of the Green Revolution era as the global yield has begun to plateau for some major cereal crops [ 230 ]. A prominent benefit of GM technology is its ability to increase crop yield within the same cultivation area with fewer inputs, mitigating the shrinking size of arable lands available for crop cultivation. According to the FAO, the arable land for food production per person will decrease from 0.24 ha in 2014 to only 0.18 ha in 2050. This does not include the additional usage of land to produce biofuel feedstock or the effects of urbanization or soil degradation [ 231 ]. Thus, there is a compelling need to produce higher agricultural yields by adopting GM technology to increase the food supply. The main impact of GM technology on improving crop yields has been through better weed control and reducing the damage caused by pests through the cultivation of herbicide-tolerant and insect-resistant crops. As a result, from 1996 to 2018, the application of GM technology has increased the global production of the main crops by producing an additional 498 million tons of corn, 278 million tons of soybean, 32.6 million tons of cotton lint, and 14.1 million tons of canola [ 232 ]. Without the cultivation of GM crops during this period, additional arable land of 12.3 million ha of soybeans, 8.1 million ha of corn, 3.1 million ha of cotton, and 0.7 million ha of canola would have been needed to cultivate the conventional crop equivalent [ 232 ]. This would have required the clearance of more areas from the tropical forests for cultivation and the use of more fertilizers, herbicides, and pesticides, as well as irrigation, to gain the same reported yield. The increase in crop yield is supported by a meta-analysis of 147 original studies from 1996 to 2014, which reported that GM technology had increased crop yields by 22%, with the yield gains larger for insect-resistant crops than for herbicide-tolerant crops [ 233 ]. Another interesting meta-analysis of data from 130 publications found that gene overexpression, or the ectopic expression of transporter genes or other gene types, in three major GM cereals (rice, wheat, and corn) had significantly increased the overall grain yield by 16.7% on average [ 234 ]. Studies on these GM crops have mainly focused on genes with probable essential roles in improving the nitrogen uptake efficiency of crops. One example is the overexpression of alanine aminotransferase ( AlaAT ) genes, which are responsible for the increase in nitrogen utilization efficiency (NUtE; the biomass or grain yield per unit of nitrogen uptake) in canola and rice [ 235 , 236 ]. Li et al. [ 234 ] further suggested that the increased yield in the GM crops might depend on the higher shoot biomass, nitrogen uptake efficiency (NUpE; plant roots capacity to acquire nitrogen from the soil), and partial factor productivity of nitrogen (PFPN; grain yield per unit of nitrogen applied in soil). In another meta-analysis study on peer-reviewed literature (from 1996 to 2016) on GM corn, the study found strong evidence of higher grain yield, ranging from 5.6 to 24.5%, higher than for the true non-GE or near-isogenic line [ 237 ]. The GM lines also contained lower concentrations of mycotoxins (~28.8%), fumonisin (~30.6%), and thricotecens (~36.5%) [ 237 ]. The evidence clearly shows the benefits of GM technology in improving crop yield and reducing the accumulation of harmful toxins in the grain. 9.2. Economic Benefits to Farmers and Consumers Through better management of weeds and pests and reduction in cost production, GM technology has significantly benefited farmers, with an additional gross income of USD 225.1 billion for the period 1996–2018 [ 232 ]. In 2018, most of the income benefits were earned by farmers in developing countries, where they received 53.7% of total income benefits, with an average of USD 4.41 received for each extra dollar invested in GM crop seeds [ 232 ]. This is consistent with previous studies showing that biotechnology applications in agriculture have brought economic benefits to numerous small-scale landholders in developing countries [ 238 , 239 ]. Moreover, GM technology not only benefits farmers and agribusinesses, but also consumers through lower costs of food supplies. It is conceivable that without the adoption of agricultural biotechnology that helped boost food supplies, commodity prices would have risen [ 240 ]. 9.3. Reduced Environmental Impacts of Agriculture The adoption of biotechnology in agriculture from 1996 to 2018 has lessened agriculture's environmental impact by reducing pesticide spraying by 775 million kg, representing a global reduction of 8.3% [ 241 ]. A meta-analysis demonstrated an overall reduction in chemical pesticide use by 37% from 1996 to 2014 due to biotechnology adoption [ 233 ]. The shift from conventional tillage (CT) to reduced tillage or no-tillage (RT/NT) farming systems in the cultivation of GM crops has resulted in a further reduction in fuel use by 12,799 million liters which have led to a significant reduction in global greenhouse gas (GHG) emissions of 34,172 million kg of carbon dioxide. Consequently, soil quality was enhanced by the retention of about 302,364 million kg of carbon dioxide [ 241 ]. 9.4. Increased Tolerance to Crop Diseases The global crop yield loss due to emerging and re-emerging pests and diseases is relatively high and was estimated to be 21.5%, 30%, 22.5%, 17.2%, and 21.4% for wheat, rice, corn, potato, and soybean, respectively [ 242 ]. The food-deficit regions of the Indo-Gangetic Plain and sub-Saharan Africa are reported to have suffered the highest crop losses [ 242 ]. Since the development of virus-resistant tobacco expressing the TMV coat protein [ 37 ], various biotechnological strategies have been applied to confer disease resistance in crops. These strategies include intervention in pathogen recognition/perception, pathogen effector binding, altering the expression of genes in plant defense signaling, targeting recessive resistance traits/susceptibility genes, interspecies transfer of dominant plant resistance genes, and utilization of antimicrobial peptides and RNAi [ 243 ]. One of the most successful stories of biotechnological application in crops to mitigate virus infection is the papaya ringspot virus (PRSV)-resistant papaya, which can resist PRSV infection through the expression of a coat protein from PRSV [ 244 ]. The development of the transgenic cultivar successfully averted the devastating loss of the papaya industry caused by PRSV in Hawaii [ 245 ]. 9.5. Nutrient Enhancement of Staple Crops Staple crops, such as rice, contain low levels of beneficial phytonutrients (nutraceuticals) and micronutrients, often below the recommended daily allowance [ 246 ]. In 2020, it was estimated that nearly 10% of the world's population (around 768 million people) were undernourished. More than half of all undernourished people live in Asia (418 million), while more than a third live in Africa (282 million) and a smaller proportion (60 million) in Latin America and the Caribbean [ 247 ]. Low- and middle-income countries rely more on starchy food staples, such as rice, banana ( Musa spp.), cassava ( Manihot esculenta ), and corn. However, the majority are deficient in beneficial phytonutrients and micronutrients. The adoption of biotechnology is believed to offer an effective and sustainable strategy to produce GM biofortified crops with specific nutrient-enriched content. This is particularly important in countries where the technology is urgently needed to help alleviate nutrient-deficiency-related illnesses [ 248 ]. There has been considerable progress in developing biofortified staple crops, predominantly via synthetic metabolic engineering [ 249 ]. The best-known example, and the first biofortified staple crop utilizing this method, is the β-carotene-enriched 'Golden Rice' [ 250 ]. The bioavailability of the β-carotene in rice, a precursor of provitamin A, could reduce vitamin A deficiency, which affects an estimated 190 million preschool-age children worldwide, of whom 91.5 million reside in Southeast Asia [ 251 ]. The Golden Rice was produced by introducing the entire β-carotene biosynthetic pathway through the multigene transformation of rice endosperm on two T-DNAs [ 51 ]. The first T-DNA carried the daffodil ( Narcissus pseudonarcissus ) phytoene synthase gene, NpPSY1 , and the bacterial ( Erwinia uredovora ) phytoene desaturase gene, EuCRT1 , controlled by an endosperm-specific glutelin (Gt1) promoter and constitutive cauliflower mosaic virus (CaMV) 35S promoter, respectively. The second T-DNA carried the daffodil lycopene β-cyclase, NpLYC-b gene under the control of a rice glutelin promoter and a selectable marker. While the β-carotene enhancement in rice was successful, the carotenoid concentration was only increased by 1.6 μg/g dry weight (DW) [ 51 ]. This prompted the production of 'Golden Rice 2′ (GR2), where the rice was engineered with corn ZmPSY and E. uredovora EuCRTI genes, both controlled by the native rice glutelin promoter [ 54 ]. The GR2 form contains a higher carotenoid accumulation of up to 23-fold (about 37 µg/g DW) than the original Golden Rice. Although GR2 is registered as safe in Australia, the USA, Canada, and New Zealand and possesses import approvals, the Philippines is the only country so far that has authorized the direct use of GR2 in food, feed, and processing. Since the production of GR2, increases in β-carotene levels through GM technology have been observed across an array of food crops, such as sorghum, corn, wheat, banana, and canola [ 252 ]. The successful generation of biofortified crops that involved the simultaneous expression of multigenes, with some generating multiple essential nutrients, was also reported. Enhancement of multiple vitamins, such as β-carotene, folate, and ascorbate, in the rice endosperm was achieved through the introduction of ZmPSY1 and Eu CRTI (carotenoid pathway), rice dehydroascorbate reductase, OsDHAR (ascorbate pathway), and folE (folate pathway) using an unlinked direct DNA transfer co-transformation strategy [ 253 ]. In a more recent study, multi-nutrient biofortified rice was developed by expressing Arabidopsis thaliana NICOTIANAMINE SYNTHASE1 ( AtNAS1 ), Phaseolus vulgaris FERRITIN ( PvFERRITIN ), bacterial CRT1, and ZmPSY in a single genetic locus that increased the levels of iron, zinc, and β-carotene content in the rice endosperm [ 254 ]. In an example that employed the multigene stacking strategy, the production of 'second generation' folate (Vitamin B 9 )-biofortified rice was achieved through simultaneous expression of four transgenes ( GTPCHI , ADCS , FPGS , and FBP ) [ 255 ]. Through this strategy, the folate content was increased significantly by 150-fold and has improved stability during post-harvest storage. Hence, this showed that the multigene stacking strategy is a highly efficient method for folate biofortification in rice, since the expression of a single transgene GTPCHI led only to about a 10-fold increase of folate concentration, whilst the co-expression of GTPCHI and ADCS resulted in a 100-fold folate enhancement [ 256 ]. 9.6. Production of Plant-Based Pharmaceuticals Global immunization coverage has declined from 86% in 2019 to 83% in 2020 due to the lack of access to immunization, aggravated by the straining of health systems due to the COVID-19 pandemic [ 257 ]. Children have been particularly affected, with the number of completely unvaccinated children increasing by 3.4 million in 2020 [ 257 ]. Vaccination coverage, specifically in underdeveloped countries, may be increased by developing a plant-based vaccine or 'edible vaccine'. This innovation offers an exciting alternative by delivering a vaccine that can be easily administered and stored without refrigerated conditions [ 258 ]. An edible vaccine is produced by integrating specific genes encoding the desired antigenic protein into the plant host genome. Once the plant-derived vaccine is consumed, the release of antigens will stimulate a specific autoimmune response via mucosal immunity. Various candidates for plant-derived vaccines using economically important crops are currently under development, such as potato [ 259 ] and banana [ 260 ] expressing hepatitis B vaccine, tomato expressing triple vaccines against shigellosis, anthrax, and cholera [ 261 ], spinach ( Spinacea oleraceae ) expressing HIV-1 vaccine [ 262 ] and carrot ( Daucus carota ) expressing Helicobacter pylori vaccine [ 263 ]. An identical technique has been applied in chloroplast transformation, allowing a much higher accumulation of antigen protein because of the multi-copy nature of the chloroplast genome compared to the single-copy nature of the nuclear genome [ 264 ]. 10. Concerns about the Effects of Agricultural Biotechnology on Human Health and the Environment 10.1. Effects on Human Health Toxicity, allergenicity, and unintended genetic effects are three common concerns associated with GM food consumption. The causative factors may include: (i) integration of the transgene and the expressed protein per se, (ii) secondary and pleiotropic effects of the expressed gene, and (iii) insertional mutagenesis in the modified organism [ 46 , 265 ]. A well-known GM product linked to several allergenicity reports was 'Starlink' corn, which incorporated the Bt gene that produces an insecticidal Cry9c protein [ 266 ]. Due to the concern about the allergenic potential of Cry9c protein in humans, the EPA restricted the use of Starlink corn in animal feed consumption in 1998. However, its residues were later found in food products, where the Starlink was intermixed with corn in the food chain. The EPA received several reports of allergic reactions to corn products that may have contained Starlink, prompting the voluntary recall of several food products containing Starlink corn in 2000 [ 267 , 268 ]. However, subsequent epidemiological investigations by The Centers for Disease Control and Prevention (CDC) concluded that there was no evidence the allergenic reactions were associated with hypersensitivity to the Cry9c protein [ 269 ]. Secondary and pleiotropic effects through the expression of introduced transgenes, many of which encode enzymes, are harder to detect and recognize since there are no direct tests available. Such pleiotropic effects could be manifested as an alteration of normal flow rates of metabolites (decreases or increases) in the metabolic pathways of the modified organism [ 270 ]. However, the unintended effects of pleiotropy may be predicted based on available knowledge of the biochemical mechanism of the encoded protein. The likelihood of pleiotropy increases if the transgenes are involved in basic cellular functions or multiple biochemical pathways [ 270 ]. Another concern is the possible unintentional mutagenic effect outside the transgene insertion site. The disruption could possibly result in the inactivation of endogenous genes, or activation of otherwise silent genes that, in turn, may generate secondary toxic compounds, and the generation of fusion proteins [ 46 ]. While it is straightforward to test and confirm the direct effect of the transgene expression as shown with the Starlink Bt protein, it is harder to identify the causal connection between the transgene insertion and the toxic effect, particularly when the metabolic activation of the toxins are several steps away from the transgene insertion. Due to concerns about the possible transgene effects in transformed plants, the production of transgene-free plants is highly desirable but remains a challenge in plant biotechnology [ 271 ]. Current strategies to remove or prevent the integration of gene editor constructs in the CRISPR/Cas9 system involve the removal of transgenes via genetic segregation, transient editor expression from DNA vectors, and DNA-independent editor delivery [ 271 ]. A common genetic segregation system usually uses the target-site genotyping approach to isolate transgenic progenies from a population of transformed plants, followed by a counter-selection strategy using visible selection markers. In rice, a CRISPR/Cas9 containing a CYP81A6-hpRNAi expression cassette was used to produce bentazon (herbicide)-hypersensitive plants [ 272 ]. The T 1 edited plants were then sprayed with bentazon, which killed the transgenic plants containing the transgenes, while the transgene-free rice seedlings grew normally [ 272 ]. Elimination of the transgene using marker gene selection is efficient. However, it raises other concerns (i.e., biosafety and biosecurity issues) regarding the use of antibiotic or herbicide-resistant transgenes in transformed crops. In the CRISPR/Cas system, transformed plants cannot survive without the transgene, requiring the screening of several generations to evaluate the presence of the transgene. Another widely used method in detecting transgene-free crops is the use of fluorescence-mediated monitoring to obtain transgene-free homozygous gene-edited crop plants [ 273 ]. In Arabidopsis, a CRISPR/Cas9 construct harboring the mCherry-expressing (red fluorescence) gene was used to visualize edited progenies in the T 1 generation. This technique may save time in identifying transgene-free plants, which usually requires the laborious work of growing, extracting, and genotyping plant genomic DNA [ 274 ]. Another interesting approach is to use the transgene killer CRISPR (TKC) system, which allows automatic transgene elimination from the edited plants. In rice, a CRISPR/Cas9 construct harboring the self-eliminating genes, barnase and CMS2 , which express toxic proteins during specific stages of plant development, was used to efficiently isolate transgene-free plants [ 275 ]. Barnase is driven by the host rice plant promoter, REG2 , specifically during the embryogenic stage, while CMS2 interrupts mitochondrial functions in the callus and vegetative stages. This strategy allows the isolation of transgene-free plants during different growth stages of T 0 plants [ 275 ]. 10.2. Long Term Effects on Genetic Diversity A wide genetic pool is vital for plant species to adapt to environmental changes. Classical and modern plant breeding techniques apply this knowledge by mixing different crop variants to create novel varieties with superior traits that would survive harsher conditions or produce higher yields. However, an estimated 75% of plant genetic diversity had diminished during the post-Green Revolution era as farmers around the world cultivated only a handful of high-yielding crop varieties in a mono-crop farming system to maximize profit [ 276 ]. The high cost of GM crop development imposed by the national and international regulatory processes has further narrowed down the biotechnological application to only high-value crops, such as soybean, corn, and cotton [ 228 , 277 ]. Over a long period, intensive cultivation, focusing on only a small number of crop species, will significantly affect the genetic pool and diversity, which is crucial for plant adaptation against climate change and virus and pest infestation [ 278 , 279 ]. 10.3. Over-Reliance on Mega-Companies in the Agricultural Market A major concern regarding GM crop cultivation is the over-reliance of farmers on large agro-biotech firms for patented seeds and agricultural chemicals. The reliance of farmers on these firms is compounded by the shrinking number of competitors in the market and the perceived oligopoly practice of a small group of mega-companies in the global agricultural market [ 280 ]. The merging of Bayer and Monsanto, for example, has resulted in Bayer owning 29% of seeds and 24% of pesticides in the global agriculture market in 2016 [ 281 ]. However, a meta-analysis of 15 studies demonstrated that two-thirds of profits of first-generation GM crops were shared downstream, while one-third accrued upstream by agribusiness companies [ 282 ], demonstrating that seed suppliers and large firms did not entirely dominate the profits. 10.1. Effects on Human Health Toxicity, allergenicity, and unintended genetic effects are three common concerns associated with GM food consumption. The causative factors may include: (i) integration of the transgene and the expressed protein per se, (ii) secondary and pleiotropic effects of the expressed gene, and (iii) insertional mutagenesis in the modified organism [ 46 , 265 ]. A well-known GM product linked to several allergenicity reports was 'Starlink' corn, which incorporated the Bt gene that produces an insecticidal Cry9c protein [ 266 ]. Due to the concern about the allergenic potential of Cry9c protein in humans, the EPA restricted the use of Starlink corn in animal feed consumption in 1998. However, its residues were later found in food products, where the Starlink was intermixed with corn in the food chain. The EPA received several reports of allergic reactions to corn products that may have contained Starlink, prompting the voluntary recall of several food products containing Starlink corn in 2000 [ 267 , 268 ]. However, subsequent epidemiological investigations by The Centers for Disease Control and Prevention (CDC) concluded that there was no evidence the allergenic reactions were associated with hypersensitivity to the Cry9c protein [ 269 ]. Secondary and pleiotropic effects through the expression of introduced transgenes, many of which encode enzymes, are harder to detect and recognize since there are no direct tests available. Such pleiotropic effects could be manifested as an alteration of normal flow rates of metabolites (decreases or increases) in the metabolic pathways of the modified organism [ 270 ]. However, the unintended effects of pleiotropy may be predicted based on available knowledge of the biochemical mechanism of the encoded protein. The likelihood of pleiotropy increases if the transgenes are involved in basic cellular functions or multiple biochemical pathways [ 270 ]. Another concern is the possible unintentional mutagenic effect outside the transgene insertion site. The disruption could possibly result in the inactivation of endogenous genes, or activation of otherwise silent genes that, in turn, may generate secondary toxic compounds, and the generation of fusion proteins [ 46 ]. While it is straightforward to test and confirm the direct effect of the transgene expression as shown with the Starlink Bt protein, it is harder to identify the causal connection between the transgene insertion and the toxic effect, particularly when the metabolic activation of the toxins are several steps away from the transgene insertion. Due to concerns about the possible transgene effects in transformed plants, the production of transgene-free plants is highly desirable but remains a challenge in plant biotechnology [ 271 ]. Current strategies to remove or prevent the integration of gene editor constructs in the CRISPR/Cas9 system involve the removal of transgenes via genetic segregation, transient editor expression from DNA vectors, and DNA-independent editor delivery [ 271 ]. A common genetic segregation system usually uses the target-site genotyping approach to isolate transgenic progenies from a population of transformed plants, followed by a counter-selection strategy using visible selection markers. In rice, a CRISPR/Cas9 containing a CYP81A6-hpRNAi expression cassette was used to produce bentazon (herbicide)-hypersensitive plants [ 272 ]. The T 1 edited plants were then sprayed with bentazon, which killed the transgenic plants containing the transgenes, while the transgene-free rice seedlings grew normally [ 272 ]. Elimination of the transgene using marker gene selection is efficient. However, it raises other concerns (i.e., biosafety and biosecurity issues) regarding the use of antibiotic or herbicide-resistant transgenes in transformed crops. In the CRISPR/Cas system, transformed plants cannot survive without the transgene, requiring the screening of several generations to evaluate the presence of the transgene. Another widely used method in detecting transgene-free crops is the use of fluorescence-mediated monitoring to obtain transgene-free homozygous gene-edited crop plants [ 273 ]. In Arabidopsis, a CRISPR/Cas9 construct harboring the mCherry-expressing (red fluorescence) gene was used to visualize edited progenies in the T 1 generation. This technique may save time in identifying transgene-free plants, which usually requires the laborious work of growing, extracting, and genotyping plant genomic DNA [ 274 ]. Another interesting approach is to use the transgene killer CRISPR (TKC) system, which allows automatic transgene elimination from the edited plants. In rice, a CRISPR/Cas9 construct harboring the self-eliminating genes, barnase and CMS2 , which express toxic proteins during specific stages of plant development, was used to efficiently isolate transgene-free plants [ 275 ]. Barnase is driven by the host rice plant promoter, REG2 , specifically during the embryogenic stage, while CMS2 interrupts mitochondrial functions in the callus and vegetative stages. This strategy allows the isolation of transgene-free plants during different growth stages of T 0 plants [ 275 ]. 10.2. Long Term Effects on Genetic Diversity A wide genetic pool is vital for plant species to adapt to environmental changes. Classical and modern plant breeding techniques apply this knowledge by mixing different crop variants to create novel varieties with superior traits that would survive harsher conditions or produce higher yields. However, an estimated 75% of plant genetic diversity had diminished during the post-Green Revolution era as farmers around the world cultivated only a handful of high-yielding crop varieties in a mono-crop farming system to maximize profit [ 276 ]. The high cost of GM crop development imposed by the national and international regulatory processes has further narrowed down the biotechnological application to only high-value crops, such as soybean, corn, and cotton [ 228 , 277 ]. Over a long period, intensive cultivation, focusing on only a small number of crop species, will significantly affect the genetic pool and diversity, which is crucial for plant adaptation against climate change and virus and pest infestation [ 278 , 279 ]. 10.3. Over-Reliance on Mega-Companies in the Agricultural Market A major concern regarding GM crop cultivation is the over-reliance of farmers on large agro-biotech firms for patented seeds and agricultural chemicals. The reliance of farmers on these firms is compounded by the shrinking number of competitors in the market and the perceived oligopoly practice of a small group of mega-companies in the global agricultural market [ 280 ]. The merging of Bayer and Monsanto, for example, has resulted in Bayer owning 29% of seeds and 24% of pesticides in the global agriculture market in 2016 [ 281 ]. However, a meta-analysis of 15 studies demonstrated that two-thirds of profits of first-generation GM crops were shared downstream, while one-third accrued upstream by agribusiness companies [ 282 ], demonstrating that seed suppliers and large firms did not entirely dominate the profits. 11. Conclusions As agricultural production approaches a bottleneck due to limited arable lands, extreme weather, and increasing food demand [ 9 ], novel tools are needed to produce more resilient, efficient, and high-yielding crops to ensure global food security. Modern biotechnology tools with improved specificity and efficiency could eventually become the main driver of agricultural improvement, overcoming the limitations of conventional practices in improving crops. Rapid emergence and innovations in SSN-based editing tools especially show promising potential for crop improvement to boost agriculture productivity. Ultimately, transparency during the development, risk assessment, and regulatory process relating to modern biotechnology tools and the resulting products are crucial to further improve their applications in agriculture and increase public trust. Will the public accept SSN-based products more than GM products using DNA recombinant technology? Will modern biotechnology tools increase the pace of crop breeding substantially to improve food security? The answers to these questions will determine the fate of modern biotechnology tools in the agricultural sector and the lives of billions of people in the near future.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8482028/

Smallpox

EARLY VACCINES Edward Jenner, the country doctor from Berkeley, Gloucester, England, is recognized as the father of smallpox vaccination. Jenner's 1796 observations, that cowpox protected against smallpox when scratched into the skin of recipients, were written up in detail and presented to the Royal Society of England in 1798 and promoted widely in letters [3]. Yet Jenner was not the first to make these observations. Some historians note that John Fewster and others, as early as 1768, living near Jenner in Thornbury, and Benjamin Jesty in 1774 in Westminster, United Kingdom, observed the benefits of cowpox inoculation for protecting humans against smallpox [4]; these observations were not documented and disseminated, however, so they remain in obscurity. Jenner promoted vaccines in England and elsewhere by letters and speeches and by giving vaccinations gratis to local residents at the "vaccine hut" outside his home (The Chantry). There was immediate fierce opposition by persons who believed that biological products from cows would result in growths resembling cows on the bodies of recipients. People of influence had an early role in supporting both inoculation and vaccination. Lady Mary Wortley Montagu, wife of the UK Ambassador to Turkey, who had had smallpox in England, observed variolators in Turkey performing inoculations. She was so impressed that she promoted the procedure via a series of letters starting in 1717 [1, 2 ]. Thomas Jefferson, Benjamin Franklin, and Benjamin Waterhouse, the latter of Harvard University, were early advocates of vaccination in the United States. For the next 100 years, technical problems tied to vaccine quality impeded the successful use of vaccine globally. Well into the 20th century, the major challenges were growing adequate quantities of vaccine of measurable potency, sterility, and durability despite differences in temperature, climate, and humidity [5]. Mode of administration was another challenge. Different scratch and inoculation techniques were used, particularly in India [ 6 ]. Throughout much of the 1800s vaccine was passed from arm to arm or dried and put on small "points" (sharp objects of ivory, steel). These methods were not reliable. Toward the end of the 19th century, animals, especially the skin of living cows, were used to grow the virus used for vaccination [5]. During the 19th century, arm-to-arm vaccination was the standard method of maintaining the product's efficacy, even during long voyages. Some practitioners put threads through the pustular material. The threads were dried and sent to the areas for populations to be vaccinated; potency certainly waned during such travel. On long sea voyages, groups of orphan children were often sent specifically to assure arm-to-arm transfer of the pustular material. In the early 1900s, an attempt to dry and preserve vaccine for shipment from France to their colonies in West Africa was described by Fasquelle and Fasquelle [7]. VACCINATION AFTER 1900 Pustular material from cows or patients with pustular disease of indeterminate origin was used for more than a century as the source of smallpox vaccine. By the beginning of the 1900s, vaccination against smallpox was being practiced in most industrialized countries. The virus now used, called vaccinia, has an obscure origin. The product may have originally been a hybrid between cowpox virus and variola virus or some other orthopoxvirus by serial passage in artificial conditions, or, as Baxby posits, vaccinia may be a laboratory survivor of a virus now extinct [ 8 ], p 214]. The various vaccinia strains globally are similar to each other genetically but differentiated from other poxviruses, including cowpox and variola viruses, by DNA mapping. By the 1950s there was improvement in vaccine quality, distribution, and public health infrastructure. Smallpox was virtually eliminated from Europe and North America by that time. Since Jenner's time opponents of vaccination have based their concerns on perceived physical harm from the procedure and breaching of individual rights. Over time, vaccination has been considered a public health good and inserted into law in the United States and elsewhere and upheld by the Supreme Court [9]. ERADICATION STRATEGY EVOLUTION High vaccination coverage had been the strategy of national and international smallpox control and elimination strategies since Jenner's findings slowly spread worldwide and became accepted in the 1800s. However, the continued existence of the disease on virtually all continents was due to fragmented and inadequate health systems. Access to remote populations was impossible in many areas and acceptance of evolving vaccine production and delivery technology was slow. Most importantly, the colonial legacy starting in the late 1800s left many areas of the world dependent on European control and resources for their health and other programs, particularly in Africa. Conservation of liquid vaccine produced mainly on cows was very difficult, because refrigeration was virtually nonexistent in the tropics until the mid- to late-1900s. In West and Central Africa and India more vaccinations were given than the censused population, yet smallpox raged because of poor vaccine quality. Massive epidemics of smallpox appeared periodically in virtually all tropical countries and several areas of temperate countries well into the 1900s, fueled by high levels of susceptibility as a result of new births, those who received poor quality vaccine, and nonimmune older persons [10]. The World Health Organization (WHO) was formed in 1948 with a mandate to develop public health policies and to coordinate surveillance, and some control and eradication initiatives. By the 1950s, many countries had passed public health laws and implemented smallpox vaccination programs, many of which were successful, particularly in the northern hemisphere. In 1959, the representative of the Soviet Union proposed a resolution for a global smallpox eradication program to the World Health Assembly [ 2 ]; this was based, in part, on the outbreaks of smallpox in several of the southern republics, which underscored the priority for development of a potent vaccine to control the outbreaks. In addition, the Soviets wished to provide vaccines to the WHO as a gift to the global program. Yet, little progress was made over the next 6 years toward global eradication [ 11 ], p 334]. Between 1959 and 1966, few funds were received or invested by WHO for smallpox eradication, and few staff were assigned to the program. The strategy was entirely reliant on attempting to achieve 80% vaccination coverage. Each country had to rely on its own manufacturers or products acquired via the WHO, mainly from Soviet donations. The Soviet product caused severe adverse reactions, which probably was responsible for poor acceptance and coverage, especially in India, where tens of millions of doses were sent [ 2 ]. Despite the 1959 resolution, WHO internal and external support for the program languished until the middle 1960s. In 1966, the United States, backed by President Lyndon Johnson, supported another resolution in support of an intensified smallpox eradication program [ 11 ], p 334]. ERADICATION AND VACCINE QUALITY By the late 1960s, when the intensified global smallpox eradication program began, it was found by WHO, that vaccine was being produced by many different countries using varying procedures. A major step was taken initially by the Smallpox Eradication Unit at WHO to advise standardized production methods and international quality control of vaccines used in the global program [ 2 ]. One of the first steps of the intensified program was to do a detailed survey of vaccine production procedures, quality, and production capacity in endemic and nonendemic countries. Questions about the relatively newly perfected freeze-dried vaccine methods, strains used, methods of growing virus, and bottling (doses per vial) were assessed [ 2 , 12 ]. Of 72 laboratories assessed, 59 replied to the WHO survey. Fifty-one laboratories (86.4%) harvested vaccinia virus from the skin of calves or sheep, and 6 (10.1%) from water buffaloes; 3 also reported using chick embryos, and 3 used tissue culture. Of the 59 laboratories, 23 (39.0%) used Lister strain (origin UK), 6 (10.2%) New York City Board of Health strain, 7 (11.9%) Paris strain, and 22 (37.3%) a variety of strains; one reported using a mixture of vaccinia and cowpox. VACCINE POTENCY, STABILITY, AND BACTERIAL CONTENT IN 1967 When the WHO requested laboratories producing smallpox vaccines to submit information on potency, heat stability and bacterial content, the following was received from 59 laboratories from 4 WHO regions. Of the 59, 31 (52.5%) reported that all 3 recent product lots were satisfactory (titers of ≥10 8.0 pock-forming units on chicken chorioallantoic membranes), and 16 (27.1%) reported vaccine stability after 4 weeks at 37°C; 12 of 53 laboratories (22.6%) sending data reported bacterial counts >500/mL, which were unacceptable by WHO standards. After the establishment of the independent WHO reference centers for smallpox vaccine testing at Connaught Laboratories in Toronto and the National Institute of Public Health in Bilthoven, the Netherlands, it was concluded that, of 39 batches submitted by producers intending to develop freeze-dried vaccine for use in their own countries and in the global programs, 25 (64.1%) failed to meet standards. The conclusion was that, in 1967, not more than 10% of the vaccine in use in endemic countries met WHO requirements. Freeze-dried vaccine, a procedure credited to Leslie Collier, rendered the vaccine stable for long periods and could be reconstituted with diluent in the field [ 12 ]. The WHO smallpox unit established (1) a manual on the production of freeze-dried vaccine, (2) a traveling set of experts who gave seminars on vaccine production in laboratories, (3) training that including hands-on demonstration of the production of reference smallpox vaccines, (4) provision of seed lots of Lister strain vaccinia, (5) development of a heat stability test, and (6) guidance for regular testing of vaccine potency and heat stability to be used by the reference centers. ADMINISTRATION OF VACCINE Another reason earlier smallpox campaigns failed was the inadequacy of the instruments used to immunize. Baxby has reviewed the variety of instruments, many of which resembled "tools of torture" [ 13 ]. Some of these required large amounts of liquid vaccine and caused maceration of the skin. This often resulted in infections in the tissues of recipients, poor success rates, and refusal of the procedure. The 2 most effective tools for injecting vaccinia virus intradermally during the eradication program were the bifurcated needle and the jet injector gun ( Figures 1 and 2 ). The bifurcated needle is a 2-pronged adaptation of a sewing needle, invented by Benjamin Rubin of Wyeth Laboratories. The sterile needle was dipped into a reconstituted vial of vaccine; a drop of vaccine was caught between the prongs. The needle is jabbed rapidly 15 times into the upper deltoid region of the arm until a small drop of blood or serum appears. The jabs should all be within a 1-cm-diameter area. Acetone is preferable to alcohol for cleansing the arm because it dries quickly; alcohol could inactivate the vaccinia virus if not dry when the multiple punctures are made. The needles are kept in a durable tubular container with a hole to shake out a sterile metal needle when needed. The needles are cleaned after use, placed in a tub of boiling water for 20 minutes, cooled, and replaced in the plastic containers, ready for reuse. Figure 1. Bifurcated needle for intradermal injection of vaccinia virus, invented by Benjamin Rubin (left) of Wyeth Laboratories. (Source: Fenner et al [ 2 ], WHO.) Figure 2. Jet injector for intradermal injection of vaccinia virus. Top left, Intradermal nozzle squirts at an angle for smallpox vaccination. Top right, Jet Injector in its case. The instruction book for West Africa was in English and French. Bottom right, Aaron Ismach, who is credited with the design of this injector. (Source: WHO, CDC) The jet injector is a pneumatic foot-activated apparatus (gun) that injects smallpox vaccine intradermally via a special nozzle. It was most effective in places where large groups of people could be assembled, such as Brazil and West and Central Africa. Aaron Ismach of the US Army is credited with the design. The apparatus has been criticized for requiring frequent maintenance and spare parts; however, in Guinea, for example, 6 vaccination teams averaged close to 2000 vaccinations per working day over a 2-year period in a rural environment where people could be assembled; meticulous attention was given to daily maintenance [14]. Not only was coverage high with both of these devices during the smallpox eradication campaign, but the success ("take") rates after vaccination were >98% in primary vaccines and 95% in those receiving revaccination (see Figure 3 ). Figure 3. Skin reaction after primary vaccination and late revaccination (several years later), performed using a multiple puncture method or a jet injector. (Source: CDC.) Smallpox vaccine multiplies in the skin's epithelium, producing a slight fever and characteristic skin reaction with redness and induration leading to a pustule by about day 7 in those receiving primary vaccines; this "Jennerian pustule" usually starts to crust and desquamate by day 14, leaving a scar. Induration may occur after revaccination; if many years have passed since vaccination, a typical Jennerian vesicle will occur ( Figure 3 ). VACCINE COMPLICATIONS During the intensified eradication program, a major effort was made to assess take rates. Only in the United States was a comprehensive nationwide survey done to look at adverse events following smallpox vaccination with the NY Board of Health seed strain. In 1970 and 1971, Lane et al published articles on complications, dividing United States recipients into those receiving smallpox vaccine for the first time (primary vaccinees) and those receiving the vaccine as "revaccinees" [ 15 , 16, 17]. Data from the 10-state survey were collected by more active ascertainment of adverse events, resulting in 5 times the number of complications reported by the passively reported events [ 15 ]. The most severe conditions were postvaccinal encephalitis and generalized vaccinia ( Figure 4 ). The death rate was about 1 in 1 million vaccinations in those receiving primary vaccines, mainly young children. Children with immunoglobulin deficiencies or severe eczema were prone to adverse events. Screening before vaccination could have decreased the number of complications. Figure 4. Accidental autoinoculation of cheek with vaccinia virus, approximately 5 days old. Primary take with "Jennerian vesicle" on arm, 10–12 days old. (Photograph courtesy of John M. Leedom, MD, CDC collection.) In 1971, the US Advisory Committee on Immunization Practices recommended that routine smallpox vaccination be stopped in the United States. This was based on the assessment of the risks of those being vaccinated and the extremely low chance of importation, even though smallpox remained endemic in East Africa and the Indian subcontinent [18, 19]. Following the certification of smallpox eradication by the World Health Assembly in 1980, the WHO advised all countries to stop routine smallpox vaccination [ 2 ]. This was accepted by virtually all countries in the early 1980s. In 2002–2003, after the terrorist airplane attacks on the Twin Towers buildings in New York City and the anthrax mailings and deaths due to a domestic bioterror event, a limited number of smallpox vaccinations were given in the United States, mainly to first responders. The US military was vaccinated, and acute myopericarditis and cardiac arrhythmias developed in 37 of >400†000 recruits [17, 18]. SMALLPOX: THE DISEASE There were 2 major manifestations of smallpox: Variola major and Variola minor. V. major was seen mainly in the Indian subcontinent, and parts of Africa and Asia during the eradication program and was the most severe form, with a 30% fatality rate. V. minor, observed in east of Africa and Latin America, was milder, with fewer lesions and a case fatality rate of <5%. The clinical and epidemiological features of smallpox have been well covered in recent and past literature and are summarized in Tables 1 and 2 [ 2 , 20 ]. The eruption evolves with all skin lesions at the same stages at a given point of time, starting from macules and papules, followed by pustules, vesicles, and finally crusts, over a 10–20-day period. Table 1. Clinical and Epidemiological Features of Smallpox and Considerations for Certification Feature Indications of Smallpox Incubation period 14–17 d Syndrome visible Yes (eruption) Recognized by public Yes (eruption) Asymptomatic carriers None Transmission mode Respiratory Vector/reservoir Human Secondary attack rate among susceptible persons High (40%–90%) Table 2. Characteristics of Surveys to Certify Eradication of Smallpox Characteristic Comment Clinical features surveyed Fever and rash; facial pockmarks Target group School-age children Diagnostic alternatives Chickenpox, monkeypox Environmental surveys for microbe No Laboratory confirmation Yes Rumor registers Yes Reward for reporting cases Yes Minimum interval since last case before certification 2 y Continuing research Yes The lesions of smallpox are focused on the peripheral (centrifugal) parts of the body ( Figure 5 ), in contrast to the rashes of diseases like chickenpox which have centripetal distribution. The most common diagnostic dilemma is in the differential diagnosis is chickenpox which can be severe in older persons and immunocompromised patients [ 20 ]. Today, in Africa the eruption of human monkeypox cannot be easily distinguished from smallpox except by laboratory testing; patients with monkeypox often have cervical and inguinal lymphadenopathy [21]. Figure 5. Smallpox on day 8 of eruption (World Health Organization photograph). THE ERADICATION PROGRAM The initial strategy of the intensified smallpox eradication program beginning in 1967 was based on at least 80% vaccination coverage of the population in each country. This strategy, while successful in northern countries where vaccine quality was monitored and people were told to be revaccinated every 3 years, was not effective in the heavily endemic areas. These countries were mainly in Africa, Latin America, and Asia. In Africa, the health infrastructure was poor, because many countries became newly independent in the late 1950s and 1960s and had limited resources. Some countries had a tradition of mobile health teams trained to detect perils, such as sleeping sickness, leprosy, and onchocerciasis, and to give smallpox vaccines. The countries did not often have managerial expertise or refrigerated repositories to conserve vaccines and other heat-sensitive biologicals [22, 23]. In the mid-1960s President Lyndon Johnson committed the United States to supporting an 18-country West and Central Africa smallpox eradication-measles control program. This followed a visit to the United States by the minister of health of Burkina Faso (then Upper Volta), who learned about the newly developed measles vaccine and wanted Africa to benefit. The funding came via the US Agency for International Development. The US Centers for Disease Control and Prevention (CDC; then the Communicable Disease Center) was the implementing organization. Of the 11 countries with the highest incidence of smallpox, 7 were in West and Central Africa. Two major innovations in this program have changed the face of public health. The first was the use of operational specialists from the CDC—managers responsible for organization, finances, equipment, supplies, logistics, transportation—who were assigned to countries with medical epidemiologists, both of whom worked closely with national counterparts [24], p 141]. Second, the concept of surveillance containment was rediscovered and refined in Nigeria [ 25 ]. Intensified surveillance and ring vaccination became the major strategies used throughout the program, especially in the Indian subcontinent [ 2 , 26, 27]. SURVEILLANCE CONTAINMENT AND RING VACCINATION Using epidemiological information, rumor notices, and village-by-village searches, persons with suspected smallpox were identified and confirmed virologically when outbreaks were few. Patients were isolated in or near their home residence; food was supplied, and a 24-hour guard hired by the program to assure the patient did not circulate until the crusts had fallen. All primary contacts of the patient since the illness began were identified and vaccinated, as were residents of neighboring houses and villages within 5 km. Detailed maps and censuses of houses and residents in the villages within 5 km of the patient were made and used to assure all occupants had been vaccinated. Cross-notification was done, by telegram or phone, to health authorities both from places where case patients had probably acquired their disease and from places they had visited since their infections first manifested. Secondary areas of priority (nearby villages, markets, schools, assembly areas) were identified on hand-drawn maps and were visited, and residents were interrogated and vaccinated. This strategy is also called "ring vaccination," as concentric circles or areas of priority were often mapped by program authorities. In the Indian subcontinent, the active case search approach was highly refined. At one time 150†000 field workers were going from village to village [26, 27]. Major increases in numbers of infected villages and cases were found, compared with routine reporting. The largest exportation of cases in the program occurred in India. The Tatanagar railway station was the source of dozens of cases in surrounding states and districts until the case tracking and containment strategy was intensified, with major assistance from Tata industries that joined the program with staff and funds [27]. In Bangladesh, it was found by facial pockmark surveys that <5% of the actual cases were being reported before the program began [ 28 ]. The last case caused by naturally occurring transmission of Variola major occurred in Bangladesh in October 1975, and the last of disease Variola minor in Somalia on 26 October 1977 ( Figures 6 and 7 ). In 1978, an outbreak of smallpox occurred in Birmingham, United Kingdom, associated with a laboratory working with poxvirus variola virus: the ducting system connecting the laboratory to a photographer's office above was the conduit of contamination [ 2 ]. Figure 6. Last case patient with Variola major smallpox (Bangladesh, 1975). (Source: World Health Organization.) Figure 7. Last case patient with naturally transmitted smallpox (October 1977). (Source: World Health Organization.) CERTIFICATION After a series of independent evaluations, including field visits to formerly endemic counties by separate international commissions, the Global Commission for the Certification of Smallpox Eradication concluded that eradication had been achieved in December 1979 ( Tables 3 and 4 ). The number of laboratories with variola virus was reduced from 76 to 6 by May 1980, when the World Health Assembly accepted the recommendations of the Global Commission that eradication had indeed been achieved [ 29 ]; there are now only 2 laboratories known to retain variola virus stocks—Vektor in Novosibirsk, Russia, and at the CDC in Atlanta. Both laboratories are visited every 2 years by biosafety experts convened by WHO, to assure that maximum biocontainment of the variola isolates is assured. Table 3. Smallpox Certification Activities in 200 Countries, 1977–1980 a Countries by Category Countries, No. Population (Billions) Total 200 4.5 Submitting statements 121 1.5 Visited by commissions 79 3.0 At special risk 44 1.8 Endemic during smallpox eradication programs 35 1.2 a A total of 17†000 specimens were collected. Table 4. Smallpox Certification Activities in India: Facial Pockmark Surveys, 1977 Age Group Survey Persons Examined, No. Total With Facial Pockmarks a Preschool children National surveys 271†897 0 International Commission 3139 0 School-age children National surveys 224†297 38 International Commission 26†665 18 Adults National surveys 1†451†125 650 International Commission 14†307 94 a Pockmarks due to smallpox contracted before 1975. CONTINUING ISSUES There remains concern that a bioterrorist event using smallpox virus could occur; this would wreak havoc globally [30]. The CDC has a smallpox response plan that focuses on national leadership, community-based planning, public health response actions, and health care facility response activities. The US national vaccine stockpile maintains 3 vaccines [ 31 , 32, 33]. ACAM2000 is a Food and Drug Administration–licensed vaccine grown on tissue culture, derived from the New York Board of Health strain of vaccine used to make Dryvax; this latter product was used widely in the eradication program. Aventis Pasteur smallpox vaccine is a vaccine supply created in the 20th century but still retaining potency. The Imvamune vaccine (Bavarian Nordic) uses modified vaccinia Ankara, a nonreplicating vaccine that requires 2 injections and is thought to elicit fewer adverse events, especially in immunologically deficient persons; it is not yet licensed. The Food and Drug Administration has recently approved Tecovirimat for the treatment of smallpox; this drug is also destined for the US national stockpile [34]. The WHO, likewise, has a smallpox preparedness plan. Their operational framework addresses education, laboratory diagnosis, biosafety and security, provision of expertise and supplies, and the strengthening of national level responses. WHO also maintains an emergency vaccine stockpile. The de novo synthesis of horsepox virus has raised concern that a terrorist could recreate variola virus [ 35 ]. Because routine smallpox vaccination stopped in the United States in 1971 and globally in the early 1980s, there is justifiable concern over population vulnerability. Constant vigilance for dealing with a return of smallpox is warranted. The 2 high security repositories of variola virus stocks, at CDC in Atlanta, and at Vektor in Novosibirsk, Russia, are inspected periodically by WHO. Periodic debates at the World Health Assembly over whether to destroy the remaining known stocks of smallpox virus have not concluded it was time to do so. Another important continuing issue is the increasing number of outbreaks and cases of human monkeypox in central and western Africa [36, 37 ]. These outbreaks will increase as population immunity falls. The smallpox vaccine protects against monkeypox. At some point, consideration must be given whether vaccinia should be used for vulnerable populations in monkeypox-endemic areas. Supplementary Data Supplementary materials are available at The Journal of Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author. The complete references are available as online Supplemental Material. Supplementary Material jiaa588_suppl_Supplementary-Material Click here for additional data file.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9803934/

Vector preference and heterogeneity in host sex ratio can affect pathogen spread in natural plant populations

Vector-borne diseases threaten human and agricultural health and are a critical component of the ecology of plants and animals. While previous studies have shown that pathogen spread can be affected by vector preferences for host infection status, less attention has been paid to vector preference for host sex, despite abundant evidence of sex-specific variation in disease burden. We investigated vector preference for host infection status and sex in the sterilizing 'anther-smut' pathogen ( Microbotryum ) of the alpine carnation, Dianthus pavonius. The pathogen is transferred among hosts by pollinators that visit infected flowers and become contaminated with spores produced by infected anthers. The host plant has a mixed breeding system with hermaphrodites and females. In experimental floral arrays, pollinators strongly preferred healthy hermaphrodites over both females and diseased plants, consistently across different guilds of pollinators and over multiple years. Using an agent-based model we showed that pollinator preferences for sex can affect pathogen spread in populations with variable sex-ratios, even if there is no preference for infection status. Our results demonstrate that vector preferences for host traits other than infection status can play a critical role in pathogen transmission dynamics when there is heterogeneity for those traits in the host population.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7755716/

Reverse Engineering Targets for Recombinant Protein Production in Corynebacterium glutamicum Inspired by a Fast-Growing Evolved Descendant

We previously reported a Corynebacterium glutamicum JH41 strain with a 58% faster growth rate through application of adaptive laboratory evolution. To verify that the fast-reproducing strain was useful as a host for recombinant protein expression, we introduced a plasmid responsible for the secretory production of a recombinant protein. The JH41 strain harboring the plasmid indeed produced the secretory recombinant protein at a 2.7-fold greater rate than its ancestral strain. To provide the reverse engineering targets responsible for boosting recombinant protein production and cell reproduction, we compared the genome sequence of the JH41 strain with its ancestral strain. Among the 15 genomic variations, a point mutation was confirmed in the 14 bases upstream of NCgl1959 (encoding a presumed siderophore-binding protein). This mutation allowed derepression of NCgl1959, thereby increasing iron consumption and ATP generation. A point mutation in the structural gene ramA (A239G), a LuxR-type global transcription regulator involved in central metabolism, allowed an increase in glucose consumption. Therefore, mutations to increase the iron and carbon consumption were concluded as being responsible for the enhanced production of recombinant protein and cell reproduction in the evolved host. Introduction Many attempts through microbial metabolic engineering have been made to overproduce metabolites by (1) improving the substrate uptake rate, (2) reducing the flux to undesired by-products, (3) introducing heterologous pathways and optimizing the activity of the enzyme, and (4) balancing product formation in cells, such as secretion of product to the extracellular medium ( Yadav et al., 2012 ). However, the metabolic manipulation of microorganisms can cause flux imbalance due to the lack of understanding of the complex regulatory mechanisms essential for operation ( Yadav et al., 2012 ). Adaptive laboratory evolution (ALE) is a reasonable approach because it not only copes with the physical defects of strains caused by genetic manipulation but also improves the productivity of cell factories ( Sandberg et al., 2019 ; Lee and Kim, 2020 ). In addition, the commercialization of next-generation sequencing (NGS) technology and the development of omics tools have enabled reverse engineering by helping us better understand mechanisms of microbial regulation ( Stella et al., 2019 ). Besides, genetically modified organism (GMO)-related issues in the food industry have increased the demand for non-GMO strains derived from ALE. Growth rate is a significant parameter in evaluating the compatibility of industrial strains. It is generally recognized that fast growth rates can achieve high cell densities and high productivity ( Feist et al., 2010 ). Genomic analyses of fast-growing mutant strains have revealed previously unknown metabolic engineering targets and methods to improve production hosts. Rugbjerg et al. (2018) reported that the introduction of the rpoB mutation into Escherichia coli increased the growth rate by 13% while also increasing the production rate of mevalonate, a building block of terpenoid products such as fragrances and bioplastics, by 71%. In addition, Wang et al. reported that the introduction of gntR1 and ramA mutations into Corynebacterium glutamicum increased the growth rate by 37% and the productivity of lysine by 100% ( Wang et al., 2018 ). Recently, we reported a fast-reproducing C. glutamicum JH41 strain, obtained through the application of ALE ( Park et al., 2020 ). The transcriptome pattern of the evolved JH41 strain revealed that the upregulation of genes in the tricarboxylic acid cycle (TCA) cycle and respiratory chain, and oxidative stress responses allowed the growth promotion. Since C. glutamicum is a useful host for recombinant protein expression ( Lee and Kim, 2018 ), it was proposed that the fast-reproducing JH41 strain might be an improved host for recombinant protein production and may also provide hidden genetic targets for reverse engineering of C. glutamicum hosts. In this report, the fast-reproducing JH41 was indeed verified as an improved host for recombinant protein production. Genomic analysis of the evolved strain was therefore performed to explain the reasons for the increased growth rate and recombinant protein production. Materials and Methods Strains and Plasmids Corynebacterium glutamicum PT strain (ancestor of JH41) and a fast-reproducing JH41 strain (descendant of PT) were used as hosts for recombinant protein production and genomic analysis ( Park et al., 2020 ). DNA manipulation followed the method described by Sambrook et al. using E. coli DH10B (Invitrogen Inc., Carlsbad, CA, United States) ( Sambrook and Russell, 2001 ). The strains and plasmids are listed in Table 1 and the primers in Supplementary Table 1 . Target gene-disrupted or gene-substituted C. glutamicum strains were prepared by double crossover using the pK mobsacB -based plasmids. Homologous arms of the target genes were constructed by connecting two fragments amplified from genomic DNA using overlap extension PCR ( Ho et al., 1989 ). Oligonucleotide synthesis and plasmid sequence confirmation were performed at a facility of Bionics Inc., (Seoul, South Korea). pfu polymerase was purchased from Solgent Inc., (Daejeon, South Korea) and restriction enzymes and T4-ligase from New England Biolabs Inc., (Ipswich, MA, United States). TABLE 1 Strains and plasmids used in this study. Strains and plasmids Description References Strains E. coli DH10B F – mcrA Δ( mrr-hsd RMS- mcrBC ) Φ80 lacZ ΔM15 Δ lacX74 recA1 endA1 araD139 Δ( ara leu )7697 galU galK rpsL nupG λ – Invitrogen BL21(DE3) E. coli F – ompT gal dcm lon hsdS B r B – m B – ; E. coli B strain), with λDE3 Novagen C. glutamicum PT Home stock of C. glutamicum wild-type, Biotin auxotroph, derived from KCTC No. 1445 Kim et al., 2016 JH41 A descendant of C. glutamicum PT, selected after 600 generation, 58% faster reproduction than PT Park et al., 2020 PT NCgl0774:35-bp PT harboring genomic NCgl0774, putative siderophore- binding lipoprotein A, inserting the 35-bp sequence This study PT NCgl1159 C–108T PT harboring genomic NCgl1159, F 0 F 1 -ATP synthase subunit A, substituting the 108th upstream sequence of NCgl1159 from C to T This study PT Δ ripA PT derivative, Δ ripA (0.8-kb deletion) This study PT Δ dtxR PT derivative, Δ dtxR (0.9-kb deletion) This study PT ramA A239G PT harboring genomic ramA , transcriptional regulator MalT, substituting to A239G This study PT SBP B C–14G PT harboring genomic SBP B (NCgl1959), putative siderophore-binding lipoprotein, substituting the 14 h upstream sequence of SBP from C to G This study PT S + R PT harboring genomic mutation of PT ramA A239G and PT SBP B C–14G This study Plasmids pET24a E. coli expression vector, P T 7 , Km R Novagen pET24a-DtxR pET24a harboring DtxR (NCgl1845) of C. glutamicum This study pSL360 E. coli and C. glutamicum shuttle vector, Km R Park et al., 2004 pCG-H36A C. glutamicum vector, TAT-signal sequence of cg1514 (start codon is changed to ATG), a strong synthetic H36 promoter Yim et al., 2016 pCG-H36A-agarase pCG-H36A derivative, harboring S. coelicolor agarase Yim et al., 2016 pCG-H36A(-SS) pCG-H36A derivative, signal peptide deleted vector, H36 promoter This study pCG-H36A(-SS)-GFP pCG-H36A(-SS) derivative, harboring eGFP This study pCG-H36A- porD pCG-H36A derivative, TAT-signal sequence of porD , H36 promoter This study pCG-H36A- porD- GFP pCG-H36A- porD derivative, harboring eGFP This study pK19 mobsacB Suicide vector for double recombination, Km R , sacB of Bacillus subtilis Schafer et al., 1994 pSL360-SBP B pSL360 harboring putative siderophore-binding lipoprotein, SBP B (NCgl1959) of C. glutamicum This study pK19 mobsacB -NCgl0774:35-bp pK19 mobsacB harboring homologous arm for insertion of NCgl0774:35-bp This study pK19 mobsacB -Δ ripA pK19 mobsacB harboring homologous arm for deletion of Δ ripA This study pK19 mobsacB -Δ dtxR pK19 mobsacB harboring homologous arm for deletion of Δ dtxR This study Media and Batch Cultures As for the preculture for DNA manipulations, lysogeny broth (LB) complex medium was used for E. coli , and brain heart infusion (BHI) medium was used for C. glutamicum . C. glutamicum was incubated at 30°C and 200 rpm in a shaking incubator, and E. coli was incubated at 37°C and 220 rpm. For batch cultures of C. glutamicum strains, a modified MCGC minimal medium [0.9% glucose, 6 g Na 2 HPO 4 , 4 g (NH4) 2 SO 4 , 3 g KH 2 PO 4 , 1 g sodium citrate dehydrate, 1 g NaCl, 0.1 g MgSO 4 –7H 2 O, 20 mg FeSO 4 –7H 2 O, 2 mg MnSO 4 –H 2 O, 2 mg FeCl 3 , 1 mg thiamine–HCl, 200 μg biotin, 35 μg CaCl 2 , 0.5 μg ZnSO 4 –7H 2 O, 0.2 μg Na 2 B 4 O 7 –10H 2 O, 0.2 μg CuCl 2 –2H 2 O, 0.1 μg (NH 4 ) 6 Mo 7 O 24 –4H 2 O per liter] was used ( Osten et al., 1989 ). A 500-ml baffled flask containing 50 ml of the modified MCGC medium was used for efficient oxygen transfer. The initial cell mass was adjusted to OD 600 = 0.1. Biomass was estimated by measuring optical density (OD) at 600 nm every hour and converting into g – DCW /L unit by the coefficient of 0.25. Secretory Production of Recombinant Protein To confirm the recombinant protein production of the hosts, pCG-H36A-agarase encoding a strong autonomous promoter and TAT-secretion signal sequence with the agarase gene from Streptomyces coelicolor was further transformed into the PT and JH41 strains ( Yim et al., 2016 ). After incubating for 24 and 48 h in a 500-ml baffled flask containing 50 ml of MCGC media, 50 ml of the culture broth was centrifuged (12,000 × g, 4°C, 5 min). The supernatant containing the secreted protein was concentrated 50-fold using a centrifugal filter (Amicon ® Ultra-15 10K, Merck Millipore, Burlington, MA, United States), and then, the buffer was changed with 50 mM Tris–HCl (pH 7.0). The prepared protein samples were diluted to final OD 600 values and then loaded on a 12% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) gel. The photograph of the gel was captured using a chemical fluorescence image analyzer (Chemi Doc MP System, Bio-Rad Laboratories, Inc., Hercules, CA, United States), and the intensity of the target band was compared using the ImageJ program. Genome Sequencing For genome sequencing, genomic DNAs of C. glutamicum PT and JH41 were isolated using a gDNA prep kit (Solgent Inc., Daejeon, South Korea) according to the manufacturer's instructions. The genome sequence was determined using the PacBio RS II (Pacific Biosciences Inc., Menlo Park, CA, United States) at a DNA sequencing facility (ChunLab Inc.). Raw sequences were assembled with PacBio SMRT Analysis ver. 2.0 software (Pacific Biosciences Inc., Menlo Park, CA, United States). Gene prediction was performed using Glimmer 3 ( Delcher et al., 2007 ), and annotations were performed by a homology search against the SEED database, Universal Protein Resource (UniProt) database, and eggNOG database ( Disz et al., 2010 ). Genome sequences of the PT and JH41 strains were compared using EzGenome ( http://ezgenome.ezbiocloud.net/ezg_browse ) and CLgenomics program (ChunLab). Analysis of single-nucleotide polymorphisms (SNPs) and gap was performed using the MAUVE ( Darling et al., 2004 ) and MegaX ( Kumar et al., 2018 ) programs. Analyses of ATP, Iron, and Glucose Concentrations Cells were harvested from batch cultures (1 ml) at 5 h by centrifugation (8,000 × g , 4°C, 5 min) and washed twice with sterile distilled water. The cell suspension (0.8 ml) was mixed with 0.2 g of glass beads (212–300 μm) in a screw-capped tube followed by homogenization for 30 s using a bead beater (Model 607, BioSpec Inc., Bartlesville, OK, United States). After chilling in an ice bath for 1 min, homogenization was repeated five more times. Cell debris was removed by centrifugation (8,000 × g , 4°C, 5 min), and the supernatant was used for measuring intracellular ATP. ATP concentration was estimated using an ENLITEN ATP Assay System Bioluminescence Detection Kit (#FF2000, Promega Inc., Madison, WI, United States). Light intensity was detected by a luminometer (GloMax ® -96 Microplate Luminometer, Promega Inc., Madison, WI, United States) after a luciferase-driven reaction ( Na et al., 2015 ). Iron consumption was estimated by measuring the remaining free iron concentration using the o -phenanthroline colorimetric method ( Serra et al., 2015 ) after cell removal. Culture broth (1 ml) was centrifuged (12,000 × g , 4°C, 5 min), and the supernatant (100 μl) was mixed with 20 μl of bromophenol blue (0.4 g/L). After pH adjustment to 3.5 by adding 0.2% H 2 SO 4 , a further 10 μl of hydroquinone (1 g/L), 20 μl of o -phenanthroline (2.5 g/L), and distilled water (up to 250 μl) were added to the mixture. After 1 h of color development at room temperature, absorbance was measured at 510 nm. The analyses of ATP and iron concentrations were determined at least three times with three biological replicates. To analyze the glucose consumption of JH41 and PT ramA A239G , the glucose concentration in the medium was measured by a reducing sugar quantification method using a 3,5-dinitrosalicylic acid (DNS) reagent ( Deshavath et al., 2020 ). After cell removal, 150 μl of supernatant and 150 μl of DNS reagent were mixed and heated at 100°C for 5 min. After standing to cool for 10 min at room temperature, 200 μl of the mixture was dispensed into 96-well plates to measure absorbance at 540 nm. D-Glucose reagent was used as a standard. Determination of DtxR Affinity on the Mutated Promoter Sequence The affinity of negative regulator diphtheria toxin repressor (DtxR) on the mutated promoter region was determined by an electrophoresis mobility shift assay using the purified DtxR protein and the PCR-amplified DNA fragment. The dtxR gene from C. glutamicum was cloned and expressed in E. coli BL21 (DE3) with an additional 12 × His tag in the expression plasmid (pET24b- dtxR , Table 1 ). The protein was further purified using Ni 2+ chelate affinity chromatography, as described previously ( Wennerhold et al., 2005 ). To test the binding of DtxR to the putative target promoter DNA, the purified DtxR protein (0–350 ng, dimeric form) was dissolved with PCR-amplified DNA fragments (399 bp, 200 ng) in a 10-μl binding buffer [50 mM Tris–HCl (pH 7.5), 5 mM MgCl 2 , 40 mM KCl, 5% (v/v) glycerol, 1 mM dithiothreitol (DTT), and 150 μM MnCl 2 ]. The reaction mixture was loaded onto a 15% native polyacrylamide gel containing 1 mM DTT and 150 μM MnCl 2 after incubation for 30 min at room temperature. Electrophoresis was performed at room temperature and 170 V using 1 × TB (89 mM Tris base, 89 mM boric acid) supplemented with 1 mM DTT and 150 μM MnCl 2 as an electrophoresis buffer. The gel was then stained with a fluorescent dye (SYBR Green I EMSA Kit, Invitrogen Inc., Carlsbad, CA, United States) according to the manufacturer's instructions. The photograph of the gel was captured using a chemical fluorescence image analyzer (Chemi Doc MP System, Bio-Rad Laboratories, Inc., Hercules, CA, United States). Transcription Quantification of rRNAs The ribosomal RNAs (rRNAs) transcription level determination method was followed as described ( Kwon et al., 2015 ). Total RNA was extracted from actively growing C. glutamicum strains using TRIzol1 reagent (Invitrogen, Carlsbad, CA, United States) and NucleoSpin1 RNA II Kit (Macherey-Nagel, Düren, Germany) after cell disruption using a Mini-Beadbeater-16 (BioSpec, Bartlesville, PA, United States) with glass beads (acid washed, 212–300 mm, Sigma-Aldrich, MO, United States). The 50 ng of total RNAs was used for complementary DNA (cDNA) synthesis using ReverTra Ace-α-1 (Toyobo, Osaka, Japan) according to the manufacturer's instructions. THUNDERBIRD SYBR1 qPCR Mix (Toyobo, Osaka, Japan) and the Step-One Plus Real-Time PCR System (Applied Biosystems, Foster City, CA, United States) were used for the gene expression analysis. Step-One Plus Software ver. 2.0 (Applied Biosystems, Foster City, CA, United States) was used to quantify the relative transcription. gapA [glyceraldehyde 3-phosphate dehydrogenase (GAPDH)] and leuA (2-isopropyl malate synthase) genes were used as internal references. The relative expression of rRNAs with the reference genes was determined using the 2 –Δ Δ CT (Livak) method. Primers for identifying transcription levels are shown in Supplementary Table 1 . Strains and Plasmids Corynebacterium glutamicum PT strain (ancestor of JH41) and a fast-reproducing JH41 strain (descendant of PT) were used as hosts for recombinant protein production and genomic analysis ( Park et al., 2020 ). DNA manipulation followed the method described by Sambrook et al. using E. coli DH10B (Invitrogen Inc., Carlsbad, CA, United States) ( Sambrook and Russell, 2001 ). The strains and plasmids are listed in Table 1 and the primers in Supplementary Table 1 . Target gene-disrupted or gene-substituted C. glutamicum strains were prepared by double crossover using the pK mobsacB -based plasmids. Homologous arms of the target genes were constructed by connecting two fragments amplified from genomic DNA using overlap extension PCR ( Ho et al., 1989 ). Oligonucleotide synthesis and plasmid sequence confirmation were performed at a facility of Bionics Inc., (Seoul, South Korea). pfu polymerase was purchased from Solgent Inc., (Daejeon, South Korea) and restriction enzymes and T4-ligase from New England Biolabs Inc., (Ipswich, MA, United States). TABLE 1 Strains and plasmids used in this study. Strains and plasmids Description References Strains E. coli DH10B F – mcrA Δ( mrr-hsd RMS- mcrBC ) Φ80 lacZ ΔM15 Δ lacX74 recA1 endA1 araD139 Δ( ara leu )7697 galU galK rpsL nupG λ – Invitrogen BL21(DE3) E. coli F – ompT gal dcm lon hsdS B r B – m B – ; E. coli B strain), with λDE3 Novagen C. glutamicum PT Home stock of C. glutamicum wild-type, Biotin auxotroph, derived from KCTC No. 1445 Kim et al., 2016 JH41 A descendant of C. glutamicum PT, selected after 600 generation, 58% faster reproduction than PT Park et al., 2020 PT NCgl0774:35-bp PT harboring genomic NCgl0774, putative siderophore- binding lipoprotein A, inserting the 35-bp sequence This study PT NCgl1159 C–108T PT harboring genomic NCgl1159, F 0 F 1 -ATP synthase subunit A, substituting the 108th upstream sequence of NCgl1159 from C to T This study PT Δ ripA PT derivative, Δ ripA (0.8-kb deletion) This study PT Δ dtxR PT derivative, Δ dtxR (0.9-kb deletion) This study PT ramA A239G PT harboring genomic ramA , transcriptional regulator MalT, substituting to A239G This study PT SBP B C–14G PT harboring genomic SBP B (NCgl1959), putative siderophore-binding lipoprotein, substituting the 14 h upstream sequence of SBP from C to G This study PT S + R PT harboring genomic mutation of PT ramA A239G and PT SBP B C–14G This study Plasmids pET24a E. coli expression vector, P T 7 , Km R Novagen pET24a-DtxR pET24a harboring DtxR (NCgl1845) of C. glutamicum This study pSL360 E. coli and C. glutamicum shuttle vector, Km R Park et al., 2004 pCG-H36A C. glutamicum vector, TAT-signal sequence of cg1514 (start codon is changed to ATG), a strong synthetic H36 promoter Yim et al., 2016 pCG-H36A-agarase pCG-H36A derivative, harboring S. coelicolor agarase Yim et al., 2016 pCG-H36A(-SS) pCG-H36A derivative, signal peptide deleted vector, H36 promoter This study pCG-H36A(-SS)-GFP pCG-H36A(-SS) derivative, harboring eGFP This study pCG-H36A- porD pCG-H36A derivative, TAT-signal sequence of porD , H36 promoter This study pCG-H36A- porD- GFP pCG-H36A- porD derivative, harboring eGFP This study pK19 mobsacB Suicide vector for double recombination, Km R , sacB of Bacillus subtilis Schafer et al., 1994 pSL360-SBP B pSL360 harboring putative siderophore-binding lipoprotein, SBP B (NCgl1959) of C. glutamicum This study pK19 mobsacB -NCgl0774:35-bp pK19 mobsacB harboring homologous arm for insertion of NCgl0774:35-bp This study pK19 mobsacB -Δ ripA pK19 mobsacB harboring homologous arm for deletion of Δ ripA This study pK19 mobsacB -Δ dtxR pK19 mobsacB harboring homologous arm for deletion of Δ dtxR This study Media and Batch Cultures As for the preculture for DNA manipulations, lysogeny broth (LB) complex medium was used for E. coli , and brain heart infusion (BHI) medium was used for C. glutamicum . C. glutamicum was incubated at 30°C and 200 rpm in a shaking incubator, and E. coli was incubated at 37°C and 220 rpm. For batch cultures of C. glutamicum strains, a modified MCGC minimal medium [0.9% glucose, 6 g Na 2 HPO 4 , 4 g (NH4) 2 SO 4 , 3 g KH 2 PO 4 , 1 g sodium citrate dehydrate, 1 g NaCl, 0.1 g MgSO 4 –7H 2 O, 20 mg FeSO 4 –7H 2 O, 2 mg MnSO 4 –H 2 O, 2 mg FeCl 3 , 1 mg thiamine–HCl, 200 μg biotin, 35 μg CaCl 2 , 0.5 μg ZnSO 4 –7H 2 O, 0.2 μg Na 2 B 4 O 7 –10H 2 O, 0.2 μg CuCl 2 –2H 2 O, 0.1 μg (NH 4 ) 6 Mo 7 O 24 –4H 2 O per liter] was used ( Osten et al., 1989 ). A 500-ml baffled flask containing 50 ml of the modified MCGC medium was used for efficient oxygen transfer. The initial cell mass was adjusted to OD 600 = 0.1. Biomass was estimated by measuring optical density (OD) at 600 nm every hour and converting into g – DCW /L unit by the coefficient of 0.25. Secretory Production of Recombinant Protein To confirm the recombinant protein production of the hosts, pCG-H36A-agarase encoding a strong autonomous promoter and TAT-secretion signal sequence with the agarase gene from Streptomyces coelicolor was further transformed into the PT and JH41 strains ( Yim et al., 2016 ). After incubating for 24 and 48 h in a 500-ml baffled flask containing 50 ml of MCGC media, 50 ml of the culture broth was centrifuged (12,000 × g, 4°C, 5 min). The supernatant containing the secreted protein was concentrated 50-fold using a centrifugal filter (Amicon ® Ultra-15 10K, Merck Millipore, Burlington, MA, United States), and then, the buffer was changed with 50 mM Tris–HCl (pH 7.0). The prepared protein samples were diluted to final OD 600 values and then loaded on a 12% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) gel. The photograph of the gel was captured using a chemical fluorescence image analyzer (Chemi Doc MP System, Bio-Rad Laboratories, Inc., Hercules, CA, United States), and the intensity of the target band was compared using the ImageJ program. Genome Sequencing For genome sequencing, genomic DNAs of C. glutamicum PT and JH41 were isolated using a gDNA prep kit (Solgent Inc., Daejeon, South Korea) according to the manufacturer's instructions. The genome sequence was determined using the PacBio RS II (Pacific Biosciences Inc., Menlo Park, CA, United States) at a DNA sequencing facility (ChunLab Inc.). Raw sequences were assembled with PacBio SMRT Analysis ver. 2.0 software (Pacific Biosciences Inc., Menlo Park, CA, United States). Gene prediction was performed using Glimmer 3 ( Delcher et al., 2007 ), and annotations were performed by a homology search against the SEED database, Universal Protein Resource (UniProt) database, and eggNOG database ( Disz et al., 2010 ). Genome sequences of the PT and JH41 strains were compared using EzGenome ( http://ezgenome.ezbiocloud.net/ezg_browse ) and CLgenomics program (ChunLab). Analysis of single-nucleotide polymorphisms (SNPs) and gap was performed using the MAUVE ( Darling et al., 2004 ) and MegaX ( Kumar et al., 2018 ) programs. Analyses of ATP, Iron, and Glucose Concentrations Cells were harvested from batch cultures (1 ml) at 5 h by centrifugation (8,000 × g , 4°C, 5 min) and washed twice with sterile distilled water. The cell suspension (0.8 ml) was mixed with 0.2 g of glass beads (212–300 μm) in a screw-capped tube followed by homogenization for 30 s using a bead beater (Model 607, BioSpec Inc., Bartlesville, OK, United States). After chilling in an ice bath for 1 min, homogenization was repeated five more times. Cell debris was removed by centrifugation (8,000 × g , 4°C, 5 min), and the supernatant was used for measuring intracellular ATP. ATP concentration was estimated using an ENLITEN ATP Assay System Bioluminescence Detection Kit (#FF2000, Promega Inc., Madison, WI, United States). Light intensity was detected by a luminometer (GloMax ® -96 Microplate Luminometer, Promega Inc., Madison, WI, United States) after a luciferase-driven reaction ( Na et al., 2015 ). Iron consumption was estimated by measuring the remaining free iron concentration using the o -phenanthroline colorimetric method ( Serra et al., 2015 ) after cell removal. Culture broth (1 ml) was centrifuged (12,000 × g , 4°C, 5 min), and the supernatant (100 μl) was mixed with 20 μl of bromophenol blue (0.4 g/L). After pH adjustment to 3.5 by adding 0.2% H 2 SO 4 , a further 10 μl of hydroquinone (1 g/L), 20 μl of o -phenanthroline (2.5 g/L), and distilled water (up to 250 μl) were added to the mixture. After 1 h of color development at room temperature, absorbance was measured at 510 nm. The analyses of ATP and iron concentrations were determined at least three times with three biological replicates. To analyze the glucose consumption of JH41 and PT ramA A239G , the glucose concentration in the medium was measured by a reducing sugar quantification method using a 3,5-dinitrosalicylic acid (DNS) reagent ( Deshavath et al., 2020 ). After cell removal, 150 μl of supernatant and 150 μl of DNS reagent were mixed and heated at 100°C for 5 min. After standing to cool for 10 min at room temperature, 200 μl of the mixture was dispensed into 96-well plates to measure absorbance at 540 nm. D-Glucose reagent was used as a standard. Determination of DtxR Affinity on the Mutated Promoter Sequence The affinity of negative regulator diphtheria toxin repressor (DtxR) on the mutated promoter region was determined by an electrophoresis mobility shift assay using the purified DtxR protein and the PCR-amplified DNA fragment. The dtxR gene from C. glutamicum was cloned and expressed in E. coli BL21 (DE3) with an additional 12 × His tag in the expression plasmid (pET24b- dtxR , Table 1 ). The protein was further purified using Ni 2+ chelate affinity chromatography, as described previously ( Wennerhold et al., 2005 ). To test the binding of DtxR to the putative target promoter DNA, the purified DtxR protein (0–350 ng, dimeric form) was dissolved with PCR-amplified DNA fragments (399 bp, 200 ng) in a 10-μl binding buffer [50 mM Tris–HCl (pH 7.5), 5 mM MgCl 2 , 40 mM KCl, 5% (v/v) glycerol, 1 mM dithiothreitol (DTT), and 150 μM MnCl 2 ]. The reaction mixture was loaded onto a 15% native polyacrylamide gel containing 1 mM DTT and 150 μM MnCl 2 after incubation for 30 min at room temperature. Electrophoresis was performed at room temperature and 170 V using 1 × TB (89 mM Tris base, 89 mM boric acid) supplemented with 1 mM DTT and 150 μM MnCl 2 as an electrophoresis buffer. The gel was then stained with a fluorescent dye (SYBR Green I EMSA Kit, Invitrogen Inc., Carlsbad, CA, United States) according to the manufacturer's instructions. The photograph of the gel was captured using a chemical fluorescence image analyzer (Chemi Doc MP System, Bio-Rad Laboratories, Inc., Hercules, CA, United States). Transcription Quantification of rRNAs The ribosomal RNAs (rRNAs) transcription level determination method was followed as described ( Kwon et al., 2015 ). Total RNA was extracted from actively growing C. glutamicum strains using TRIzol1 reagent (Invitrogen, Carlsbad, CA, United States) and NucleoSpin1 RNA II Kit (Macherey-Nagel, Düren, Germany) after cell disruption using a Mini-Beadbeater-16 (BioSpec, Bartlesville, PA, United States) with glass beads (acid washed, 212–300 mm, Sigma-Aldrich, MO, United States). The 50 ng of total RNAs was used for complementary DNA (cDNA) synthesis using ReverTra Ace-α-1 (Toyobo, Osaka, Japan) according to the manufacturer's instructions. THUNDERBIRD SYBR1 qPCR Mix (Toyobo, Osaka, Japan) and the Step-One Plus Real-Time PCR System (Applied Biosystems, Foster City, CA, United States) were used for the gene expression analysis. Step-One Plus Software ver. 2.0 (Applied Biosystems, Foster City, CA, United States) was used to quantify the relative transcription. gapA [glyceraldehyde 3-phosphate dehydrogenase (GAPDH)] and leuA (2-isopropyl malate synthase) genes were used as internal references. The relative expression of rRNAs with the reference genes was determined using the 2 –Δ Δ CT (Livak) method. Primers for identifying transcription levels are shown in Supplementary Table 1 . Results Effect of Fast-Growing JH41 Host on the Secretory Production of Recombinant Protein Because protein is the major macromolecule in cells, the rapidly reproducing JH41 strain was supposed to be able to produce a protein at an enhanced rate. The fact that no extracellular protease activity or economical purification of secretory protein was found in C. glutamicum host ( Lee and Kim, 2018 ) further encouraged us to hypothesize that the rapidly reproducing JH41 strain would become an improved C. glutamicum host for secretory protein production. To verify this hypothesis, a plasmid designed to secrete foreign agarase (pCG-H36A-agarase) was introduced into JH41 and PT strains. The secretory recombinant protein (agarase, 34 kDa) showed a major band in the SDS-PAGE ( Figure 1A ). The band intensities of the recombinant protein from the JH41 host were 7,490 AU at 24 h and 17,231 AU at 48 h, while those from the PT host were 4,780 AU at 24 h and 6,212 AU at 48 h ( Figure 1B ). The JH41 host produced the secretory recombinant protein (56% at 24 h and 177% at 48 h) at a greater rate than the PT host. In order to confirm the production effect of other proteins in the JH41 host, a plasmid expressing green fluorescent protein (GFP) in cytoplasm (pCG-H36A[-SS]-GFP, no signal sequence) and a plasmid secreting GFP by a TAT-signal sequence (pCG-H36A- porD -GFP) were introduced into the JH41 host ( Supplementary Figure 1 ). The productions of cytoplasmic GFP and secreted GFP from the JH41 host were greater than those from the PT host by 5–16% and 13–116%, respectively. Therefore, the rapid-reproducing JH41 strain was an enhanced host for the production of recombinant proteins. FIGURE 1 Effects of host on the secretory production of recombinant protein. (A) Image of sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE). Lane 1, PT + pCG-H36A (empty vector) at 24 h; lane 2, PT + pCG-H36A-agarase at 24 h; lane 3, JH41 + pCG-H36A (empty vector) at 24 h; lane 4, JH41 + pCG-H36A-agarase at 24 h; lane 5, PT + pCG-H36A (empty vector) at 48 h; lane 6, PT + pCG-H36A-agarase at 48 h; lane 7, JH41 + pCG-H36A (empty vector) at 48 h; lane 8, JH41 + pCG-H36A-agarase at 48 h. Arrow indicates the secreted agarase band (34 kDa). The image is from the representative experiment from three independent experiments. (B) Band density estimation of the SDS-PAGE image using ImageJ software. a: lane 2; b: lane 4; c: lane 6; d: lane 8. Genome Analysis of the JH41 Host To understand the reasons for enhanced cell reproduction and to provide reverse engineering targets responsible for the improved recombinant protein production, the JH41 genome was resequenced and deposited [National Center for Biotechnology Information (NCBI) accession number PRJNA554987 ]. A total 15 genomic mutations were found to be accumulated in JH41 compared to its ancestral PT ( Table 2 ). Among them, four were found in the non-coding region, including an internal transcribed spacer (ITS) mutation, and 11 in the coding region. Genomic mutations were categorized into the clusters of orthologous groups of proteins (COG) as four to the function unknown group (S), two to the transcription group (K), two to the inorganic ion transport and metabolism group (P), one to the defense mechanism (V), one to the signal transduction mechanism (T), and one to the energy production and conversion group (C). The remaining three mutations did not show protein function or were not protein-encoding regions. A 35-bp repeated mutation in the structural gene encoding a putative siderophore-binding lipoprotein A (SBP A, NCgl0774) and a substitution mutation (C→G) at the 14 bases upstream of the start codon of the gene encoding a putative siderophore-binding lipoprotein B (SBP B, NCgl1959) were identified. A condensed mutation was also found in one of the operons encoding rRNA primary transcript. One mutation was found in the 16S rRNA structural gene, and another 23-point mutations were densely observed between the 16S and 23S rRNA sequences, which are called 16S–23S internal transcribed spacer (ITS) ( Osorio et al., 2005 ). TABLE 2 Genomic mutations of the JH41 host over ancestral host. PT position a JH41 position a COG functional category b Locus tag Mutation region Variants Function (gene name) mRNA fold based on transcriptome c) 80,225 1,076,088 S NCgl2726 Non-coding (-57 bp of NCgl2726) Substitution (G→A) Tripartite tricarboxylate transporter substrate binding protein 3.05 461,943 1,457,809 - - 16S rRNA coding region Substitution (A→G) Bacterial small subunit ribosomal RNA; 16S ribosomal RNA – 462,350 1,458,217 Non-coding (16S-23S ITS of 1 st rRNA operon) Similar to 16S–23S ITS of 5th rRNA operon rRNA operon 16S-23S ITS (internal transcribed spacer) – 1,191,984 2,187,859 T NCgl0722 Coding (372 bp) Insertion (C) Histidine protein kinase ( mtrB ) 1.27 1,251,401 2,247,312 P NCgl0774 Coding (15 bp) Repeat (35 bp, AAAACTCTCCTCCGCA CTGATCGTGCTTCTCG CAG) ABC transporter system; A putative Fe 3+ -siderophore-binding lipoprotein 0.03 1,660,001 2,655,914 C NCgl1159 Non-coding (-108 bp of NCgl1159) Substitution (C→T) ATP synthase F 0 F 1 subunit A 0.27 1,968,380 2,964,295 K NCgl1435 Coding (201 bp) Substitution (A→G) Transcriptional regulator 1.17 2,004,381 3,000,298 – – Coding (67 bp) Substitution (G→A) Hypothetical protein – 2,326,488 217,999 P NCgl1959 Non-coding (-14 bp of NCgl1959) Substitution (C→G) ABC transporter system; a putative Fe 3+ -siderophore-binding lipoprotein 5.1 2,521,778 413,291 S NCgl2129 Coding (110 bp) Substitution (A→T) Uncharacterized protein, cell membrane; membrane; signal; transmembrane; transmembrane helix. 1.43 2,589,950 481,463 S NCgl0891 Coding (709 bp) Substitution (T→G) HNH endonucleases 1.08 2,589,952 481,465 S NCgl0891 Coding (711 bp) Substitution (G→T) HNH endonucleases 1.08 2,708,045 599,558 K NCgl2298 Coding (757 bp) Substitution (G→A) HTH-type transcriptional repressor ( osrR ) 1.27 2,743,443 634,956 V NCgl2331 Coding (610 bp) Substitution (G→A) Penicillin binding protein 0.53 2,899,770 791,283 K NCgl2472 Coding (239 bp) Substitution (T→C) Transcriptional regulator MalT ( ramA ) 1.17 a Genome sequence data NCBI bioproject number: PRJNA554987 . b COG functional categories (M, cell wall/membrane/envelope biogenesis; T, signal transduction mechanisms; V, defense mechanisms; K, transcription; L, replication, recombination and repair; C, energy production and conversion; E, amino acid transport and metabolism; P, inorganic ion transport and metabolism; S, function unknown). c Data adopted from the transcriptome (NCBI bioproject number: PRJNA556334 ) ( Park et al., 2020 ). Mutation Effect on Iron Consumption The mutations observed in two presumed siderophore-binding lipoproteins (NCgl0774: 35-bp repeated mutation; NCgl1959: point substitution mutation at -14 bp) suggested their possible role in the rapid reproduction of JH41. To explore the functional significance of these putative iron-importing systems ( Ribeiro and Simões, 2019 ), the iron consumption of JH41 was profiled ( Figure 2B ). Iron concentration in the medium decreased more rapidly in the JH41 strain. Since iron is an essential element for bacteria and acts as a redox center for several cytochromes and iron–sulfur proteins in the electron transfer chain ( Hider and Kong, 2010 ), it was assumed that the rapidly reproducing JH41 might harbor greater energy than its ancestor. To validate this, ATP concentrations of the fast-growing JH41 and ancestral PT were analyzed. Indeed, the intracellular ATP concentrations of JH41 were greater than those of the PT strain with a range of 19–35% during the active growing phase ( Figure 2C ). FIGURE 2 Profiles of growth, iron consumption, and intracellular ATP concentration of the JH41 host. (A) Batch culture growth profiles of PT and JH41 hosts. Cultures were in a 500-ml baffled flask containing 50 ml of modified MCGC medium. Each data point represents the mean ± SD; n = 3 biologically independent samples. (PT: white; JH41: black). (B) Iron consumption in PT and JH41 hosts. Samples were taken from the batch cultures (500-ml baffled flask containing 50 ml of modified MCGC medium). Iron concentration of the medium was measured by the o -phenanthroline colorimetric method. Error bars show a statistical mean ± SD; n = 3 biologically independent samples. (PT: white; JH41: black). (C) Intracellular ATP concentrations in PT and JH41 strains. Samples were taken from the batch cultures (500-ml baffled flask containing 50 ml of modified MCGC medium). Intracellular ATP was measured by the luciferin–luciferase reaction immediately after disrupting cells. Error bars show a statistical mean ± SD; n = 3 biologically independent samples (PT: white; JH41: black). To assess which of the putative siderophore-binding lipoproteins affected iron consumption, a PT strain harboring the genomic NCgl0774 gene with the precise 35-bp repeated sequence (PT NCgl0774:35-bp) was constructed for complementation tests. The constructed PT NCgl0774:35-bp strain showed no differences compared with the PT strain in the growth profile ( Supplementary Figure 2 ). This result indicated that the 35-bp repeat mutation on the putative siderophore-binding protein A was not a prominent factor contributing to the rapid reproduction of JH41. Another point mutation found in the JH41 strain was located at -14 bp of the start codon of NCgl1959 encoding a putative siderophore-binding lipoprotein B, in the middle of the consensus sequence of the DtxR binding site ( TTAGG C AAGG C TA C C T TT TTGCCTATG; bold: DtxR binding consensus sequence; underline: mutation C→G; italic: start codon), where DtxR is a repressor of the genes responsible for iron uptake and protection against free radical damage ( Wennerhold and Bott, 2006 ). The mutation on the DtxR binding sequence was assumed to derepress the downstream NCgl1959. To this end, we also found that the transcriptome data show a 5.1-fold upregulation of NCgl1959 in JH41 [ Table 2 ; messenger RNA (mRNA) fold based on transcriptome column]. To verify the hypothesis, the binding affinity between the NCgl1959 regulatory region and DtxR was investigated by gel shift assay ( Figure 3A ). The binding affinity of the His-tag purified DtxR protein on the DNA fragment carrying the point mutation was only 40% of the affinity of the original sequence, which suggested that the mutation would have caused the derepression of NCgl1959 by reducing affinity to the transcription repressor, DtxR. Further, complementation studies were performed by constitutively expressing NCgl1959, the putative siderophore-binding lipoprotein B, on a plasmid in the PT strain. The PT strain expressing NCgl1959 (PT + pSL360-SBP B) showed more rapid reproduction (75% increase) ( Figure 3B ), greater iron consumption ( Figure 3C ), and higher ATP concentrations ( Figure 3D ) than the parental PT strain. Therefore, the derepression or overexpression of the putative siderophore-binding protein (NCgl1959) enabled C. glutamicum to consume more iron and further increase intracellular ATP concentration. FIGURE 3 Effects of the mutation at 14 bases upstream of NCgl1959. (A) Native polyacrylamide gel image showing the affinity of DtxR protein to the normal DNA fragment (up) and to the mutated DNA fragment (down). L: size marker; lane 1: DNA only; lane 2: DNA + DtxR (87 ng); lane 3: DNA + DtxR (350 ng). The DNA fragments were stained with SYBR Green. The chromatogram on the right side represents the intensity of bands by ImageJ software. The image is from the representative experiment from three independent experiments. (B) Growth profile of PT strain harboring plasmid constitutively expressing NCgl1959. Each data point is the mean ± SD; n = 3 biologically independent samples. (PT + empty vector: white; PT + pSL360-SBP B: black). (C) Iron consumption in PT strain harboring plasmid constitutively expressing NCgl1959. The iron concentration in the medium was measured by the o -phenanthroline colorimetric method. All strains were cultured in a 500-ml baffled flask containing 50 ml of modified MCGC medium. Bar heights and error bars show a statistical mean ± SD; n = 3 biologically independent samples (PT + empty vector: white; PT + pSL360-SBP B: black). (D) Intracellular ATP concentrations in PT strain harboring plasmid constitutively expressing NCgl1959. The intracellular ATP luciferin–luciferase reaction was measured immediately after harvest during the actively growing phase. Point and error bars show a statistical mean ± SD; n = 3 biologically independent samples (PT + empty vector: white; PT + pSL360-SBP B: black). The point mutation found at the -108 region of the NCgl1159 (subunit A of F 0 F 1 -ATP synthase) might be involved in ATP generation ( Table 2 ). The growth profile of the PT host containing the genomic mutation at the -108 region of the NCgl1159 (PT NCgl1159 C–108T ), however, was not different from that of the PT host ( Supplementary Figure 3 ). Investigating further, known regulators related to iron metabolism were found to be disrupted in the PT genome. Along with DtxR, the repressor of the putative siderophore-binding lipoprotein B expression, RipA, which is known to be a gene-enhancing regulator for iron-consuming proteins and is repressed by DtxR ( Wennerhold et al., 2005 ), were chosen as the disruption targets. Part of the coding sequences of DtxR (1.9→1.0 kb) and RipA (1.7→0.9 kb) were removed from the genome of PT strain using a double crossover method. The PT Δ dtxR strain showed a 34% faster growth rate and the PT Δ ripA (47% slower growth rate) compared with PT strain ( Figure 4 ). The disruption of RipA (leading downregulation of iron-consuming genes) led the cell to double more slowly, and the disruption of DtxR (leading to overexpression of iron-importing genes) resulted in the faster doubling. FIGURE 4 Effect of genomic disruptions of iron-relevant regulator genes ( ripA, dtxR ) on the growth of PT strain. The PT Δ ripA and PT Δ dtxR strains were incubated at 30°C, 200 rpm in a shaking incubator. A 500-ml baffled flask containing 50 ml of the modified MCGC media was used for efficient oxygen transfer. The exponential growth rate (μ) in batch culture was determined by linear regression of log biomass concentrations over each process time. Each data point is the mean ± SD; n = 3 biologically independent samples (PT: white circle; PT Δ ripA : black circle; PT Δ dtxR : black rhombus). Mutation Effect on Glucose Consumption Another notable mutation in JH41 was on the NCgl2472 gene (A239G), encoding the LuxR-type global transcription regulator RamA protein, which controls the expression of genes involved in sugar uptake, glycolysis, acetate, and many other metabolic pathways ( Auchter et al., 2011 ). Wang et al. and Graf et al. reported a ramA mutation found in another fast-growing strain of C. glutamicum , and the mutation was described as a key mutation because it enabled substrate uptake and increased metabolic flux ( Wang et al., 2018 ; Graf et al., 2019 ). To verify that the ramA A239G mutation in the JH41 strain also contributed to the enhanced cell division, the genomic point mutation was introduced to PT strain by homologous recombination to construct PT ramA A239G strain. Indeed, the PT ramA A239G strain consumed glucose at 0.89 ± 0.03 g –glucose /g – DCW ⋠h, whereas PT and JH41 consumed it at 0.64 ± 0.04 and 1.11 ± 0.01 g –glucose /g – D CW ⋠h, respectively ( Figure 5 ). Therefore, the ramA A239G mutation led the PT strain to increase the rates of glucose consumption by 39% and growth by 40%. The increase in glucose uptake would have provided the increased precursor metabolites and energy to C. glutamicum cells. FIGURE 5 Effects of genomic ramA A239G mutation in PT strain on growth rate and glucose consumption. The growth rate and glucose consumption profiles of (A) PT, (B) JH41, and (C) PT ramA A239G . Batch cultures were performed in a 500-ml baffled flask containing 50 ml of 0.9% glucose modified MCGC medium. Cell densities of C. glutamicum PT, JH41, and PT ramA A239G were measured every 2 h. Glucose consumption was measured by 3,5-dinitrosalicylic acid (DNS) reduction sugar quantification method. Each data point represents mean ± SD; n = 3 biologically independent samples. (biomass: black; glucose concentration: white). Mutation Effect on Ribosome Components The ITS mutation (condensed 23 point mutations, Table 2 ) located in the non-coding region of rRNA primary transcript in the genome of JH41 was assumed to influence the maturation of rRNA from its pre-transcript, and this might contribute to the increased number of ribosomes. To confirm this hypothesis, concentrations of the mature 5S, 16S, and 23S rRNAs in the JH41 strain were compared with those in the PT strain by quantitative PCR ( Table 3 ). The fast-growing JH41 strain harbored higher concentrations of mature rRNAs (45% more 16S rRNA, 72% more 23S rRNA, 37% more 5S rRNA) compared to the PT strain. Along with increases in the three rRNAs, our previous paper also reported that most of the mRNAs encoding ribosomal proteins were upregulated ( Supplementary Table 2 ) ( Park et al., 2020 ). These results enabled us to conclude the ITS mutation in rRNA primary transcript supported the fast growth of JH41 by increases in matured rRNAs and ribosomal proteins, and therefore, the increase in ribosome components might be a reverse engineering target to enhance the production of recombinant protein even though the genomic introduction of the multiple mutations in the one ITS region among the 6 rRNA operons would be technically difficult. TABLE 3 Levels of intracellular rRNAs in JH41. Reference gene Target gene Fold difference (2 –ΔΔCT ) a gapA 16S rRNA 1.45 ± 0.37 23S rRNA 1.72 ± 0.64 5S rRNA 1.37 ± 0.42 leuA 16S rRNA 1.47 ± 0.07 23S rRNA 1.30 ± 0.10 5S rRNA 1.18 ± 0.15 a (rRNA ratio in JH41)/(rRNA ratio in PT). Values represent the statistical mean ± SD; n = 3 biologically independent samples. Combinatorial Effect of SBP B and RamA Mutations on Recombinant Protein In the above results, the mutations on the SBP B upregulation (SBP B C–14G ) and the glucose-uptake regulator ( ramA A239G ) were the clear genetic factors affecting the growth rate of JH41. To evaluate the combinatorial effect of the two mutations on the growth profile and recombinant protein production, the genome of the PT host was mutated to have either the SBP B C–14G mutation (PT SBP B C–14G host) or the ramA A239G mutation (PT ramA A239G host) or to have both mutations (PT SBP B C–14G ramA A239G , in short, PT S + R). The growth rate of the SBP B genome-edited host (PT SBP B C–14G ) was 0.56 h –1 (t D = 1.23 h) and that of the RamA genome-edited host (PT ramA A239G ) was 0.52 h –1 (t D = 1.33 h), which were 51 and 41% faster doubling, respectively, than that of PT host (growth rate = 0.37 h –1 , t D = 1.87 h). The double mutant of ramA A239G and SBP B C–14G host (PT S + R) showed the fasted growth rate of 0.66 h –1 (t D = 1.05 h), which was 78% higher than that of PT and even faster than that of the evolved JH41 host (0.61 h –1 , t D = 1.13 h) ( Figure 6A ). FIGURE 6 Effects of genomic ramA A239G and SBP B C–14G mutation in PT strain on growth rate and the secretory production of recombinant protein. (A) Batch culture growth profiles of PT and JH41 strains. The growth rate profiles of PT, JH41, PT ramA A239G , PT SBP B C–14G , and PT S + R. Batch cultures were performed in a 500-ml baffled flask containing 50 ml of 0.9% glucose modified MCGC medium. Cell densities of C. glutamicum PT, JH41, PT ramA A239G , PT SBP B C–14G , and PT S + R were measured every 1 h. Each data point represents mean ± SD; n = 3 biologically independent samples. (PT: white circle; JH41: black circle; PT ramA A239G : black triangle; PT SBP B C–14G : white triangle; PT S + R: Black square). (B) Image of sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE). Lane 1, PT + pCG-H36A (empty vector); lane 2, JH41 + pCG-H36A-agarase; lane 3, PT SBP B C–14G + pCG-H36A (empty vector); lane 4, PT SBP B C–14G + pCG-H36A-agarase; lane 5, PT ramA A239G + pCG-H36A (empty vector); lane 6, PT ramA A239G + pCG-H36A-agarase; lane 7, PT S + R + pCG-H36A (empty vector); lane 8, PT S + R + pCG-H36A-agarase. Agarase production was confirmed after 24 h cultivation. Arrow indicates the secreted agarase band (34 kDa). The image is from the representative experiment from three independent experiments. (C) Band density estimation of the SDS-PAGE image using ImageJ software. a: lane 1; b: lane 2; c: lane 4; d: lane 6; e: lane 8. The secretory protein productions in the PT SBP B C–14G and the PT ramA A239G hosts are also shown to be greater than that from the PT host ( Figures 6B,C ). The SDS-PAGE band intensity of the secretory recombinant proteins from the PT SBP B C–14G and the PT ramA A239G hosts were 8,149 and 6,146 AU, respectively, where those from PT host was 4,130 AU. The fastest growing PT S + R hosts, however, secreted the recombinant protein at only 4,064 AU, which is even lower than that of the PT host. Therefore, the mutations in ramA A239G and SBP B C–14G , respectively, resulted in the increased growth rates and protein production. Meanwhile, the combination of the two genes showed a remarkable increase in the growth rate but was found to have a negative effect on protein production. The iron metabolism regulator-deleted strains (PT Δ dtxR and PT Δ ripA ) also decreased the recombinant agarase production by 34 and 37%, respectively, than that from PT host ( Supplementary Figure 4 ). Effect of Fast-Growing JH41 Host on the Secretory Production of Recombinant Protein Because protein is the major macromolecule in cells, the rapidly reproducing JH41 strain was supposed to be able to produce a protein at an enhanced rate. The fact that no extracellular protease activity or economical purification of secretory protein was found in C. glutamicum host ( Lee and Kim, 2018 ) further encouraged us to hypothesize that the rapidly reproducing JH41 strain would become an improved C. glutamicum host for secretory protein production. To verify this hypothesis, a plasmid designed to secrete foreign agarase (pCG-H36A-agarase) was introduced into JH41 and PT strains. The secretory recombinant protein (agarase, 34 kDa) showed a major band in the SDS-PAGE ( Figure 1A ). The band intensities of the recombinant protein from the JH41 host were 7,490 AU at 24 h and 17,231 AU at 48 h, while those from the PT host were 4,780 AU at 24 h and 6,212 AU at 48 h ( Figure 1B ). The JH41 host produced the secretory recombinant protein (56% at 24 h and 177% at 48 h) at a greater rate than the PT host. In order to confirm the production effect of other proteins in the JH41 host, a plasmid expressing green fluorescent protein (GFP) in cytoplasm (pCG-H36A[-SS]-GFP, no signal sequence) and a plasmid secreting GFP by a TAT-signal sequence (pCG-H36A- porD -GFP) were introduced into the JH41 host ( Supplementary Figure 1 ). The productions of cytoplasmic GFP and secreted GFP from the JH41 host were greater than those from the PT host by 5–16% and 13–116%, respectively. Therefore, the rapid-reproducing JH41 strain was an enhanced host for the production of recombinant proteins. FIGURE 1 Effects of host on the secretory production of recombinant protein. (A) Image of sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE). Lane 1, PT + pCG-H36A (empty vector) at 24 h; lane 2, PT + pCG-H36A-agarase at 24 h; lane 3, JH41 + pCG-H36A (empty vector) at 24 h; lane 4, JH41 + pCG-H36A-agarase at 24 h; lane 5, PT + pCG-H36A (empty vector) at 48 h; lane 6, PT + pCG-H36A-agarase at 48 h; lane 7, JH41 + pCG-H36A (empty vector) at 48 h; lane 8, JH41 + pCG-H36A-agarase at 48 h. Arrow indicates the secreted agarase band (34 kDa). The image is from the representative experiment from three independent experiments. (B) Band density estimation of the SDS-PAGE image using ImageJ software. a: lane 2; b: lane 4; c: lane 6; d: lane 8. Genome Analysis of the JH41 Host To understand the reasons for enhanced cell reproduction and to provide reverse engineering targets responsible for the improved recombinant protein production, the JH41 genome was resequenced and deposited [National Center for Biotechnology Information (NCBI) accession number PRJNA554987 ]. A total 15 genomic mutations were found to be accumulated in JH41 compared to its ancestral PT ( Table 2 ). Among them, four were found in the non-coding region, including an internal transcribed spacer (ITS) mutation, and 11 in the coding region. Genomic mutations were categorized into the clusters of orthologous groups of proteins (COG) as four to the function unknown group (S), two to the transcription group (K), two to the inorganic ion transport and metabolism group (P), one to the defense mechanism (V), one to the signal transduction mechanism (T), and one to the energy production and conversion group (C). The remaining three mutations did not show protein function or were not protein-encoding regions. A 35-bp repeated mutation in the structural gene encoding a putative siderophore-binding lipoprotein A (SBP A, NCgl0774) and a substitution mutation (C→G) at the 14 bases upstream of the start codon of the gene encoding a putative siderophore-binding lipoprotein B (SBP B, NCgl1959) were identified. A condensed mutation was also found in one of the operons encoding rRNA primary transcript. One mutation was found in the 16S rRNA structural gene, and another 23-point mutations were densely observed between the 16S and 23S rRNA sequences, which are called 16S–23S internal transcribed spacer (ITS) ( Osorio et al., 2005 ). TABLE 2 Genomic mutations of the JH41 host over ancestral host. PT position a JH41 position a COG functional category b Locus tag Mutation region Variants Function (gene name) mRNA fold based on transcriptome c) 80,225 1,076,088 S NCgl2726 Non-coding (-57 bp of NCgl2726) Substitution (G→A) Tripartite tricarboxylate transporter substrate binding protein 3.05 461,943 1,457,809 - - 16S rRNA coding region Substitution (A→G) Bacterial small subunit ribosomal RNA; 16S ribosomal RNA – 462,350 1,458,217 Non-coding (16S-23S ITS of 1 st rRNA operon) Similar to 16S–23S ITS of 5th rRNA operon rRNA operon 16S-23S ITS (internal transcribed spacer) – 1,191,984 2,187,859 T NCgl0722 Coding (372 bp) Insertion (C) Histidine protein kinase ( mtrB ) 1.27 1,251,401 2,247,312 P NCgl0774 Coding (15 bp) Repeat (35 bp, AAAACTCTCCTCCGCA CTGATCGTGCTTCTCG CAG) ABC transporter system; A putative Fe 3+ -siderophore-binding lipoprotein 0.03 1,660,001 2,655,914 C NCgl1159 Non-coding (-108 bp of NCgl1159) Substitution (C→T) ATP synthase F 0 F 1 subunit A 0.27 1,968,380 2,964,295 K NCgl1435 Coding (201 bp) Substitution (A→G) Transcriptional regulator 1.17 2,004,381 3,000,298 – – Coding (67 bp) Substitution (G→A) Hypothetical protein – 2,326,488 217,999 P NCgl1959 Non-coding (-14 bp of NCgl1959) Substitution (C→G) ABC transporter system; a putative Fe 3+ -siderophore-binding lipoprotein 5.1 2,521,778 413,291 S NCgl2129 Coding (110 bp) Substitution (A→T) Uncharacterized protein, cell membrane; membrane; signal; transmembrane; transmembrane helix. 1.43 2,589,950 481,463 S NCgl0891 Coding (709 bp) Substitution (T→G) HNH endonucleases 1.08 2,589,952 481,465 S NCgl0891 Coding (711 bp) Substitution (G→T) HNH endonucleases 1.08 2,708,045 599,558 K NCgl2298 Coding (757 bp) Substitution (G→A) HTH-type transcriptional repressor ( osrR ) 1.27 2,743,443 634,956 V NCgl2331 Coding (610 bp) Substitution (G→A) Penicillin binding protein 0.53 2,899,770 791,283 K NCgl2472 Coding (239 bp) Substitution (T→C) Transcriptional regulator MalT ( ramA ) 1.17 a Genome sequence data NCBI bioproject number: PRJNA554987 . b COG functional categories (M, cell wall/membrane/envelope biogenesis; T, signal transduction mechanisms; V, defense mechanisms; K, transcription; L, replication, recombination and repair; C, energy production and conversion; E, amino acid transport and metabolism; P, inorganic ion transport and metabolism; S, function unknown). c Data adopted from the transcriptome (NCBI bioproject number: PRJNA556334 ) ( Park et al., 2020 ). Mutation Effect on Iron Consumption The mutations observed in two presumed siderophore-binding lipoproteins (NCgl0774: 35-bp repeated mutation; NCgl1959: point substitution mutation at -14 bp) suggested their possible role in the rapid reproduction of JH41. To explore the functional significance of these putative iron-importing systems ( Ribeiro and Simões, 2019 ), the iron consumption of JH41 was profiled ( Figure 2B ). Iron concentration in the medium decreased more rapidly in the JH41 strain. Since iron is an essential element for bacteria and acts as a redox center for several cytochromes and iron–sulfur proteins in the electron transfer chain ( Hider and Kong, 2010 ), it was assumed that the rapidly reproducing JH41 might harbor greater energy than its ancestor. To validate this, ATP concentrations of the fast-growing JH41 and ancestral PT were analyzed. Indeed, the intracellular ATP concentrations of JH41 were greater than those of the PT strain with a range of 19–35% during the active growing phase ( Figure 2C ). FIGURE 2 Profiles of growth, iron consumption, and intracellular ATP concentration of the JH41 host. (A) Batch culture growth profiles of PT and JH41 hosts. Cultures were in a 500-ml baffled flask containing 50 ml of modified MCGC medium. Each data point represents the mean ± SD; n = 3 biologically independent samples. (PT: white; JH41: black). (B) Iron consumption in PT and JH41 hosts. Samples were taken from the batch cultures (500-ml baffled flask containing 50 ml of modified MCGC medium). Iron concentration of the medium was measured by the o -phenanthroline colorimetric method. Error bars show a statistical mean ± SD; n = 3 biologically independent samples. (PT: white; JH41: black). (C) Intracellular ATP concentrations in PT and JH41 strains. Samples were taken from the batch cultures (500-ml baffled flask containing 50 ml of modified MCGC medium). Intracellular ATP was measured by the luciferin–luciferase reaction immediately after disrupting cells. Error bars show a statistical mean ± SD; n = 3 biologically independent samples (PT: white; JH41: black). To assess which of the putative siderophore-binding lipoproteins affected iron consumption, a PT strain harboring the genomic NCgl0774 gene with the precise 35-bp repeated sequence (PT NCgl0774:35-bp) was constructed for complementation tests. The constructed PT NCgl0774:35-bp strain showed no differences compared with the PT strain in the growth profile ( Supplementary Figure 2 ). This result indicated that the 35-bp repeat mutation on the putative siderophore-binding protein A was not a prominent factor contributing to the rapid reproduction of JH41. Another point mutation found in the JH41 strain was located at -14 bp of the start codon of NCgl1959 encoding a putative siderophore-binding lipoprotein B, in the middle of the consensus sequence of the DtxR binding site ( TTAGG C AAGG C TA C C T TT TTGCCTATG; bold: DtxR binding consensus sequence; underline: mutation C→G; italic: start codon), where DtxR is a repressor of the genes responsible for iron uptake and protection against free radical damage ( Wennerhold and Bott, 2006 ). The mutation on the DtxR binding sequence was assumed to derepress the downstream NCgl1959. To this end, we also found that the transcriptome data show a 5.1-fold upregulation of NCgl1959 in JH41 [ Table 2 ; messenger RNA (mRNA) fold based on transcriptome column]. To verify the hypothesis, the binding affinity between the NCgl1959 regulatory region and DtxR was investigated by gel shift assay ( Figure 3A ). The binding affinity of the His-tag purified DtxR protein on the DNA fragment carrying the point mutation was only 40% of the affinity of the original sequence, which suggested that the mutation would have caused the derepression of NCgl1959 by reducing affinity to the transcription repressor, DtxR. Further, complementation studies were performed by constitutively expressing NCgl1959, the putative siderophore-binding lipoprotein B, on a plasmid in the PT strain. The PT strain expressing NCgl1959 (PT + pSL360-SBP B) showed more rapid reproduction (75% increase) ( Figure 3B ), greater iron consumption ( Figure 3C ), and higher ATP concentrations ( Figure 3D ) than the parental PT strain. Therefore, the derepression or overexpression of the putative siderophore-binding protein (NCgl1959) enabled C. glutamicum to consume more iron and further increase intracellular ATP concentration. FIGURE 3 Effects of the mutation at 14 bases upstream of NCgl1959. (A) Native polyacrylamide gel image showing the affinity of DtxR protein to the normal DNA fragment (up) and to the mutated DNA fragment (down). L: size marker; lane 1: DNA only; lane 2: DNA + DtxR (87 ng); lane 3: DNA + DtxR (350 ng). The DNA fragments were stained with SYBR Green. The chromatogram on the right side represents the intensity of bands by ImageJ software. The image is from the representative experiment from three independent experiments. (B) Growth profile of PT strain harboring plasmid constitutively expressing NCgl1959. Each data point is the mean ± SD; n = 3 biologically independent samples. (PT + empty vector: white; PT + pSL360-SBP B: black). (C) Iron consumption in PT strain harboring plasmid constitutively expressing NCgl1959. The iron concentration in the medium was measured by the o -phenanthroline colorimetric method. All strains were cultured in a 500-ml baffled flask containing 50 ml of modified MCGC medium. Bar heights and error bars show a statistical mean ± SD; n = 3 biologically independent samples (PT + empty vector: white; PT + pSL360-SBP B: black). (D) Intracellular ATP concentrations in PT strain harboring plasmid constitutively expressing NCgl1959. The intracellular ATP luciferin–luciferase reaction was measured immediately after harvest during the actively growing phase. Point and error bars show a statistical mean ± SD; n = 3 biologically independent samples (PT + empty vector: white; PT + pSL360-SBP B: black). The point mutation found at the -108 region of the NCgl1159 (subunit A of F 0 F 1 -ATP synthase) might be involved in ATP generation ( Table 2 ). The growth profile of the PT host containing the genomic mutation at the -108 region of the NCgl1159 (PT NCgl1159 C–108T ), however, was not different from that of the PT host ( Supplementary Figure 3 ). Investigating further, known regulators related to iron metabolism were found to be disrupted in the PT genome. Along with DtxR, the repressor of the putative siderophore-binding lipoprotein B expression, RipA, which is known to be a gene-enhancing regulator for iron-consuming proteins and is repressed by DtxR ( Wennerhold et al., 2005 ), were chosen as the disruption targets. Part of the coding sequences of DtxR (1.9→1.0 kb) and RipA (1.7→0.9 kb) were removed from the genome of PT strain using a double crossover method. The PT Δ dtxR strain showed a 34% faster growth rate and the PT Δ ripA (47% slower growth rate) compared with PT strain ( Figure 4 ). The disruption of RipA (leading downregulation of iron-consuming genes) led the cell to double more slowly, and the disruption of DtxR (leading to overexpression of iron-importing genes) resulted in the faster doubling. FIGURE 4 Effect of genomic disruptions of iron-relevant regulator genes ( ripA, dtxR ) on the growth of PT strain. The PT Δ ripA and PT Δ dtxR strains were incubated at 30°C, 200 rpm in a shaking incubator. A 500-ml baffled flask containing 50 ml of the modified MCGC media was used for efficient oxygen transfer. The exponential growth rate (μ) in batch culture was determined by linear regression of log biomass concentrations over each process time. Each data point is the mean ± SD; n = 3 biologically independent samples (PT: white circle; PT Δ ripA : black circle; PT Δ dtxR : black rhombus). Mutation Effect on Glucose Consumption Another notable mutation in JH41 was on the NCgl2472 gene (A239G), encoding the LuxR-type global transcription regulator RamA protein, which controls the expression of genes involved in sugar uptake, glycolysis, acetate, and many other metabolic pathways ( Auchter et al., 2011 ). Wang et al. and Graf et al. reported a ramA mutation found in another fast-growing strain of C. glutamicum , and the mutation was described as a key mutation because it enabled substrate uptake and increased metabolic flux ( Wang et al., 2018 ; Graf et al., 2019 ). To verify that the ramA A239G mutation in the JH41 strain also contributed to the enhanced cell division, the genomic point mutation was introduced to PT strain by homologous recombination to construct PT ramA A239G strain. Indeed, the PT ramA A239G strain consumed glucose at 0.89 ± 0.03 g –glucose /g – DCW ⋠h, whereas PT and JH41 consumed it at 0.64 ± 0.04 and 1.11 ± 0.01 g –glucose /g – D CW ⋠h, respectively ( Figure 5 ). Therefore, the ramA A239G mutation led the PT strain to increase the rates of glucose consumption by 39% and growth by 40%. The increase in glucose uptake would have provided the increased precursor metabolites and energy to C. glutamicum cells. FIGURE 5 Effects of genomic ramA A239G mutation in PT strain on growth rate and glucose consumption. The growth rate and glucose consumption profiles of (A) PT, (B) JH41, and (C) PT ramA A239G . Batch cultures were performed in a 500-ml baffled flask containing 50 ml of 0.9% glucose modified MCGC medium. Cell densities of C. glutamicum PT, JH41, and PT ramA A239G were measured every 2 h. Glucose consumption was measured by 3,5-dinitrosalicylic acid (DNS) reduction sugar quantification method. Each data point represents mean ± SD; n = 3 biologically independent samples. (biomass: black; glucose concentration: white). Mutation Effect on Ribosome Components The ITS mutation (condensed 23 point mutations, Table 2 ) located in the non-coding region of rRNA primary transcript in the genome of JH41 was assumed to influence the maturation of rRNA from its pre-transcript, and this might contribute to the increased number of ribosomes. To confirm this hypothesis, concentrations of the mature 5S, 16S, and 23S rRNAs in the JH41 strain were compared with those in the PT strain by quantitative PCR ( Table 3 ). The fast-growing JH41 strain harbored higher concentrations of mature rRNAs (45% more 16S rRNA, 72% more 23S rRNA, 37% more 5S rRNA) compared to the PT strain. Along with increases in the three rRNAs, our previous paper also reported that most of the mRNAs encoding ribosomal proteins were upregulated ( Supplementary Table 2 ) ( Park et al., 2020 ). These results enabled us to conclude the ITS mutation in rRNA primary transcript supported the fast growth of JH41 by increases in matured rRNAs and ribosomal proteins, and therefore, the increase in ribosome components might be a reverse engineering target to enhance the production of recombinant protein even though the genomic introduction of the multiple mutations in the one ITS region among the 6 rRNA operons would be technically difficult. TABLE 3 Levels of intracellular rRNAs in JH41. Reference gene Target gene Fold difference (2 –ΔΔCT ) a gapA 16S rRNA 1.45 ± 0.37 23S rRNA 1.72 ± 0.64 5S rRNA 1.37 ± 0.42 leuA 16S rRNA 1.47 ± 0.07 23S rRNA 1.30 ± 0.10 5S rRNA 1.18 ± 0.15 a (rRNA ratio in JH41)/(rRNA ratio in PT). Values represent the statistical mean ± SD; n = 3 biologically independent samples. Combinatorial Effect of SBP B and RamA Mutations on Recombinant Protein In the above results, the mutations on the SBP B upregulation (SBP B C–14G ) and the glucose-uptake regulator ( ramA A239G ) were the clear genetic factors affecting the growth rate of JH41. To evaluate the combinatorial effect of the two mutations on the growth profile and recombinant protein production, the genome of the PT host was mutated to have either the SBP B C–14G mutation (PT SBP B C–14G host) or the ramA A239G mutation (PT ramA A239G host) or to have both mutations (PT SBP B C–14G ramA A239G , in short, PT S + R). The growth rate of the SBP B genome-edited host (PT SBP B C–14G ) was 0.56 h –1 (t D = 1.23 h) and that of the RamA genome-edited host (PT ramA A239G ) was 0.52 h –1 (t D = 1.33 h), which were 51 and 41% faster doubling, respectively, than that of PT host (growth rate = 0.37 h –1 , t D = 1.87 h). The double mutant of ramA A239G and SBP B C–14G host (PT S + R) showed the fasted growth rate of 0.66 h –1 (t D = 1.05 h), which was 78% higher than that of PT and even faster than that of the evolved JH41 host (0.61 h –1 , t D = 1.13 h) ( Figure 6A ). FIGURE 6 Effects of genomic ramA A239G and SBP B C–14G mutation in PT strain on growth rate and the secretory production of recombinant protein. (A) Batch culture growth profiles of PT and JH41 strains. The growth rate profiles of PT, JH41, PT ramA A239G , PT SBP B C–14G , and PT S + R. Batch cultures were performed in a 500-ml baffled flask containing 50 ml of 0.9% glucose modified MCGC medium. Cell densities of C. glutamicum PT, JH41, PT ramA A239G , PT SBP B C–14G , and PT S + R were measured every 1 h. Each data point represents mean ± SD; n = 3 biologically independent samples. (PT: white circle; JH41: black circle; PT ramA A239G : black triangle; PT SBP B C–14G : white triangle; PT S + R: Black square). (B) Image of sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE). Lane 1, PT + pCG-H36A (empty vector); lane 2, JH41 + pCG-H36A-agarase; lane 3, PT SBP B C–14G + pCG-H36A (empty vector); lane 4, PT SBP B C–14G + pCG-H36A-agarase; lane 5, PT ramA A239G + pCG-H36A (empty vector); lane 6, PT ramA A239G + pCG-H36A-agarase; lane 7, PT S + R + pCG-H36A (empty vector); lane 8, PT S + R + pCG-H36A-agarase. Agarase production was confirmed after 24 h cultivation. Arrow indicates the secreted agarase band (34 kDa). The image is from the representative experiment from three independent experiments. (C) Band density estimation of the SDS-PAGE image using ImageJ software. a: lane 1; b: lane 2; c: lane 4; d: lane 6; e: lane 8. The secretory protein productions in the PT SBP B C–14G and the PT ramA A239G hosts are also shown to be greater than that from the PT host ( Figures 6B,C ). The SDS-PAGE band intensity of the secretory recombinant proteins from the PT SBP B C–14G and the PT ramA A239G hosts were 8,149 and 6,146 AU, respectively, where those from PT host was 4,130 AU. The fastest growing PT S + R hosts, however, secreted the recombinant protein at only 4,064 AU, which is even lower than that of the PT host. Therefore, the mutations in ramA A239G and SBP B C–14G , respectively, resulted in the increased growth rates and protein production. Meanwhile, the combination of the two genes showed a remarkable increase in the growth rate but was found to have a negative effect on protein production. The iron metabolism regulator-deleted strains (PT Δ dtxR and PT Δ ripA ) also decreased the recombinant agarase production by 34 and 37%, respectively, than that from PT host ( Supplementary Figure 4 ). Discussion The evolved C. glutamicum JH41 strain with increased growth rate was able to produce 2.7-fold greater secretory recombinant protein than PT host ( Figure 1 ). The confirmed genetic factors that contributed to the increased growth rates and became the reverse engineering targets to enhance production of recombinant proteins were (1) SBP B C–14G that enhanced iron transport leading the increase of intracellular energy and (2) ramA A239G that enhanced glucose uptake leading to the increase in energy and metabolite pool. High levels of intracellular ATP have been known to improve target products by improving substrate uptake, cell growth, biosynthesis, and resistance to toxic compounds ( Hara and Kondo, 2015 ). The authors also reported that the productions of recombinant proteins (green fluorescence protein and alkaline protease) were increased in the highly energized E. coli host by the overexpression of phosphoenolpyruvate carboxykinase (Pck) ( Kim et al., 2012 ). A high level of intracellular energy can provide favorable conditions for energy-consuming reactions such as charging amino acids to transfer RNA (tRNA), translational elongation, folding, and secretion. In addition, a high intracellular energy state can upregulate the genes involved in the biosynthetic pathways for amino acid supply ( Kwon et al., 2008 ). The point substitution mutation in the regulatory region of NCgl1959 in JH41 strain, resulted in the increase in iron consumption by derepressing the cognate genes ( Figure 3 ). Although the function of NCgl1959 protein in C. glutamicum has not yet been clearly identified, the NCgl1959 protein was estimated to have a similar 3D structure with the known siderophore-binding lipoproteins ( E. coli FhuD and Bacillus cereus YfiY) ( Clarke et al., 2000 ; Zawadzka et al., 2009 ) ( Supplementary Figure 5 ). One might speculate that NCgl1959 would play the role of iron complex uptake. The imported iron could be used for the assembly of electron transfer chain components (i.e., iron–sulfur reaction center for NADH dehydrogenase; heme–iron for cytochrome c oxidase and cytochromes) and redox-containing proteins (i.e., heme-dependent catalase and peroxidase) ( Park et al., 2020 ), and therefore, it is considered that the enhanced iron consumption in JH41 strain would contribute to generating more energy to the respiratory chain, along with a simultaneous increase in defense capability against oxidative stress. In particular, RamA has been shown to act as a positive regulator of the TCA cycle genes, sdhCAB (succinate dehydrogenase operon) and acn (aconitase) ( Bussmann et al., 2009 ; Teramoto et al., 2011 ). The point mutation ramA A239G (corresponding to RamA Y80C ) of JH41 strain was located at the GAF-2 domain (amino acid positions 8 and 146), the same domain of the reports from Graf et al. and Wang et al. ( Wang et al., 2018 ; Graf et al., 2019 ), and the GAF domain has been found in the cyclic nucleotide phosphodiesterase superfamily in other species ( Fawcett et al., 2000 ; Schultz, 2009 ). Therefore, we considered that the ramA A239G would have enabled the JH41 strain to increase glucose consumption by altering the cyclic nucleotide-mediated signals, thereby resulting in increased substrate uptake. The increase in carbon source in the JH41 host would have been favorable cellular conditions for the recombinant protein production by providing enhanced pools of precursor metabolites, along with more energy. We do not understand why the combination of the two mutations of ramA A239G and SBP B C–14G showed a faster growth rate exceeding JH41 and less production of recombinant protein ( Figure 6 ). The H36 promoter for the recombinant protein production was one of the most powerful synthetic promoter known in C. glutamicum ( Yim et al., 2013 ), and it was able to promote the secretory production of agarase from the PT ramA A239G and the PT SBP B C–14G along with the JH41 hosts. There might be unknown factor responsible for this mismatch between fast growth and less protein production ( Nordholt et al., 2017 ). Further basic studies on cellular physiology are required to understand this unexpected phenomenon. The condensed 23 point mutations occurred in the ITS region of the first rRNA operon among the six rRNA operons of C. glutamicum JH41. The multimutated first ITS sequence was similar to the sequence of the ITS in the fifth rRNA operon. A tRNA gene is located between 16S rRNA and 23S rRNA in E. coli rRNA pretranscript, while C. glutamicum does not contain tRNA in this sequence region. Although the processing mechanism of the rRNA pretranscript in C. glutamicum remains unknown, the ITS mutation in JH41 might have facilitated RNase processing, as in the case of E. coli . Therefore, the processing of the polycistronic primary transcript might have been increased to form mature ribosomes by the ITS mutation, having observed the increases in rRNAs and ribosomal proteins. The increase in mature ribosome components would contribute to the enhancement of recombinant protein productions, as observed in JH41. Additional genetic factors enhancing growth and recombinant protein production remain veiled. It is worth mentioning another complementation on the genomic mutation on the coding region of NCgl2298 ( osrR , G757A = D253N), an HTH-type transcription regulator known to upregulate the tolerance genes against oxidative stress and heat stress ( Hong et al., 2016 ). The mutation on the osrR was expected to release the stress tension that was provoked by the enhanced iron and carbon uptakes. The growth of the PT NCgl2298 G757A was, however, no better than that of the PT strain ( Supplementary Figure 6A ). The combination of mutations (i.e., the mutations on ramA + osrR ) would be required for further understanding. Besides, a mutation in the -109 region of NCgl1159 (F 0 F 1 -ATP synthase subunit A), which was expected to affect cell growth by altering ATP generation, did not increase the growth rate ( Supplementary Figure 3 ). The transcription levels of ATP synthase components in JH41 were only half those in our previous paper ( Park et al., 2020 ), so the half-expressed ATP synthase might have been somehow balanced for the enhanced energy generation under the increased iron and carbon uptake conditions. Another possible mutation to consider is NCgl2331, a penicillin-binding protein involved in cell defense. Substituting A for the 610th G of the NCgl2331 gene in the PT strain resulted in marginal increase (14%) in growth rate ( Supplementary Figure 6B ). Penicillin-binding proteins are membrane-associated macromolecules and are expected to play a key role in cell biosynthesis and cell division ( Macheboeuf et al., 2006 ). In this respect, our discovery offers essential cues for the construction of an efficient cell factory for the production of recombinant proteins and expands the biological knowledge base. Studies on the reverse engineering of the other mutations including the ITS condensed mutations might lead us to further improvement on C. glutamicum host. Data Availability Statement The JH41 genome was resequenced and deposited in NCBI accession number: PRJNA554987 . Author Contributions ML, JP, and KP performed the experiments. ML and PK designed the experiments. JK and PK raised grant and supervised. ML and PK wrote the manuscript. All authors contributed to the article and approved the submitted version. Conflict of Interest The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Supplementary Material The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fbioe.2020.588070/full#supplementary-material Click here for additional data file.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1287610/

Comparison of Minisatellite Polymorphisms in the Bacillus cereus Complex: a Simple Assay for Large-Scale Screening and Identification of Strains Most Closely Related to Bacillus anthracis

Polymorphism of five tandem repeats that are monomorphic in Bacillus anthracis was investigated in 230 isolates of the B. cereus group and in 5 sequenced B. cereus genomes in search for markers allowing identification of B. cereus and B. thuringiensis strains most closely related to B. anthracis . Using this multiple-locus variable number of tandem repeat analysis (MLVA), a cluster of 30 strains was selected for further characterization. Eventually, six of these were characterized by multilocus sequence type analysis. One of the strains is only six point mutations (of almost 3,000 bp) away from B. anthracis and was also proposed to be closest to B. anthracis by MLVA analysis. However, this strain remains separated from B. anthracis by a number of significant genetic events observed in B. anthracis , including the loss of the hemolysin activity, the presence of four prophages, and the presence of the two virulence plasmids, pXO1 and pXO2. One particular minisatellite marker provides an efficient assay to identify the subset of B. cereus and B. thuringiensis strains closely related to B. anthracis . Based on these results, a very simple assay is proposed that allows the screening of hundreds of strains from the B. cereus complex, with modest equipment and at a low cost, to eventually fill the gap with B. anthracis and better understand the origin and making of this dangerous pathogen.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4776515/

Proteomic comparison defines novel markers to characterize heterogeneous populations of extracellular vesicle subtypes

Significance The last decade has seen a rapid expansion of interest in extracellular vesicles (EVs), proposed to mediate cell–cell communication in patho/physiological conditions. Although heterogeneity of EVs has become obvious, as highlighted recently by the International Society for Extracellular Vesicles, the field is lacking specific tools to distinguish EVs of different intracellular origins, and thus probably different functions. Here, thanks to a comprehensive comparison of different types of EVs isolated from a single cell type, we define proteins generically present in EVs, small EV-specific and -excluded ones, and a few specific of endosome-derived exosomes or nonexosomal small EVs. This work will allow proper evaluation of the molecular mechanisms of biogenesis and secretion and the respective functions of subtypes of EVs.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2812294/

GC Content-Based Pan-Pox Universal PCR Assays for Poxvirus Detection ▿

Chordopoxviruses of the subfamily Chordopoxvirinae , family Poxviridae , infect vertebrates and consist of at least eight genera with broad host ranges. For most chordopoxviruses, the number of viral genes and their relative order are highly conserved in the central region. The GC content of chordopoxvirus genomes, however, evolved into two distinct types: those with genome GC content of more than 60% and those with a content of less than 40% GC. Two standard PCR assays were developed to identify chordopoxviruses based on whether the target virus has a low or high GC content. In design of the assays, the genus Avipoxvirus , which encodes major rearrangements of gene clusters, was excluded. These pan-pox assays amplify DNA from more than 150 different isolates and strains, including from primary clinical materials, from all seven targeted genera of chordopoxviruses and four unclassified new poxvirus species. The pan-pox assays represent an important advance for the screening and diagnosis of human and animal poxvirus infections, and the technology used is accessible to many laboratories worldwide.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3456558/

Clinical evaluation of the Emergency Medical Services (EMS) ambulance dispatch-based syndromic surveillance system, New York City

Since 1998, the New York City Department of Health has used New York City Emergency Medical Services (EMS) ambulance dispatch data to monitor for a communitywide rise in influenzalike illness (ILI) as an early detection system for bioterrorism. A clinical validation study was conducted during peak influenza season at six New York City emergency deparments (EDs) to compare patients with ILI brought in by ambulance with other patients to examine potential biases associated with ambulance dispatch-based surveillance. We also examined the utility of 4 EMS call types (selected from 52) for case detection of ILI. Clinical ILI was defined as fever (temperature higher than 100°F) on history or exam, along with either cough or sore throat. Of the 2,294 ED visits reviewed, 522 patients (23%) met the case definition for ILI, 64 (12%) of whom arrived by ambulance. Patients with ILI brought in by ambulance were older, complained of more severe symptoms, and were more likely to undergo diagnostic testing, be diagnosed with pneumonia, and be admitted to the hospital than patients who arrived by other means. The median duration of symptoms prior to presenting to the ED, however, was the same for both groups (48 hours). The selected call types had a sensitivity of 58% for clinical ILI, and a predictive value positive of 22%. Individuals with symptoms consistent with the prodrome of inhalational anthrax were likely to utilize the EMS system and usually did so early in the course of illness. While EMS-based surveillance is more sensitive for severe illness and for illness affecting older individuals, there is not necessarily a loss of timeliness associated with EMS-based (versus ED-based) surveillance.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4531518/

Surveillance systems for neglected tropical diseases: global lessons from China’s evolving schistosomiasis reporting systems, 1949–2014

Though it has been a focus of the country's public health surveillance systems since the 1950s, schistosomiasis represents an ongoing public health challenge in China. Parallel, schistosomiasis-specific surveillance systems have been essential to China's decades-long campaign to reduce the prevalence of the disease, and have contributed to the successful elimination in five of China's twelve historically endemic provinces, and to the achievement of morbidity and transmission control in the other seven. More recently, an ambitious goal of achieving nation-wide transmission interruption by 2020 has been proposed. This paper details how schistosomiasis surveillance systems have been structured and restructured within China's evolving public health system, and how parallel surveillance activities have provided an information system that has been integral to the characterization of, response to, and control of the disease. With the ongoing threat of re-emergence of schistosomiasis in areas previously considered to have achieved transmission control, a critical examination of China's current surveillance capabilities is needed to direct future investments in health information systems and to enable improved coordination between systems in support of ongoing control. Lessons drawn from China's experience are applied to the current global movement to reduce the burden of helminthiases, where surveillance capacity based on improved diagnostics is urgently needed. Introduction Schistosomiasis poses a major public health threat in many tropical and sub-tropical countries—the disease is endemic to 78 countries worldwide and, in 2011, at least 243 million people were infected and needed treatment [ 1 ]. In China, despite substantial progress achieved reducing the burden of schistosomiasis over the past two decades, the disease remains an ongoing public health challenge [ 2 , 3 ]. Based on the most recent data, nearly 250,000 people are estimated to be infected in 171 counties in China [ 4 ]. Since the 1950s, endemic areas have been identified in 454 counties in 12 provinces [ 3 , 4 ]. As of 2012, transmission had been eliminated in 5 of the 12 endemic provinces, and morbidity and transmission control had been attained in the other 7 (see official control categories and their criteria in Table 1 ; [ 5 , 6 ]). The country has set an ambitious goal of achieving nation-wide transmission interruption by 2020, however several obstacles lie in the way, including re-emergence in previously controlled areas [ 7 ], as well as challenges achieving lasting interruption of transmission in remaining endemic foci [ 3 , 4 ]. Table 1 Criteria for assigning county-level schistosomiasis transmission status Transmission status Criteria Infection control Human infection prevalence <5% Bovine infection prevalence <5% <10 acute cases over 2 week period in a village Transmission control Human infection prevalence <1% Bovine infection prevalence <1% No acute cases No infected snails ( O. hupensis ) for two consecutive years Transmission interruption No human cases for five consecutive years No bovine cases for five consecutive years No snails ( O. hupensis ) for two consecutive years Elimination No new infection in humans or bovines for five years after reaching transmission interruption Note: All criteria are evaluated at the administrative village-level. China's multiple, evolving surveillance systems for schistosomiasis have been integral to the country's decades-long campaign to reduce the prevalence of the disease. Chinese surveillance for schistosomiasis began in the middle of the 20th century after the founding of the People's Republic, with identification of endemic areas through both hospital-based reporting and limited epidemiological surveys. Currently, there are four parallel surveillance systems specific to schistosomiasis (Table 2 )— routine surveillance, national surveys, sentinel surveillance and the national infectious disease reporting system ( NIDRS ). These systems serve distinct, but complementary purposes. The routine surveillance system (termed 'routine surveys' in some literature), was initiated in the 1950s. The other three schistosomiasis surveillance systems created in the late 1980s were strengthened considerably in 1992 when the Chinese government received a long-term World Bank loan aimed at improving schistosomiasis morbidity control. The loan resulted in significant investment in schistosomiasis research [ 8 ] and disease control efforts, as well as financial support for the national and routine surveillance systems [ 8 , 9 ]. Table 2 Summary of key characteristics of current schistosomiasis surveillance systems in China NIDRS Sentinel Routine National Active/passive Passive Active Active Active Year initiated 1950s (1989 for schistosomiasis) 1990 1950s 1989 Reporting unit Individuals in hospitals Sentinel village Village Village Coverage All hospitals Nine sentinel villages All villages in endemic counties 1% of villages in endemic provinces Time frame Real-time, within 24 hours of patient diagnosis Yearly All villages sampled over ~3 years, reporting occurs at completion of each village's survey Periodic every 6-9 years: 1989, 1995, 2004 Purpose Aid understanding of disease patterns; provide evidence for policy-making Longitudinally and objectively monitor how the schistosome-endemic situation changes over time Evaluate control measures Clarify the endemic status of schistosomiasis as established by the previous national survey Information collected Individual cases (demographics, patient residence, diagnosis, treatment and hospital) Snail habitat, human infection prevalence and intensity, bovine infection prevalence Snail habitat, human infection prevalence Human infection prevalence and intensity, bovine infection prevalence, snail habitat Diagnostics Clinical and laboratory IHA screen then Kato-Katz and miracidium hatch IHA screen then Kato-Katz ELISA screen then Kato-Katz Major changes 2004: Replaced paper-based monthly or yearly reporting with internet-based real-time reporting system 2011: Added Miracidium Hatch Test to diagnostic procedure 2011: Replaced yearly reporting with internet-based parasitic disease reporting that occurs after completion of each village survey 2004: Inclusion criteria expanded to include areas with prevalence >0.5% from previous criterion of >1% Strengths Inexpensive, Algorithms can be created to automatically detect outbreaks of emerging or reemerging disease Provides longitudinal measures of disease prevalence and intensity Provides greatest coverage since it samples all endemic villages in the province Provides a nationwide estimate of schistosomiasis prevalence Limitations Underreporting of chronic cases; potential underreporting of acute cases due to political pressure; potential information bias associated with variable clinical and diagnostic capacities of reporting sites; non-response bias associated with reporter fatigue Sampling occurs at limited sites (20 in 1989, 80 in 2005); longitudinal follow-up over decades can yield non-response bias resulting from participation and reporter fatigue; potential selection bias associated with the choice of villages to sample longitudinally Variable clinical and diagnostic capabilities can lead to information bias; potential selection bias associated with choice of survey sites, since those in charge of surveillance are also in charge of control efforts; potential reporting bias as funding can be tied to disease control success; potential non-response bias resulting from participation fatigue and temporary rural-to-urban migration Occurs rarely and survey methods change, making it difficult to assess temporal patterns; only includes 1% of endemic villages in each province IHA: indirect hemagglutination assay; Village: administrative village, with typical population of ~1000 people. Schistosomiasis, like many neglected tropical diseases and other human helminthiases in particular, presents several surveillance challenges. First, the disease is typically concentrated in low-income, rural areas where health infrastructure is limited. Second, the clinical presentation of the disease is rarely acute, and like many chronic diseases, long-term infections can evade clinical detection and eventually lead to severe health sequelae later in life. Thus reliance on passive hospital or clinic-based reporting can grossly underestimate the number of infections. China's long history conducting schistosomiasis surveillance through multiple systems presents a unique opportunity to examine how their complementary structures, designs and sampling approaches—and their evolution over time—have provided essential epidemiological information to support the control of a neglected tropical disease. The goal of this paper is to detail how China's schistosomiasis surveillance systems have been structured and restructured as part of China's evolving public health system, yielding a combined information system that has been integral to the country's progress in reducing the burden of the disease. We discuss the diagnostic approaches underlying these systems, as well as additional tools necessary for cost-effective schistosomiasis surveillance in endemic settings that are approaching transmission control and elimination but face the threat of re-emergence. Finally, we discuss ways in which China's experience with schistosomiasis surveillance can offer insights to other countries as they develop and strengthen their own infectious disease surveillance systems. Diagnostic methods for detecting schistosomiasis infections in humans and animals China's schistosomiasis surveillance programs target three key hosts: humans, bovines and snails. Assays to detect schistosomiasis infections in human hosts are used to provide a direct measure of disease burden in the population, whereas surveys of bovines and snails (spatial distribution and infections) provide a means to evaluate the transmission potential of a given setting, and the particular role played by non-human reservoirs. Detection of human infections Detection of schistosomiasis infections in humans in China is typically accomplished by immunologic or coprologic assays. Immunologic assays commonly used in China include Schistosoma japonicum -specific immunoglobulin G standardized enzyme-linked immunosorbent assay (ELISA) and the indirect hemagglutination assay (IHA). Diagnostic immunological test kits have historically been subsidized by the national government, which has contributed to their widespread use. From a practical standpoint, immunologic methods are desirable as they require only a blood sample from a finger or ear-stick. Immunoassays generally have high sensitivity, but low specificity, due in part to the inability of the assay to distinguish between past and current infections [ 10 ]. The immunological tests are generally used as a first-step screening method in large population surveys: those who test positive by immunoassay are then tested with a coprologic exam. China relies on the Kato-Katz technique as the primary coprologic schistosomiasis diagnostic, following guidance of the World Health Organization [ 11 ]. The method involves the microscopic examination of a stool sample smeared across three slides using a light microscope. Slides are examined by trained laboratory technicians who count Schistosoma eggs in standardized 41.7 mg smears and calculate eggs per gram (EPG) of the original sample, providing a widely accepted proxy measure of infection intensity [ 12 ]. The assay lacks sensitivity in low infection-intensity regions [ 13 , 14 ]. Another common coprologic assay is the miracidium hatch test, which detects the presence of viable schistosome eggs in stool by inducing miracidia to hatch in an aqueous solution. Typically performed in county-level anti-schistosomiasis stations, 30 g stool samples are strained through a 200 μm nylon mesh to concentrate schistosome eggs and remove small particles. The enmeshed material is then suspended in an aqueous solution, and placed in a lit, temperature-controlled environment where the sample is examined at two, five and eight hours for the presence of miracidia using low magnification [ 14 ]. Detection of bovine infections Detection of S. japonicum in the bovine population serves as a complement to human surveillance, as bovines are a key non-human reservoir of S. japonicum in China. Thus detection of infections in bovines offers a means of monitoring the role of non-human hosts in introducing or sustaining transmission, particularly in agricultural settings where bovines are abundant [ 14 ]. To test bovines for infection, a variant of the miracidium hatch test is used with modified observation times. Because the high water content of bovine stool can promote rapid hatching, the suspended sample is typically examined at shorter time intervals (one, two, and four hours) after preparation relative to the human exam [ 15 ]. Quantification of intermediate host densities and infections Assessments of the density and infection status of the intermediate snail host Oncomelania hupensis have long played an important role in China's surveillance for schistosomiasis. Snail density surveys consist of collecting and counting snails from historical or newly identified snail habitat. Specimens are drawn from a consistent kuang -sized (0.11 m 2 ) sampling frame at random or equal interval points along streams, irrigation canals, fields, and lakes in endemic and some formerly endemic areas [ 7 ]. The infection status of sampled specimens is determined by microscopic examination for the presence of cercariae. This method is limited in that it is only capable of detecting patent infections, and even in regions where human infection prevalence is high, the prevalence of S. japonicum infections in snails is very low (<<1%; [ 16 ]), and thus screening of a large number of snail hosts is needed. In areas where human infection prevalence is low, infected snails are rare and a poor indicator of the presence of human infections [ 14 ]. Future diagnostic needs As schistosomiasis infection prevalence declines in China, more sensitive diagnostic methods are needed [ 14 ]. Recent diagnostic innovations include methods that detect S. japonicum DNA in human and bovine stool or serum, as well as in snails [ 17 – 19 ], but these assays have not yet been widely adopted. Such approaches offer potentially greater sensitivity than existing methods, particularly in areas where infection intensities are low, which is a key technological need for schistosomiasis elimination [ 20 ]. Issues of cost, standardization, laboratory infrastructure and technical training remain to be addressed for this new class of molecular diagnostic techniques. History of institutional support for schistosomiasis surveillance in China In 1956, a massive nationwide schistosomiasis control campaign was launched, which bore the slogan "Schistosomiasis has to be eliminated" and was subsequently memorialized by Chairman Mao Zedong's poem "Farewell to the God of Plague" in which he mourned the advance of the disease and championed its subsequent retreat [ 21 , 22 ]. Medical school graduates were required to spend time in rural areas, both to provide care directly as village doctors, as well as to train local residents to serve as paramedics, or so-called barefoot doctors [ 23 , 24 ]. At the same time, China established administrative and professional health organizations at varying levels of government to carry out both surveillance and control tasks aimed at eliminating specific diseases. These included an Office of Endemic Disease (OED) in each province, which designed and managed surveillance and control programs, as well as units targeting specific infections. For schistosomiasis, county-level anti-schistosomiasis stations were established under the supervision of OEDs and were responsible for carrying out surveys on humans and the snail intermediate host, control activities and health education. These stations established the necessary infrastructure and recording procedures for documenting infection and treatment statuses, disease progression of patients, and distribution and abundance of the intermediate host of the pathogen, and hence played a central role in the first standardized schistosomiasis surveillance activities [ 7 ]. In the late 1970s, reforms to China's health systems led to the dismantling of the cooperative health system in place of a market-based system, and by 1985, barefoot doctors either became village doctors who worked on a fee-for-service basis, or they shifted to other professions [ 25 ]. This substantial change in government policy led to a dramatic decline in primary health care coverage in rural areas. The restructuring of the public health system also involved the establishment of Epidemic Prevention Stations (EPS) at provincial, prefecture and county administrative levels, with a vertical structure for disease reporting and for the provision of technical guidance [ 26 ]. Provincial, prefecture, and county EPS were financed by the health bureau at the same administrative level, resulting in little administrative control of higher-level EPS over lower-level institutions (although technical guidance did flow from higher- to lower-level units). Prefecture and county EPS primarily reported to their local health bureau instead of provincial EPS or the Ministry of Health. This structure would change dramatically with the most recent major overhaul of China's heath systems at the turn of the 21 st century [ 26 , 27 ]. In 2002, the National Center for Disease Control and Prevention and Control (China CDC) was established from the former Chinese Academy of Preventive Medicine [ 27 ]. With the founding of China CDC, nearly all institutions responsible for disease control were integrated into a single system that linked national, provincial, prefecture and county levels. All levels of EPS were converted to CDCs and linked with other CDCs in a vertical structure that persists to this day (Figure 1 ; [ 27 ]). Figure 1 Structure of China's Public Health System after 2002 [ 44 ] . China's evolving schistosomiasis surveillance systems Surveillance was spotty and inconsistent in the years following the founding of People's Republic of China, and not until 1989 did robust schistosomiasis surveillance programs emerge [ 28 ]. A massive, national survey in 1989 confirmed endemic counties and assessed schistosomiasis prevalence status in 12 provinces based on human infections and an assessment of intermediate host habitat [ 29 ]. Starting in 1992, additional schistosomiasis surveillance systems came online, including annual, locally administered surveys in endemic villages, as well as intensive national surveys to estimate the prevalence of schistosomiasis across provinces and in several high priority areas [ 30 ]. In 2003, the SARS epidemic exposed key weaknesses in China's ability to detect and manage infectious disease epidemics [ 31 ]. In response, the Chinese government invested heavily in an enhanced National Infectious Disease Reporting system (NIDRS), which launched in 2004. The overhauled NIDRS was fully internet-based, and the number of reportable diseases was increased to reflect the country's need for key surveillance data to support infectious disease prevention and control programs. At the same time, China's entire public health system was being streamlined as described above, leading to the formation of China CDC [ 32 ]. Coincident with this massive transformation of China's public health institutions, a large World Bank loan that supported schistosomiasis control programs was ending, and its completion in 2001 resulted in a significant loss of funding for schistosomiasis surveillance. Surveillance and control efforts decreased, contributing to the re-emergence of schistosomiasis in several locations where it had previously been controlled [ 7 , 33 ]. In response, the Chinese Ministry of Health included schistosomiasis on a list of four infectious diseases targeted with high priority for surveillance and control, along with tuberculosis, HIV/AIDS and Hepatitis B. This was a major policy change, establishing schistosomiasis control as a national priority and elevating the position of the disease on the list of nationally notifiable diseases (i.e., from class C to class B; Table 3 ). Improvements in schistosomiasis surveillance followed, as increased commitment and funding flowed from the central government. Table 3 Infectious diseases covered by mandatory NIDRS reporting [ 31 ] Rank # of diseases Reporting time Diseases Class A 2 2 hours Plague, Cholera Class B 26 24 hours § SARS § , HIV/AIDS, Viral hepatitis (A, B, C, E, other), Poliomyelitis § , Human avian influenza § , Measles, Epidemic hemorrhagic fever, Rabies, Epidemic Japanese encephalitis B, Dengue fever, Anthrax § , Tuberculosis, Dysentery (viral or amebic), Epidemic cerebrospinal meningitis, Typhoid and Paratyphoid, Pertussis, Diphtheria, Tetanus neonatorum, Scarlet fever, Brucellosis, Gonorrhea, Syphilis, Leptospirosis, Schistosomiasis, Malaria, H1N1 swine flu Class C 11 24 hours Influenza, Mumps, Rubella, Acute hemorrhagic conjunctivitis, Leprosy, Typhus, Leishmaniasis, Echinococcosis, Filariasis, Infectious Diarrhea other than cholera, dysentery or typhoid and paratyphoid, Foot-and-mouth disease § Cases of SARS, poliomyelitis, pulmonary anthrax and human infection with highly pathogenic avian influenza must be reported within 2 hours. In the following sections, we describe the purpose, sampling approach, and diagnostics used in each of the four schistosomiasis surveillance systems— routine surveys , national surveys, sentinel surveys and NIDRS —and we describe how each surveillance system was impacted by changes in China's health systems following the SARS epidemic in 2004. Routine surveys Routine schistosomiasis surveys involve periodic infection screening in every endemic village in the seven endemic provinces. Staff at county-level anti-schistosomiasis stations have historically conducted these screenings in collaboration with local village governments and hospitals [ 7 ]. Routine surveys generate data used by the Ministry of Health to evaluate control measures, and provide information used by provincial-level schistosomiasis control steering committees to determine which counties have satisfied the official criteria for transmission control (Table 1 ). The primary outcomes and other characteristics of routine surveys are summarized in Table 2 . Changes since 2004 With the substantial increase in central government support for schistosomiasis control since 2004 [ 34 ], local anti-schistosomiasis stations began receiving funding directly from the central government to carry out and improve routine surveys, and to implement interventions based on survey findings. Because anti-schistosomiasis stations are responsible for both surveillance and control, and the funding they receive from the national government is in part correlated to control success, reporting bias may be present in routine surveillance data. The national and provincial health bureaus set annual goals (e.g., prevalence reduction targets) for anti-schistosomiasis stations, which were folded into annual control performance reviews for the stations. To improve reliability of these surveys, a real-time, internet-based reporting system was established in 2011 to file, store and manage routine survey data [ 35 ], and quality control checks are routinely performed on these electronic records to detect and reduce reporting biases. Sampling methodology Routine surveys in a given endemic county are typically conducted over the course of two, three or more years depending on the level of endemicity in—and the control status of—the county. For counties that have achieved transmission control (Table 1 ), all villages are surveyed over a three-year period (i.e., one-third of villages in the county are surveyed each year). In counties that have achieved transmission interruption, all villages are surveyed over four- or five-year periods, and in counties that have yet to achieve transmission control, all villages are surveyed over two years. Villages are randomly selected and grouped into two (for counties not yet achieving transmission control), three (for counties achieving transmission control), or four or five (for counties achieving transmission interruption) subgroups. Each year, one subgroup of villages is surveyed, and this is repeated annually until all subgroups are surveyed. At that point, villages are randomly grouped into new subgroups, and the multi-year survey process is repeated. In villages that are part of the sentinel schistosomiasis surveillance system (see below), routine surveys are not conducted and the necessary information is shared from the sentinel system. In a given village, routine surveillance involves intermediate host surveys in the spring followed by surveys of the human population in the fall, at the end of the transmission season. The target human sample is >90% of residents between the ages of 6 and 65 years old [ 9 ]. Sampling of cattle and other alternative mammalian hosts are also included in routine survey efforts, but these are the responsibility of the local animal husbandry department and the degree of coordination between the veterinary and public health efforts is highly variable [ 7 ]. All residents are requested to present themselves at a designated location for screening on the day of the survey. Villager participation in these surveys has been inconsistent. Residents may not feel ill and therefore may see no need to be tested for schistosomiasis. In endemic areas, surveys have been completed every year or every other year for >50 years, leading to participation fatigue. To overcome this challenge, health workers have conducted health education campaigns that emphasize the impacts of undetected, chronic infections, and the importance of routine schistosomiasis examinations. When residents are missed by a routine survey screening, public health workers will seek them out in their homes and request their participation. Sampling and screening protocols for routine surveys are standardized by the Ministry of Health to improve consistency between counties and provinces, though variations have been noted. In Sichuan Province, for example, different diagnostics have been used between counties, some carrying out serologic exams only, some serologic followed by stool exams and some stool exams only. Thus, there are limitations when directly comparing routine survey data between counties. Relationship to routine surveillance systems for other parasites and diseases In China, routine surveys are unique to schistosomiasis, providing spatial and temporal data on infection patterns that are generally unavailable for other parasitic diseases. The surveys are very resource-intensive, and thus they have not been adopted, for instance, for soil-transmitted helminths (STHs). A sentinel STH system exists and resembles routine schistosomiasis surveys with less exhaustive sampling. The standardized training of county-level staff that carry out routine survey screenings—and the infrastructure developed to support them—have provided ongoing quantitative data on the distribution of cases at fine spatial scales [ 36 ]. These data are unique to schistosomiasis, and have provided essential information to schistosomiasis reduction programs throughout China. National surveys In 1989, China carried out the first national schistosomiasis survey with the purpose of estimating the prevalence of the disease in each endemic province. The resulting information served as an important baseline dataset for the World Bank loan for schistosomiasis control (see Introduction), and the 1989 data were subsequently compared with the 1995 national survey in an economic valuation of the loan-financed programs [ 37 ]. National surveys are large, periodic cross-sectional studies that generate estimates of prevalence among humans and domestic animals across endemic settings. They are carried out every 6–9 years in China, and their primary outcomes and other characteristics are summarized in Table 2 . Changes since 2004 The most recent national schistosomiasis survey in 2004 included considerable technical improvements over previous surveys, such as an additional sampling stratum targeting low intensity infection areas; use of ultrasonography to screen a small number of subjects tested for schistosome-induced fibrosis; use of a subset of surveyed villages to assess the sensitivity of the Kato-Katz test as performed by Chinese technicians [ 28 ]; and sampling to estimate snail density and snail infection prevalence in one county within each province [ 37 ]. Importantly, the 2004 national survey included all formerly endemic areas that had achieved infection control and transmission control (see Table 1 ), whereas previous surveys focused on formerly endemic areas that had achieved infection control only. These and other changes to the national survey sampling design in 2004 provided important information as China oriented its control programs towards national elimination, but also generated survey results that were not directly comparable to the 1989 or 1995 surveys [ 28 ]. Sampling methodology National surveys are mandated by the Chinese Ministry of Health, but are designed, managed and overseen by the OEDs in the seven endemic provinces [ 28 ]. The sampling unit is endemic administrative villages, as defined by a combination of data from the preceding national survey and preceding years of routine survey data. The sampling design is based on a stratified cluster random sampling across three strata: (1) province, (2) environment/ecotype, and (3) estimated local prevalence based on the most recent routine survey data (Figure 2 ). All residents 6 to 65 years old are eligible to participate in national surveys, and the ELISA method is used for initial screening, followed by Kato-Katz exams to confirm and quantify infection for all ELISA positives [ 28 ]. Figure 2 Sampling design for the 1989 and 1995 national schistosomiasis surveys [ 36 ] . A stratified cluster random sampling design with three strata was used. The first sampling stratum includes the seven schistosomiasis-endemic provinces (Jiangsu, Anhui, Jiangxi, Hubei, Hunan, Sichuan, and Yunnan) and one controlled province (Zhejiang). For endemic provinces, environmental and ecosystem characteristics were used to define the first sub-stratum (eco-type), and the level of prevalence defined the second sub-stratum (prevalence). A target survey population was drawn from each of the second sub-strata, with the characteristics defined in the figure. For areas that have achieved control (Zhejiang province in the 1989 and 1995 national surveys), the target survey population was drawn from two administrative villages from each historically endemic county. Because national surveys use standardized sampling methodologies, the results are comparable between provinces. What is more, the Ministry of Health uses the national surveys' representative population sampling strategy to estimate prevalence across each endemic province while decreasing the required size of the sample. However, the sampled population, chosen in part based on historical prevalence, may not be representative [ 28 ]. Relationship to national surveys for other parasitic diseases The Chinese government has historically carried out national surveys for other human parasitic diseases [ 38 ]. Between 1988 and 1992, and again between 2001 and 2004, national surveys tested for 56 different parasites, including protozoa and helminths [ 28 , 39 ], using a random sampling scheme and a set of diagnostics similar to the national schistosomiasis surveys [ 40 ]. However, sampling for these national surveys was not informed by routine surveillance data. Thus, in the absence of recent prevalence data, socioeconomic factors—including variables such as sanitation access, income and literacy—were used to stratify the population for sampling. Ad hoc provincial surveys following national survey protocols As an example of how new surveillance capacity can follow on existing efforts, some provinces carry out provincial schistosomiasis sampling surveys borrowing sampling methodology from the 1989 and 1995 national surveys, but with the purpose of including a larger population within the province to boost sensitivity. Provincial OEDs design provincial survey sampling protocols, including human, bovine and intermediate host surveys, and surveys are carried out by anti-schistosomiasis stations in each county with the support of the EPS. Endemic villages from each county are sampled randomly within three categories based on infection prevalence and infection intensity determined from prior routine surveys, as follows. One third of endemic villages are randomly selected from each of three categories of infection prevalence based on results from routine surveys: less than five percent prevalence; between 5 and 15 percent; and >15 percent. In settings with low variability in infection prevalence between villages, sampling is based on infection intensities, with one third of villages sampled each from light, medium and heavy infection categories, again based on the results of recent routine surveys. In Sichuan's provincial survey in 2001, for instance, 1,188 villages were sampled, involving >1,810,000 people ranging from age 5 to 65. The survey took six months to complete, from April to October, and all cattle were screened in each sampled village. For comparison, for the national surveys in Sichuan, >52,000 people from 39 villages, and >51,000 people from 40 villages, were surveyed respectively in 1989 and 1995. The more intensive sampling regimen of the provincial survey provided more detailed spatial coverage than the national survey. National sentinel surveillance The sentinel schistosomiasis surveillance system was initiated in 1989 with 20 sites in the seven endemic provinces, selected to represent different eco-epidemiological zones of transmission [ 9 ]. The purpose of the sentinel system is to monitor longitudinal prevalence and infection intensity. Sentinel surveys consist of human and bovine infection screening—along with snail surveys—with the purpose of capturing temporal changes in prevalence and intensity of infection [ 9 ]. For example, in Sichuan Province there are four counties enrolled in sentinel surveillance—Xichang, Dangling, Guanghan and Pujiang—where snail sampling takes place two times a year, in late spring and in fall, and human and animal surveys are conducted in the fall [ 9 ]. Within each county, one village is selected each year where at least 90% of village residents over the age of six are sampled. Changes since 2004 The sentinel system expanded in 2005 to include 80 sites across China, reflecting the Chinese government's increased commitment and funding to schistosomiasis surveillance and control. In Sichuan Province, the number of sentinel sites increased from four to nine, and of these, six are in hilly regions, one is from a plateau region, and two are from mountainous areas. Hilly areas are represented with the most sentinel sites because infections in Sichuan are most common in these regions. The nine Sichuan sites have remained constant since 2005, and there is a commitment to continue longitudinal tracking of schistosomiasis infections in these sites regardless of their transmission trajectories in order to take maximum advantage of a consistent longitudinal dataset. Diagnostic protocols for the sentinel system have changed to accommodate the changing transmission profiles of sentinel sites, particularly challenges in detecting low-intensity infections. As of 2011, for instance, sentinel sites in Sichuan Province test all those classified as positive using a serologic assay with both Kato-Katz and a miracidium hatch test. These multiple rounds of diagnoses, and a history of intensive surveillance generally in these historically endemic regions, have led to community-wide surveillance fatigue in some sentinel sites. Costs of implementing the ongoing sentinel surveillance are likely to increase as it becomes more challenging to achieve high levels of community participation, necessitating more intensive follow-up with residents and more sensitive diagnostic methods. Sampling methodology The number of sentinel sites sustained by the system is based on the level of funding provided by the central government, and provinces with higher prevalence and infection intensity are allotted more resources (and thus sites). To select sites, national and provincial level OEDs consider two factors: variation in infection intensity (sites are selected from communities with heavy, medium and light infections based on infection intensity measurements from prior routine surveillance); and eco-epidemiological zone (e.g., plateau, lake, hilly or mountainous regions). Sentinel system screening has been conducted by Kato-Katz since 1989, and the IHA test was added in 2000, with patients found positive by IHA confirmed with the Kato-Katz exam. If acute or advanced cases are found, epidemiological investigations are conducted to collect more detailed information about each case. The overall design of the sentinel system was developed and approved by a national steering committee, yielding a system that is consistent among all surveillance sites, and making the data comparable between sites in all seven endemic provinces. It is sometimes necessary to move sites when implementing an intensive longitudinal surveillance system. For instance, after >10 years at one site in Sichuan, the population exhibited fatigue from yearly surveys and a new site was substituted to increase community participation. In one instance, as infection intensity and prevalence decreased dramatically across Sichuan Province, a very low prevalence sentinel site was replaced with a high prevalence site. Substituting sites interrupts the sentinel surveillance time-series, and data at new sites provide few insights into temporal trends until the series is allowed to accumulate over a number of years. National infectious disease reporting system The National infectious disease reporting system ( NIDRS ) was established in the 1950s [ 31 ], but it was not until 1989, with the passing of China's Law on Preventing and Treating Infectious Diseases, that the reporting of selected infectious diseases was mandated by law [ 32 ]. The NIDRS has historically involved a vertical reporting structure where hospitals reported the aggregated number of cases monthly by post to the county health department, who in turn report aggregate cases to the prefecture, then to the province, and finally to the central government. The information is used by Chinese public health officials to understand national patterns of infectious disease transmission, and to develop appropriate prevention and control programs [ 41 ]. Changes since 2004 Following the SARS outbreak, China's Law on Preventing and Treating Infectious Diseases was revised, and the NIDRS was transformed into an internet-based, real-time system [ 31 ]. The new system involves the routing of individual cases (rather than strictly aggregated data) up the hierarchy of public health institutions. The system provides for timely reporting of schistosomiasis cases to public health officials, allowing for rapid identification of, and response to, suspected re-emergence of the disease. The reporting of individual cases also enhances the spatial and temporal resolution of disease reporting, allowing for assessment of, for example, seasonal trends and persistent hot-spots of transmission [ 31 ]. Sampling methodology Cases of schistosomiasis diagnosed in clinics or hospitals are reported directly to the NIDRS . Reporting includes both acute and chronic cases, however acute schistosomiasis contributes a disproportionate number of reported schistosomiasis cases. Acute schistosomiasis is a severe allergic reaction to the migrating schistosomule in the blood stream following infection and typically occurs only in naïve populations such as children or military personnel deployed to a new region. The rapid onset of severe symptoms typical of acute schistosomiasis is more likely to prompt health-seeking behavior, therefore resulting in more frequent capture by NIDRS [ 42 ]. Because the symptoms of chronic S. japonicum infection are non-specific and often sub-clinical, chronic infections are poorly captured by clinic-based surveillance. Thus NIDRS schistosomiasis surveillance is considered key for identifying areas of disease emergence (or re-emergence), but not for assessing infection prevalence. What is more, diagnosis of a reported case is coded in the NIDRS as using either a clinical or a laboratory diagnostic, without further differentiation such as between stool and serologic exams [ 43 ]. The primary outcomes and other characteristics of the NIDRS system are listed in Table 2 . Schistosomiasis diagnostic and clinical capabilities, and therefore reporting capacity, differ considerably across China, and the transition to the internet-based NIDRS revealed several limitations of surveillance capacity in China [ 31 ]. The availability of computing resources and internet coverage required to connect directly to NIDRS varied within and between provinces, limiting the generation and transmission of electronic case reports [ 44 ]. Hospitals that lack sufficient information technology resources were required to send paper forms to the county CDC for entry into the online system, introducing a reporting delay. A national initiative was launched to equip such hospitals with computer workstations and network coverage, resulting in the delivery of >4,000 computers to hospitals and clinics in Sichuan Province in 2005, for example [ 45 ]. Other challenges were revealed relating to disparities in clinical resources. For instance, a disproportionate number of the cases of schistosomiasis present in NIDRS have been reported by county-level hospitals, in part because diagnostic capabilities are poorer at township-level hospitals, and diagnostic capabilities for parasitic diseases are sometimes entirely absent at private clinics and village healthcare centers [ 46 ]. Relationship to reporting systems for other diseases At present, the NIDRS covers 39 infectious diseases in three classes based on their public health importance (Table 3 ; [ 47 ]). The Class A list includes two diseases, cholera and plague, that are required to be reported within two hours of diagnosis. Pulmonary anthrax, SARS, poliomyelitis, and human infection with highly pathogenic avian influenza are in Class B, and are also required to be reported within two hours. Other Class B and Class C diseases must be reported within 24 hours of diagnosis. Class C diseases include newly identified diseases or diseases with emerging public health importance, including foot and mouth disease. Schistosomiasis was listed as a Class B nationally reportable infectious disease in 2004 [ 47 ], in part to rapidly capture reports of acute cases that may signal an outbreak of infection [ 43 ]. Future directions in surveillance Ideally, surveillance systems provide data that support evidence-based policy decisions and contribute to effective and efficient disease control [ 48 , 49 ]. Additionally, effective surveillance systems must respond to advances in technology and changes in programmatic goals. While multiple, complementary systems can provide rich spatial and temporal information on the distribution of cases, resource constraints necessitate surveillance systems that minimize redundancies, particularly in low- and middle-income countries. Key factors in the evolution of China's schistosomiasis surveillance over the past six decades include the World Bank loan (and the rise in cases following its termination in 2001); changes in the epidemiology of schistosomiasis; changes in available technologies for diagnosis and treatment [ 50 ]; and major public health events such as the 2003 SARS epidemic. The SARS epidemic in particular revealed weaknesses in China's public health system with respect to the detection of, and prompt response to, disease outbreaks [ 31 ]. Following the epidemic, the country's infectious disease surveillance infrastructure underwent a massive overhaul, which included construction of new infectious disease facilities and laboratories that were crucial to improving the quality of surveillance data, as well as facilitation of improved data sharing and of more efficient coordination between public health agencies. The overhaul also included a complete redesign of the real-time infectious disease reporting system [ 51 ], as well as the Ministry of Health's listing of schistosomiasis among diseases such as HIV/AIDS, tuberculosis and hepatitis B as a high priority for control [ 25 ]. This classification raised the profile of the disease among public health officials, increased the commitment of health care workers and government agencies to schistosomiasis elimination, and stimulated heightened surveillance and control efforts [ 43 , 52 ]. Throughout these major transitions, China has made remarkable progress towards controlling schistosomiasis. The country's evolving, parallel surveillance systems have each played a distinct role in characterizing the dynamic state of schistosomiasis transmission across diverse regions of the country. Thus far, the systems and the information they provide have been used independently; numerous reports, based on routine surveys, sentinel surveys and national surveys have been published and used to develop control strategies and monitor progress towards achieving elimination. More recently, there has been an effort to expand the use of these data beyond endemic disease control. Routine surveillance data, for instance, has been used to provide evidence of re-emerging schistosomiasis in Sichuan Province in areas that had previously achieved local control [ 7 ]. Surveillance data from multiple systems can be fruitfully combined, as evidenced by integration of national and NIDRS surveillance data from all provinces to estimate the incidence, prevalence and disease burden of schistosomiasis, as well as malaria, hookworm and other water, sanitation, and hygiene-associated infections [ 53 ]. The analysis revealed large regional disparities in the burden of these infections, particularly among children, who experience the greatest risk of water, sanitation and hygiene-attributable disease in China. In a separate study, combined national and NIDRS data were used to estimate the future burden of these infections under climate change [ 54 ]. While this work demonstrates the value of China's parallel surveillance systems, understanding of how data from these systems can be effectively combined, to characterize and inform the many dimensions of disease transmission and control, remains limited. The overlapping nature of these systems, both spatially and temporally, provides a unique opportunity to explore how a country can maximize the value of multiple, distinctive information sources to inform public health campaigns and investments. This in an area where future research is needed both in China and elsewhere. Furthermore, as many provinces move towards schistosomiasis elimination and case detection becomes ever more challenging and expensive, it will be even more important to determine how the data, sampling approaches, and detection methods these systems provide can be used to optimize the performance of surveillance efforts. Successful schistosomiasis elimination will require the ability to efficiently and accurately identify the few remaining infected individuals and non-human reservoirs among large uninfected populations. This will be complicated by decreased diagnostic sensitivity owing to reduced infection intensity, testing fatigue among residents as a result of long-standing repeated examinations, and ever-increasing health care costs associated with diagnosis. To be successful, China's surveillance system must be prepared to modify sampling methods, reporting approaches and diagnostics in response to new scientific and technological advances and changing schistosomiasis infection patterns. Recent attempts to establish diagnostic standards through the systematic evaluation of an array of immunological assays [ 55 ] and new mobile reporting methods, discussed below, are a promising step in this direction. Further, as China's national disease burden continues its shift from infectious diseases to chronic, non-communicable diseases (NCDs), additional strain will be placed on surveillance personnel and institutions currently devoted to infectious disease, and opportunities for integrating the surveillance and care of NCDs and infectious diseases will be needed [ 56 , 57 ]. Care must be taken to ensure that the resources are adequate and to avoid surveillance fatigue among healthcare workers and those in sentinel villages. For China, research and innovations needed to succeed in continuing control and elimination of schistosomiasis include the development of diagnostic tools with higher sensitivity for the detection of human infections, and new ways of assessing human susceptibility to infection. In addition, effective ways to deploy recently developed tools for detecting the infectious stage of schistosomes (cercariae) in the natural environment [ 30 , 58 , 59 ] will be needed. Likewise, mobile disease reporting technologies (e.g. use of mobile phones or positioning systems with wireless capability for spatially explicit reporting of cases or vectors) show considerable promise and have been recently applied for surveillance of other neglected tropical diseases [ 60 ]. These technologies have been used in field surveys of schistosomiasis endemic areas, to file reports of acute cases and infected snails into a centralized surveillance database [ 9 ], and to coordinate responses to disease outbreaks after the 2008 earthquake in Sichuan province [ 16 , 61 ]. Yet effective integration of mobile technologies into large-scale surveillance procedures is still relatively new, and will need further evaluation. At the same time, new approaches will be needed to estimate the cost-effectiveness of the combination of existing surveillance systems in achieving specific surveillance endpoints, such as a characterization of spatial and temporal variability in cases. While considerable effort has been devoted to evaluating surveillance systems from a functional standpoint [ 62 – 66 ] —assessing timeliness [ 67 ], record accuracy [ 68 , 69 ] and enhancing operational efficiency [ 70 – 72 ] —few analytical tools exist for quantitatively assessing how changes to design translate into improved (or inferior) estimates of the underlying rates of disease taking into account system sensitivity, coverage, bias and other factors. Development of such tools in China would provide a useful precedent as schistosomiasis surveillance is increasingly brought online in other countries. Global lessons Globally, there is a major movement underway to reduce the burden of helminthiases. The priorities for the global control and elimination of schistosomiasis, along with STHs, were articulated by the WHO in 2013 [ 73 ]. The surveillance needs of these global efforts are immense. Nearly 240 million people, most of whom live in sub-Saharan Africa, are estimated to require preventive chemotherapy for schistosomiasis alone as of 2009, yet coverage is reaching only 8% of those in need [ 74 , 75 ]. While many factors contribute to this deficit, the lack of surveillance capabilities in nearly all schistosomiasis-endemic countries is a major obstacle to efficiently expanding coverage [ 74 ]. Accurate, timely, country-level data are generally unavailable, and targeting drug delivery for schistosomiasis control in these settings is predominantly based on risk maps generated using ad-hoc epidemiological data (e.g., research data, historical drug delivery data, etc.), rather than surveillance [ 76 ]. China's history with surveillance of schistosomiasis can serve as a guide for countries, in sub-Saharan Africa and elsewhere, that require low-cost, simple surveillance and diagnostic technologies. Given the current efforts to roll-out preventative chemotherapy, there will be a need to identify areas where disease burden is declining and mass-treatment efforts can be transitioned to targeted treatment and monitoring, as well as areas where disease burden is not declining, where more broad control efforts are needed. Throughout Africa, where over 90% of schistosomiasis cases occur [ 77 ], the capacity to conduct population-based surveys is limited [ 78 ], and current control recommendations do not include surveillance activities [ 74 ]. There is some movement towards developing the infrastructure for parallel active surveillance systems—sentinel systems to track progress of control strategies coupled with regular provincial surveys—in highly endemic areas. China's sentinel systems offer a model for deploying targeted surveillance resources and monitoring control progress over time. Similar to China's provincial surveys, which borrow from standardized national survey protocols, province-level schistosomiasis mapping of all endemic districts has recently begun in a few countries (e.g. Ethiopia), and is rapidly scaling up for other neglected tropical diseases such as trachoma, with the benefit of funds provided by the UK Department for International Development and The Bill & Melinda Gates Foundation [ 79 ]. These efforts are moving incrementally towards the electronic collection and reporting of epidemiological data, and may benefit from lessons learned during China's 2011 adoption of a simple, internet-based system to file, store and manage routine survey data. Crucially, this system provides a model for electronic data collection even in the absence of a sophisticated NIDRS capacity. As many low-resource countries aspire for control and even elimination of schistosomiasis, understanding the synergy of parallel systems will prove cost-effective both by reducing unnecessary redundancy in data collection and in maximizing programming dollars. The systems described in China have distinct utility, but how these systems may be translated to other areas of schistosomiasis endemicity is not yet known and must be explored. Looking forward, targeted surveillance capacity is needed to understand and respond to the transmission consequences of rapid environmental, social, and technological changes, such as the construction of dams and changes in agricultural practices, that can result in an upsurge in schistosomiasis cases [ 54 , 77 , 80 – 82 ]. China's sentinel system offers a model for deploying targeted surveillance capacity in regions undergoing such changes [ 83 ]. Optimal surveillance of schistosomiasis and other neglected tropical diseases must evolve with changing programmatic needs and technical advances. Currently, real-time electronic surveillance is limited by lack of broadband internet infrastructure in many rural schistosomiasis-endemic areas. However, mobile disease reporting technologies offer new opportunities in areas where traditional means of communication are limited [ 6 ], and may become increasingly relevant as attention turns to elimination. In the context of elimination, new diagnostic tools are urgently needed that offer rapid, sensitive, low-cost alternatives to current methods. China is currently dealing with this issue [ 14 ], and as other settings move towards achieving elimination, the traditional diagnostics that are currently relied upon by many surveillance systems will need to be replaced [ 84 ]. Recently developed methods offer improved opportunities for point-of-care surveillance (e.g., [ 85 ]), and future investments will be needed to ensure advances like these are pursued and, ultimately, incorporated into surveillance practice. Diagnostic methods for detecting schistosomiasis infections in humans and animals China's schistosomiasis surveillance programs target three key hosts: humans, bovines and snails. Assays to detect schistosomiasis infections in human hosts are used to provide a direct measure of disease burden in the population, whereas surveys of bovines and snails (spatial distribution and infections) provide a means to evaluate the transmission potential of a given setting, and the particular role played by non-human reservoirs. Detection of human infections Detection of schistosomiasis infections in humans in China is typically accomplished by immunologic or coprologic assays. Immunologic assays commonly used in China include Schistosoma japonicum -specific immunoglobulin G standardized enzyme-linked immunosorbent assay (ELISA) and the indirect hemagglutination assay (IHA). Diagnostic immunological test kits have historically been subsidized by the national government, which has contributed to their widespread use. From a practical standpoint, immunologic methods are desirable as they require only a blood sample from a finger or ear-stick. Immunoassays generally have high sensitivity, but low specificity, due in part to the inability of the assay to distinguish between past and current infections [ 10 ]. The immunological tests are generally used as a first-step screening method in large population surveys: those who test positive by immunoassay are then tested with a coprologic exam. China relies on the Kato-Katz technique as the primary coprologic schistosomiasis diagnostic, following guidance of the World Health Organization [ 11 ]. The method involves the microscopic examination of a stool sample smeared across three slides using a light microscope. Slides are examined by trained laboratory technicians who count Schistosoma eggs in standardized 41.7 mg smears and calculate eggs per gram (EPG) of the original sample, providing a widely accepted proxy measure of infection intensity [ 12 ]. The assay lacks sensitivity in low infection-intensity regions [ 13 , 14 ]. Another common coprologic assay is the miracidium hatch test, which detects the presence of viable schistosome eggs in stool by inducing miracidia to hatch in an aqueous solution. Typically performed in county-level anti-schistosomiasis stations, 30 g stool samples are strained through a 200 μm nylon mesh to concentrate schistosome eggs and remove small particles. The enmeshed material is then suspended in an aqueous solution, and placed in a lit, temperature-controlled environment where the sample is examined at two, five and eight hours for the presence of miracidia using low magnification [ 14 ]. Detection of bovine infections Detection of S. japonicum in the bovine population serves as a complement to human surveillance, as bovines are a key non-human reservoir of S. japonicum in China. Thus detection of infections in bovines offers a means of monitoring the role of non-human hosts in introducing or sustaining transmission, particularly in agricultural settings where bovines are abundant [ 14 ]. To test bovines for infection, a variant of the miracidium hatch test is used with modified observation times. Because the high water content of bovine stool can promote rapid hatching, the suspended sample is typically examined at shorter time intervals (one, two, and four hours) after preparation relative to the human exam [ 15 ]. Quantification of intermediate host densities and infections Assessments of the density and infection status of the intermediate snail host Oncomelania hupensis have long played an important role in China's surveillance for schistosomiasis. Snail density surveys consist of collecting and counting snails from historical or newly identified snail habitat. Specimens are drawn from a consistent kuang -sized (0.11 m 2 ) sampling frame at random or equal interval points along streams, irrigation canals, fields, and lakes in endemic and some formerly endemic areas [ 7 ]. The infection status of sampled specimens is determined by microscopic examination for the presence of cercariae. This method is limited in that it is only capable of detecting patent infections, and even in regions where human infection prevalence is high, the prevalence of S. japonicum infections in snails is very low (<<1%; [ 16 ]), and thus screening of a large number of snail hosts is needed. In areas where human infection prevalence is low, infected snails are rare and a poor indicator of the presence of human infections [ 14 ]. Future diagnostic needs As schistosomiasis infection prevalence declines in China, more sensitive diagnostic methods are needed [ 14 ]. Recent diagnostic innovations include methods that detect S. japonicum DNA in human and bovine stool or serum, as well as in snails [ 17 – 19 ], but these assays have not yet been widely adopted. Such approaches offer potentially greater sensitivity than existing methods, particularly in areas where infection intensities are low, which is a key technological need for schistosomiasis elimination [ 20 ]. Issues of cost, standardization, laboratory infrastructure and technical training remain to be addressed for this new class of molecular diagnostic techniques. History of institutional support for schistosomiasis surveillance in China In 1956, a massive nationwide schistosomiasis control campaign was launched, which bore the slogan "Schistosomiasis has to be eliminated" and was subsequently memorialized by Chairman Mao Zedong's poem "Farewell to the God of Plague" in which he mourned the advance of the disease and championed its subsequent retreat [ 21 , 22 ]. Medical school graduates were required to spend time in rural areas, both to provide care directly as village doctors, as well as to train local residents to serve as paramedics, or so-called barefoot doctors [ 23 , 24 ]. At the same time, China established administrative and professional health organizations at varying levels of government to carry out both surveillance and control tasks aimed at eliminating specific diseases. These included an Office of Endemic Disease (OED) in each province, which designed and managed surveillance and control programs, as well as units targeting specific infections. For schistosomiasis, county-level anti-schistosomiasis stations were established under the supervision of OEDs and were responsible for carrying out surveys on humans and the snail intermediate host, control activities and health education. These stations established the necessary infrastructure and recording procedures for documenting infection and treatment statuses, disease progression of patients, and distribution and abundance of the intermediate host of the pathogen, and hence played a central role in the first standardized schistosomiasis surveillance activities [ 7 ]. In the late 1970s, reforms to China's health systems led to the dismantling of the cooperative health system in place of a market-based system, and by 1985, barefoot doctors either became village doctors who worked on a fee-for-service basis, or they shifted to other professions [ 25 ]. This substantial change in government policy led to a dramatic decline in primary health care coverage in rural areas. The restructuring of the public health system also involved the establishment of Epidemic Prevention Stations (EPS) at provincial, prefecture and county administrative levels, with a vertical structure for disease reporting and for the provision of technical guidance [ 26 ]. Provincial, prefecture, and county EPS were financed by the health bureau at the same administrative level, resulting in little administrative control of higher-level EPS over lower-level institutions (although technical guidance did flow from higher- to lower-level units). Prefecture and county EPS primarily reported to their local health bureau instead of provincial EPS or the Ministry of Health. This structure would change dramatically with the most recent major overhaul of China's heath systems at the turn of the 21 st century [ 26 , 27 ]. In 2002, the National Center for Disease Control and Prevention and Control (China CDC) was established from the former Chinese Academy of Preventive Medicine [ 27 ]. With the founding of China CDC, nearly all institutions responsible for disease control were integrated into a single system that linked national, provincial, prefecture and county levels. All levels of EPS were converted to CDCs and linked with other CDCs in a vertical structure that persists to this day (Figure 1 ; [ 27 ]). Figure 1 Structure of China's Public Health System after 2002 [ 44 ] . China's evolving schistosomiasis surveillance systems Surveillance was spotty and inconsistent in the years following the founding of People's Republic of China, and not until 1989 did robust schistosomiasis surveillance programs emerge [ 28 ]. A massive, national survey in 1989 confirmed endemic counties and assessed schistosomiasis prevalence status in 12 provinces based on human infections and an assessment of intermediate host habitat [ 29 ]. Starting in 1992, additional schistosomiasis surveillance systems came online, including annual, locally administered surveys in endemic villages, as well as intensive national surveys to estimate the prevalence of schistosomiasis across provinces and in several high priority areas [ 30 ]. In 2003, the SARS epidemic exposed key weaknesses in China's ability to detect and manage infectious disease epidemics [ 31 ]. In response, the Chinese government invested heavily in an enhanced National Infectious Disease Reporting system (NIDRS), which launched in 2004. The overhauled NIDRS was fully internet-based, and the number of reportable diseases was increased to reflect the country's need for key surveillance data to support infectious disease prevention and control programs. At the same time, China's entire public health system was being streamlined as described above, leading to the formation of China CDC [ 32 ]. Coincident with this massive transformation of China's public health institutions, a large World Bank loan that supported schistosomiasis control programs was ending, and its completion in 2001 resulted in a significant loss of funding for schistosomiasis surveillance. Surveillance and control efforts decreased, contributing to the re-emergence of schistosomiasis in several locations where it had previously been controlled [ 7 , 33 ]. In response, the Chinese Ministry of Health included schistosomiasis on a list of four infectious diseases targeted with high priority for surveillance and control, along with tuberculosis, HIV/AIDS and Hepatitis B. This was a major policy change, establishing schistosomiasis control as a national priority and elevating the position of the disease on the list of nationally notifiable diseases (i.e., from class C to class B; Table 3 ). Improvements in schistosomiasis surveillance followed, as increased commitment and funding flowed from the central government. Table 3 Infectious diseases covered by mandatory NIDRS reporting [ 31 ] Rank # of diseases Reporting time Diseases Class A 2 2 hours Plague, Cholera Class B 26 24 hours § SARS § , HIV/AIDS, Viral hepatitis (A, B, C, E, other), Poliomyelitis § , Human avian influenza § , Measles, Epidemic hemorrhagic fever, Rabies, Epidemic Japanese encephalitis B, Dengue fever, Anthrax § , Tuberculosis, Dysentery (viral or amebic), Epidemic cerebrospinal meningitis, Typhoid and Paratyphoid, Pertussis, Diphtheria, Tetanus neonatorum, Scarlet fever, Brucellosis, Gonorrhea, Syphilis, Leptospirosis, Schistosomiasis, Malaria, H1N1 swine flu Class C 11 24 hours Influenza, Mumps, Rubella, Acute hemorrhagic conjunctivitis, Leprosy, Typhus, Leishmaniasis, Echinococcosis, Filariasis, Infectious Diarrhea other than cholera, dysentery or typhoid and paratyphoid, Foot-and-mouth disease § Cases of SARS, poliomyelitis, pulmonary anthrax and human infection with highly pathogenic avian influenza must be reported within 2 hours. In the following sections, we describe the purpose, sampling approach, and diagnostics used in each of the four schistosomiasis surveillance systems— routine surveys , national surveys, sentinel surveys and NIDRS —and we describe how each surveillance system was impacted by changes in China's health systems following the SARS epidemic in 2004. Routine surveys Routine schistosomiasis surveys involve periodic infection screening in every endemic village in the seven endemic provinces. Staff at county-level anti-schistosomiasis stations have historically conducted these screenings in collaboration with local village governments and hospitals [ 7 ]. Routine surveys generate data used by the Ministry of Health to evaluate control measures, and provide information used by provincial-level schistosomiasis control steering committees to determine which counties have satisfied the official criteria for transmission control (Table 1 ). The primary outcomes and other characteristics of routine surveys are summarized in Table 2 . Changes since 2004 With the substantial increase in central government support for schistosomiasis control since 2004 [ 34 ], local anti-schistosomiasis stations began receiving funding directly from the central government to carry out and improve routine surveys, and to implement interventions based on survey findings. Because anti-schistosomiasis stations are responsible for both surveillance and control, and the funding they receive from the national government is in part correlated to control success, reporting bias may be present in routine surveillance data. The national and provincial health bureaus set annual goals (e.g., prevalence reduction targets) for anti-schistosomiasis stations, which were folded into annual control performance reviews for the stations. To improve reliability of these surveys, a real-time, internet-based reporting system was established in 2011 to file, store and manage routine survey data [ 35 ], and quality control checks are routinely performed on these electronic records to detect and reduce reporting biases. Sampling methodology Routine surveys in a given endemic county are typically conducted over the course of two, three or more years depending on the level of endemicity in—and the control status of—the county. For counties that have achieved transmission control (Table 1 ), all villages are surveyed over a three-year period (i.e., one-third of villages in the county are surveyed each year). In counties that have achieved transmission interruption, all villages are surveyed over four- or five-year periods, and in counties that have yet to achieve transmission control, all villages are surveyed over two years. Villages are randomly selected and grouped into two (for counties not yet achieving transmission control), three (for counties achieving transmission control), or four or five (for counties achieving transmission interruption) subgroups. Each year, one subgroup of villages is surveyed, and this is repeated annually until all subgroups are surveyed. At that point, villages are randomly grouped into new subgroups, and the multi-year survey process is repeated. In villages that are part of the sentinel schistosomiasis surveillance system (see below), routine surveys are not conducted and the necessary information is shared from the sentinel system. In a given village, routine surveillance involves intermediate host surveys in the spring followed by surveys of the human population in the fall, at the end of the transmission season. The target human sample is >90% of residents between the ages of 6 and 65 years old [ 9 ]. Sampling of cattle and other alternative mammalian hosts are also included in routine survey efforts, but these are the responsibility of the local animal husbandry department and the degree of coordination between the veterinary and public health efforts is highly variable [ 7 ]. All residents are requested to present themselves at a designated location for screening on the day of the survey. Villager participation in these surveys has been inconsistent. Residents may not feel ill and therefore may see no need to be tested for schistosomiasis. In endemic areas, surveys have been completed every year or every other year for >50 years, leading to participation fatigue. To overcome this challenge, health workers have conducted health education campaigns that emphasize the impacts of undetected, chronic infections, and the importance of routine schistosomiasis examinations. When residents are missed by a routine survey screening, public health workers will seek them out in their homes and request their participation. Sampling and screening protocols for routine surveys are standardized by the Ministry of Health to improve consistency between counties and provinces, though variations have been noted. In Sichuan Province, for example, different diagnostics have been used between counties, some carrying out serologic exams only, some serologic followed by stool exams and some stool exams only. Thus, there are limitations when directly comparing routine survey data between counties. Relationship to routine surveillance systems for other parasites and diseases In China, routine surveys are unique to schistosomiasis, providing spatial and temporal data on infection patterns that are generally unavailable for other parasitic diseases. The surveys are very resource-intensive, and thus they have not been adopted, for instance, for soil-transmitted helminths (STHs). A sentinel STH system exists and resembles routine schistosomiasis surveys with less exhaustive sampling. The standardized training of county-level staff that carry out routine survey screenings—and the infrastructure developed to support them—have provided ongoing quantitative data on the distribution of cases at fine spatial scales [ 36 ]. These data are unique to schistosomiasis, and have provided essential information to schistosomiasis reduction programs throughout China. National surveys In 1989, China carried out the first national schistosomiasis survey with the purpose of estimating the prevalence of the disease in each endemic province. The resulting information served as an important baseline dataset for the World Bank loan for schistosomiasis control (see Introduction), and the 1989 data were subsequently compared with the 1995 national survey in an economic valuation of the loan-financed programs [ 37 ]. National surveys are large, periodic cross-sectional studies that generate estimates of prevalence among humans and domestic animals across endemic settings. They are carried out every 6–9 years in China, and their primary outcomes and other characteristics are summarized in Table 2 . Changes since 2004 The most recent national schistosomiasis survey in 2004 included considerable technical improvements over previous surveys, such as an additional sampling stratum targeting low intensity infection areas; use of ultrasonography to screen a small number of subjects tested for schistosome-induced fibrosis; use of a subset of surveyed villages to assess the sensitivity of the Kato-Katz test as performed by Chinese technicians [ 28 ]; and sampling to estimate snail density and snail infection prevalence in one county within each province [ 37 ]. Importantly, the 2004 national survey included all formerly endemic areas that had achieved infection control and transmission control (see Table 1 ), whereas previous surveys focused on formerly endemic areas that had achieved infection control only. These and other changes to the national survey sampling design in 2004 provided important information as China oriented its control programs towards national elimination, but also generated survey results that were not directly comparable to the 1989 or 1995 surveys [ 28 ]. Sampling methodology National surveys are mandated by the Chinese Ministry of Health, but are designed, managed and overseen by the OEDs in the seven endemic provinces [ 28 ]. The sampling unit is endemic administrative villages, as defined by a combination of data from the preceding national survey and preceding years of routine survey data. The sampling design is based on a stratified cluster random sampling across three strata: (1) province, (2) environment/ecotype, and (3) estimated local prevalence based on the most recent routine survey data (Figure 2 ). All residents 6 to 65 years old are eligible to participate in national surveys, and the ELISA method is used for initial screening, followed by Kato-Katz exams to confirm and quantify infection for all ELISA positives [ 28 ]. Figure 2 Sampling design for the 1989 and 1995 national schistosomiasis surveys [ 36 ] . A stratified cluster random sampling design with three strata was used. The first sampling stratum includes the seven schistosomiasis-endemic provinces (Jiangsu, Anhui, Jiangxi, Hubei, Hunan, Sichuan, and Yunnan) and one controlled province (Zhejiang). For endemic provinces, environmental and ecosystem characteristics were used to define the first sub-stratum (eco-type), and the level of prevalence defined the second sub-stratum (prevalence). A target survey population was drawn from each of the second sub-strata, with the characteristics defined in the figure. For areas that have achieved control (Zhejiang province in the 1989 and 1995 national surveys), the target survey population was drawn from two administrative villages from each historically endemic county. Because national surveys use standardized sampling methodologies, the results are comparable between provinces. What is more, the Ministry of Health uses the national surveys' representative population sampling strategy to estimate prevalence across each endemic province while decreasing the required size of the sample. However, the sampled population, chosen in part based on historical prevalence, may not be representative [ 28 ]. Relationship to national surveys for other parasitic diseases The Chinese government has historically carried out national surveys for other human parasitic diseases [ 38 ]. Between 1988 and 1992, and again between 2001 and 2004, national surveys tested for 56 different parasites, including protozoa and helminths [ 28 , 39 ], using a random sampling scheme and a set of diagnostics similar to the national schistosomiasis surveys [ 40 ]. However, sampling for these national surveys was not informed by routine surveillance data. Thus, in the absence of recent prevalence data, socioeconomic factors—including variables such as sanitation access, income and literacy—were used to stratify the population for sampling. Ad hoc provincial surveys following national survey protocols As an example of how new surveillance capacity can follow on existing efforts, some provinces carry out provincial schistosomiasis sampling surveys borrowing sampling methodology from the 1989 and 1995 national surveys, but with the purpose of including a larger population within the province to boost sensitivity. Provincial OEDs design provincial survey sampling protocols, including human, bovine and intermediate host surveys, and surveys are carried out by anti-schistosomiasis stations in each county with the support of the EPS. Endemic villages from each county are sampled randomly within three categories based on infection prevalence and infection intensity determined from prior routine surveys, as follows. One third of endemic villages are randomly selected from each of three categories of infection prevalence based on results from routine surveys: less than five percent prevalence; between 5 and 15 percent; and >15 percent. In settings with low variability in infection prevalence between villages, sampling is based on infection intensities, with one third of villages sampled each from light, medium and heavy infection categories, again based on the results of recent routine surveys. In Sichuan's provincial survey in 2001, for instance, 1,188 villages were sampled, involving >1,810,000 people ranging from age 5 to 65. The survey took six months to complete, from April to October, and all cattle were screened in each sampled village. For comparison, for the national surveys in Sichuan, >52,000 people from 39 villages, and >51,000 people from 40 villages, were surveyed respectively in 1989 and 1995. The more intensive sampling regimen of the provincial survey provided more detailed spatial coverage than the national survey. National sentinel surveillance The sentinel schistosomiasis surveillance system was initiated in 1989 with 20 sites in the seven endemic provinces, selected to represent different eco-epidemiological zones of transmission [ 9 ]. The purpose of the sentinel system is to monitor longitudinal prevalence and infection intensity. Sentinel surveys consist of human and bovine infection screening—along with snail surveys—with the purpose of capturing temporal changes in prevalence and intensity of infection [ 9 ]. For example, in Sichuan Province there are four counties enrolled in sentinel surveillance—Xichang, Dangling, Guanghan and Pujiang—where snail sampling takes place two times a year, in late spring and in fall, and human and animal surveys are conducted in the fall [ 9 ]. Within each county, one village is selected each year where at least 90% of village residents over the age of six are sampled. Changes since 2004 The sentinel system expanded in 2005 to include 80 sites across China, reflecting the Chinese government's increased commitment and funding to schistosomiasis surveillance and control. In Sichuan Province, the number of sentinel sites increased from four to nine, and of these, six are in hilly regions, one is from a plateau region, and two are from mountainous areas. Hilly areas are represented with the most sentinel sites because infections in Sichuan are most common in these regions. The nine Sichuan sites have remained constant since 2005, and there is a commitment to continue longitudinal tracking of schistosomiasis infections in these sites regardless of their transmission trajectories in order to take maximum advantage of a consistent longitudinal dataset. Diagnostic protocols for the sentinel system have changed to accommodate the changing transmission profiles of sentinel sites, particularly challenges in detecting low-intensity infections. As of 2011, for instance, sentinel sites in Sichuan Province test all those classified as positive using a serologic assay with both Kato-Katz and a miracidium hatch test. These multiple rounds of diagnoses, and a history of intensive surveillance generally in these historically endemic regions, have led to community-wide surveillance fatigue in some sentinel sites. Costs of implementing the ongoing sentinel surveillance are likely to increase as it becomes more challenging to achieve high levels of community participation, necessitating more intensive follow-up with residents and more sensitive diagnostic methods. Sampling methodology The number of sentinel sites sustained by the system is based on the level of funding provided by the central government, and provinces with higher prevalence and infection intensity are allotted more resources (and thus sites). To select sites, national and provincial level OEDs consider two factors: variation in infection intensity (sites are selected from communities with heavy, medium and light infections based on infection intensity measurements from prior routine surveillance); and eco-epidemiological zone (e.g., plateau, lake, hilly or mountainous regions). Sentinel system screening has been conducted by Kato-Katz since 1989, and the IHA test was added in 2000, with patients found positive by IHA confirmed with the Kato-Katz exam. If acute or advanced cases are found, epidemiological investigations are conducted to collect more detailed information about each case. The overall design of the sentinel system was developed and approved by a national steering committee, yielding a system that is consistent among all surveillance sites, and making the data comparable between sites in all seven endemic provinces. It is sometimes necessary to move sites when implementing an intensive longitudinal surveillance system. For instance, after >10 years at one site in Sichuan, the population exhibited fatigue from yearly surveys and a new site was substituted to increase community participation. In one instance, as infection intensity and prevalence decreased dramatically across Sichuan Province, a very low prevalence sentinel site was replaced with a high prevalence site. Substituting sites interrupts the sentinel surveillance time-series, and data at new sites provide few insights into temporal trends until the series is allowed to accumulate over a number of years. National infectious disease reporting system The National infectious disease reporting system ( NIDRS ) was established in the 1950s [ 31 ], but it was not until 1989, with the passing of China's Law on Preventing and Treating Infectious Diseases, that the reporting of selected infectious diseases was mandated by law [ 32 ]. The NIDRS has historically involved a vertical reporting structure where hospitals reported the aggregated number of cases monthly by post to the county health department, who in turn report aggregate cases to the prefecture, then to the province, and finally to the central government. The information is used by Chinese public health officials to understand national patterns of infectious disease transmission, and to develop appropriate prevention and control programs [ 41 ]. Changes since 2004 Following the SARS outbreak, China's Law on Preventing and Treating Infectious Diseases was revised, and the NIDRS was transformed into an internet-based, real-time system [ 31 ]. The new system involves the routing of individual cases (rather than strictly aggregated data) up the hierarchy of public health institutions. The system provides for timely reporting of schistosomiasis cases to public health officials, allowing for rapid identification of, and response to, suspected re-emergence of the disease. The reporting of individual cases also enhances the spatial and temporal resolution of disease reporting, allowing for assessment of, for example, seasonal trends and persistent hot-spots of transmission [ 31 ]. Sampling methodology Cases of schistosomiasis diagnosed in clinics or hospitals are reported directly to the NIDRS . Reporting includes both acute and chronic cases, however acute schistosomiasis contributes a disproportionate number of reported schistosomiasis cases. Acute schistosomiasis is a severe allergic reaction to the migrating schistosomule in the blood stream following infection and typically occurs only in naïve populations such as children or military personnel deployed to a new region. The rapid onset of severe symptoms typical of acute schistosomiasis is more likely to prompt health-seeking behavior, therefore resulting in more frequent capture by NIDRS [ 42 ]. Because the symptoms of chronic S. japonicum infection are non-specific and often sub-clinical, chronic infections are poorly captured by clinic-based surveillance. Thus NIDRS schistosomiasis surveillance is considered key for identifying areas of disease emergence (or re-emergence), but not for assessing infection prevalence. What is more, diagnosis of a reported case is coded in the NIDRS as using either a clinical or a laboratory diagnostic, without further differentiation such as between stool and serologic exams [ 43 ]. The primary outcomes and other characteristics of the NIDRS system are listed in Table 2 . Schistosomiasis diagnostic and clinical capabilities, and therefore reporting capacity, differ considerably across China, and the transition to the internet-based NIDRS revealed several limitations of surveillance capacity in China [ 31 ]. The availability of computing resources and internet coverage required to connect directly to NIDRS varied within and between provinces, limiting the generation and transmission of electronic case reports [ 44 ]. Hospitals that lack sufficient information technology resources were required to send paper forms to the county CDC for entry into the online system, introducing a reporting delay. A national initiative was launched to equip such hospitals with computer workstations and network coverage, resulting in the delivery of >4,000 computers to hospitals and clinics in Sichuan Province in 2005, for example [ 45 ]. Other challenges were revealed relating to disparities in clinical resources. For instance, a disproportionate number of the cases of schistosomiasis present in NIDRS have been reported by county-level hospitals, in part because diagnostic capabilities are poorer at township-level hospitals, and diagnostic capabilities for parasitic diseases are sometimes entirely absent at private clinics and village healthcare centers [ 46 ]. Relationship to reporting systems for other diseases At present, the NIDRS covers 39 infectious diseases in three classes based on their public health importance (Table 3 ; [ 47 ]). The Class A list includes two diseases, cholera and plague, that are required to be reported within two hours of diagnosis. Pulmonary anthrax, SARS, poliomyelitis, and human infection with highly pathogenic avian influenza are in Class B, and are also required to be reported within two hours. Other Class B and Class C diseases must be reported within 24 hours of diagnosis. Class C diseases include newly identified diseases or diseases with emerging public health importance, including foot and mouth disease. Schistosomiasis was listed as a Class B nationally reportable infectious disease in 2004 [ 47 ], in part to rapidly capture reports of acute cases that may signal an outbreak of infection [ 43 ]. Future directions in surveillance Ideally, surveillance systems provide data that support evidence-based policy decisions and contribute to effective and efficient disease control [ 48 , 49 ]. Additionally, effective surveillance systems must respond to advances in technology and changes in programmatic goals. While multiple, complementary systems can provide rich spatial and temporal information on the distribution of cases, resource constraints necessitate surveillance systems that minimize redundancies, particularly in low- and middle-income countries. Key factors in the evolution of China's schistosomiasis surveillance over the past six decades include the World Bank loan (and the rise in cases following its termination in 2001); changes in the epidemiology of schistosomiasis; changes in available technologies for diagnosis and treatment [ 50 ]; and major public health events such as the 2003 SARS epidemic. The SARS epidemic in particular revealed weaknesses in China's public health system with respect to the detection of, and prompt response to, disease outbreaks [ 31 ]. Following the epidemic, the country's infectious disease surveillance infrastructure underwent a massive overhaul, which included construction of new infectious disease facilities and laboratories that were crucial to improving the quality of surveillance data, as well as facilitation of improved data sharing and of more efficient coordination between public health agencies. The overhaul also included a complete redesign of the real-time infectious disease reporting system [ 51 ], as well as the Ministry of Health's listing of schistosomiasis among diseases such as HIV/AIDS, tuberculosis and hepatitis B as a high priority for control [ 25 ]. This classification raised the profile of the disease among public health officials, increased the commitment of health care workers and government agencies to schistosomiasis elimination, and stimulated heightened surveillance and control efforts [ 43 , 52 ]. Throughout these major transitions, China has made remarkable progress towards controlling schistosomiasis. The country's evolving, parallel surveillance systems have each played a distinct role in characterizing the dynamic state of schistosomiasis transmission across diverse regions of the country. Thus far, the systems and the information they provide have been used independently; numerous reports, based on routine surveys, sentinel surveys and national surveys have been published and used to develop control strategies and monitor progress towards achieving elimination. More recently, there has been an effort to expand the use of these data beyond endemic disease control. Routine surveillance data, for instance, has been used to provide evidence of re-emerging schistosomiasis in Sichuan Province in areas that had previously achieved local control [ 7 ]. Surveillance data from multiple systems can be fruitfully combined, as evidenced by integration of national and NIDRS surveillance data from all provinces to estimate the incidence, prevalence and disease burden of schistosomiasis, as well as malaria, hookworm and other water, sanitation, and hygiene-associated infections [ 53 ]. The analysis revealed large regional disparities in the burden of these infections, particularly among children, who experience the greatest risk of water, sanitation and hygiene-attributable disease in China. In a separate study, combined national and NIDRS data were used to estimate the future burden of these infections under climate change [ 54 ]. While this work demonstrates the value of China's parallel surveillance systems, understanding of how data from these systems can be effectively combined, to characterize and inform the many dimensions of disease transmission and control, remains limited. The overlapping nature of these systems, both spatially and temporally, provides a unique opportunity to explore how a country can maximize the value of multiple, distinctive information sources to inform public health campaigns and investments. This in an area where future research is needed both in China and elsewhere. Furthermore, as many provinces move towards schistosomiasis elimination and case detection becomes ever more challenging and expensive, it will be even more important to determine how the data, sampling approaches, and detection methods these systems provide can be used to optimize the performance of surveillance efforts. Successful schistosomiasis elimination will require the ability to efficiently and accurately identify the few remaining infected individuals and non-human reservoirs among large uninfected populations. This will be complicated by decreased diagnostic sensitivity owing to reduced infection intensity, testing fatigue among residents as a result of long-standing repeated examinations, and ever-increasing health care costs associated with diagnosis. To be successful, China's surveillance system must be prepared to modify sampling methods, reporting approaches and diagnostics in response to new scientific and technological advances and changing schistosomiasis infection patterns. Recent attempts to establish diagnostic standards through the systematic evaluation of an array of immunological assays [ 55 ] and new mobile reporting methods, discussed below, are a promising step in this direction. Further, as China's national disease burden continues its shift from infectious diseases to chronic, non-communicable diseases (NCDs), additional strain will be placed on surveillance personnel and institutions currently devoted to infectious disease, and opportunities for integrating the surveillance and care of NCDs and infectious diseases will be needed [ 56 , 57 ]. Care must be taken to ensure that the resources are adequate and to avoid surveillance fatigue among healthcare workers and those in sentinel villages. For China, research and innovations needed to succeed in continuing control and elimination of schistosomiasis include the development of diagnostic tools with higher sensitivity for the detection of human infections, and new ways of assessing human susceptibility to infection. In addition, effective ways to deploy recently developed tools for detecting the infectious stage of schistosomes (cercariae) in the natural environment [ 30 , 58 , 59 ] will be needed. Likewise, mobile disease reporting technologies (e.g. use of mobile phones or positioning systems with wireless capability for spatially explicit reporting of cases or vectors) show considerable promise and have been recently applied for surveillance of other neglected tropical diseases [ 60 ]. These technologies have been used in field surveys of schistosomiasis endemic areas, to file reports of acute cases and infected snails into a centralized surveillance database [ 9 ], and to coordinate responses to disease outbreaks after the 2008 earthquake in Sichuan province [ 16 , 61 ]. Yet effective integration of mobile technologies into large-scale surveillance procedures is still relatively new, and will need further evaluation. At the same time, new approaches will be needed to estimate the cost-effectiveness of the combination of existing surveillance systems in achieving specific surveillance endpoints, such as a characterization of spatial and temporal variability in cases. While considerable effort has been devoted to evaluating surveillance systems from a functional standpoint [ 62 – 66 ] —assessing timeliness [ 67 ], record accuracy [ 68 , 69 ] and enhancing operational efficiency [ 70 – 72 ] —few analytical tools exist for quantitatively assessing how changes to design translate into improved (or inferior) estimates of the underlying rates of disease taking into account system sensitivity, coverage, bias and other factors. Development of such tools in China would provide a useful precedent as schistosomiasis surveillance is increasingly brought online in other countries. Global lessons Globally, there is a major movement underway to reduce the burden of helminthiases. The priorities for the global control and elimination of schistosomiasis, along with STHs, were articulated by the WHO in 2013 [ 73 ]. The surveillance needs of these global efforts are immense. Nearly 240 million people, most of whom live in sub-Saharan Africa, are estimated to require preventive chemotherapy for schistosomiasis alone as of 2009, yet coverage is reaching only 8% of those in need [ 74 , 75 ]. While many factors contribute to this deficit, the lack of surveillance capabilities in nearly all schistosomiasis-endemic countries is a major obstacle to efficiently expanding coverage [ 74 ]. Accurate, timely, country-level data are generally unavailable, and targeting drug delivery for schistosomiasis control in these settings is predominantly based on risk maps generated using ad-hoc epidemiological data (e.g., research data, historical drug delivery data, etc.), rather than surveillance [ 76 ]. China's history with surveillance of schistosomiasis can serve as a guide for countries, in sub-Saharan Africa and elsewhere, that require low-cost, simple surveillance and diagnostic technologies. Given the current efforts to roll-out preventative chemotherapy, there will be a need to identify areas where disease burden is declining and mass-treatment efforts can be transitioned to targeted treatment and monitoring, as well as areas where disease burden is not declining, where more broad control efforts are needed. Throughout Africa, where over 90% of schistosomiasis cases occur [ 77 ], the capacity to conduct population-based surveys is limited [ 78 ], and current control recommendations do not include surveillance activities [ 74 ]. There is some movement towards developing the infrastructure for parallel active surveillance systems—sentinel systems to track progress of control strategies coupled with regular provincial surveys—in highly endemic areas. China's sentinel systems offer a model for deploying targeted surveillance resources and monitoring control progress over time. Similar to China's provincial surveys, which borrow from standardized national survey protocols, province-level schistosomiasis mapping of all endemic districts has recently begun in a few countries (e.g. Ethiopia), and is rapidly scaling up for other neglected tropical diseases such as trachoma, with the benefit of funds provided by the UK Department for International Development and The Bill & Melinda Gates Foundation [ 79 ]. These efforts are moving incrementally towards the electronic collection and reporting of epidemiological data, and may benefit from lessons learned during China's 2011 adoption of a simple, internet-based system to file, store and manage routine survey data. Crucially, this system provides a model for electronic data collection even in the absence of a sophisticated NIDRS capacity. As many low-resource countries aspire for control and even elimination of schistosomiasis, understanding the synergy of parallel systems will prove cost-effective both by reducing unnecessary redundancy in data collection and in maximizing programming dollars. The systems described in China have distinct utility, but how these systems may be translated to other areas of schistosomiasis endemicity is not yet known and must be explored. Looking forward, targeted surveillance capacity is needed to understand and respond to the transmission consequences of rapid environmental, social, and technological changes, such as the construction of dams and changes in agricultural practices, that can result in an upsurge in schistosomiasis cases [ 54 , 77 , 80 – 82 ]. China's sentinel system offers a model for deploying targeted surveillance capacity in regions undergoing such changes [ 83 ]. Optimal surveillance of schistosomiasis and other neglected tropical diseases must evolve with changing programmatic needs and technical advances. Currently, real-time electronic surveillance is limited by lack of broadband internet infrastructure in many rural schistosomiasis-endemic areas. However, mobile disease reporting technologies offer new opportunities in areas where traditional means of communication are limited [ 6 ], and may become increasingly relevant as attention turns to elimination. In the context of elimination, new diagnostic tools are urgently needed that offer rapid, sensitive, low-cost alternatives to current methods. China is currently dealing with this issue [ 14 ], and as other settings move towards achieving elimination, the traditional diagnostics that are currently relied upon by many surveillance systems will need to be replaced [ 84 ]. Recently developed methods offer improved opportunities for point-of-care surveillance (e.g., [ 85 ]), and future investments will be needed to ensure advances like these are pursued and, ultimately, incorporated into surveillance practice.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7301433/

Incidence and treatment costs of severe bacterial infections among people who inject heroin: A cohort study in South London, England

Highlights • People who inject heroin have extremely high risk of severe bacterial infections. • Women have higher risk than men. • Among those starting treatment for heroin dependence, high risk persists for at least ten years. • Annual hospital treatment costs for injecting-related bacterial infections in London are £4.5m. Background People who inject drugs often get bacterial infections. Few longitudinal studies have reported the incidence and treatment costs of these infections. Methods For a cohort of 2335 people who inject heroin entering treatment for drug dependence between 2006 and 2017 in London, England, we reported the rates of hospitalisation or death with primary causes of cutaneous abscess, cellulitis, phlebitis, septicaemia, osteomyelitis, septic arthritis, endocarditis, or necrotising fasciitis. We compared these rates to the general population. We also used NHS reference costs to calculate the cost of admissions. Results During a median of 8.0 years of follow-up, 24 % of patients (570/2335) had a severe bacterial infection, most commonly presenting with cutaneous abscesses or cellulitis. Bacterial infections accounted for 13 % of all hospital admissions. The rate was 73 per 1000 person-years (95 % CI 69–77); 50 times the general population, and the rate remained high throughout follow-up. The rate of severe bacterial infections for women was 1.50 (95 % CI 1.32–1.69) times the rate for men. The mean cost per admission was £4980, and we estimate that the annual cost of hospital treatment for people who inject heroin in London is £4.5 million. Conclusions People who inject heroin have extreme and long-term risk of severe bacterial infections. 1 Introduction Bacterial infections are common among people who inject illicit drugs. Cutaneous abscesses, cellulitis and other localised infections are some of the most frequent reasons for medical care in this population. These infection can be serious, sometimes requiring hospital treatment or leading to complications such as invasive infections and amputations. Cross-sectional studies show that between 7 % and 37 % of people who inject drugs report a soft tissue infection in the last 6–12 months, and lifetime prevalence may be as high as 70 % ( Coull et al., 2014 ; Larney et al., 2017 ). Invasive bacterial infections at sites such as the bones, joints, heart, and blood, are also more common among people who inject drugs than the general population, and have a high mortality risk ( Frontera and Gradon, 2000 ; Peterson et al., 2014 ). Time-series data suggest that the number of hospital admissions for injecting-related bacterial infections is increasing in the US and the UK ( Ciccarone et al., 2016 ; Lewer et al., 2017 ). Several elements of a causal pathway have been established. These include colonisation of the skin; transferring bacteria onto drugs when they are transported (for example in the mouth); frequent breaking of the skin when injecting (and sometimes through other injuries) ( Phillips et al., 2017 ); and longer-term damage to the skin, soft tissue and veins with acids and particulate matter in drug preparations, which increases vulnerability to infection ( Harris et al., 2019 ; Hope et al., 2008 ; Murphy et al., 2001 ; Packer et al., 2019 ). Clusters of unusual infections such as anthrax and botulism have been observed among people who use illicit drugs ( Trayner et al., 2018 ), but there is likely to be many more infections and deaths caused by common species such as streptococci and staphylococci from the patient's own skin or mouth ( Gordon, 2005 ). Cross-sectional studies have identified that women, people who inject subcutaneously, homeless people, and those who inject stimulants have raised prevalence ( Hope et al., 2008 , 2016 ; Murphy et al., 2001 ). These groups may inject more frequently or have poorer access to sterile and sharp injecting equipment. However, relatively few studies have reported the incidence and costs of bacterial infections among people who inject drugs, particularly in comparison to the large number of studies of blood-borne viral infections in this population ( Degenhardt et al., 2016 ). This may be because most existing studies into bacterial infection use cross-sectional self-report data. In contrast to this, we used longitudinal electronic health record data to estimate the rate and treatment costs for severe bacterial infections in a cohort of people who inject heroin in South London, England. 2 Materials and methods 2.1 Data source We used data from the Clinical Records Interactive Search (CRIS) resource at the South London and Maudsley NHS Foundation Trust Biomedical Research Centre. This is a research repository of anonymised data derived from the electronic health record system of a mental healthcare provider in South London, England ( Perera et al., 2016 ). The study population was 2335 patients aged 18–64 entering community-based substance use treatment between 1 January 2006 and 31 March 2017, with reported use of heroin and drug injection. Patients were linked using NHS number, date of birth, sex and postcode to inpatient hospital admissions data from the national Hospital Episode Statistics for England database, and to mortality data from the UK Office for National Statistics. Linkage was conducted by NHS Digital, a public sector statistical agency. The end of follow-up was the participant's 65th birthday, death, or 31 March 2017. Some patients have long periods of engagement with the drug treatment service, while other only attend one appointment, but data linkage was available for all patients regardless of their engagement with the service. We treated admissions within two days of discharge after a previous admission as a single admission. We also accessed hospital admission data for all residents in the healthcare provider's catchment area of the London Boroughs of Croydon, Lambeth, Lewisham and Southwark (the 'comparison group'). 2.2 Outcome measures We defined severe bacterial infections as hospital admission with a primary cause of cutaneous abscess (ICD-10 code L02), cellulitis (L03), phlebitis or thrombophlebitis (I80), sepsis or septicaemia (A40, A41), endocarditis (I30.1, I39, I33.0, 140.0, I41.0), septic arthritis or osteomyelitis (M86, M00, M463, M46.5), and necrotising fasciitis (M72.6). We also counted all-cause hospital admissions. 2.3 Participant characteristics Data were derived from routinely collected information. Drugs used by patients and 'route of administration' (i.e. whether the patient reports injecting) were from the treatment service's National Drug Treatment Monitoring System (NDTMS) data set. This is a standardised patient assessment conducted periodically by drug treatment services in England ( Marsden et al., 2009 ; Public Health England, 2018a ). We identified that patients injected heroin if this was recorded on any NDTMS record during follow-up. In most of these cases, heroin injection was recorded on the earliest record, but we also included patients where it was recorded later, since some patients do not initially disclose injection. Date of birth and sex were taken from the healthcare provider's central patient database. For descriptive purposes, we reported: (a) drugs other than heroin were listed for at least 10 % of participants, (b) whether homelessness or unstable housing was listed in patient databases, and (c) whether serious mental illness, defined as a diagnosis or bipolar disorder or schizophrenia, was listed in patient databases. 2.4 Statistical analysis To calculate an expected number of admissions, we first calculated admission rates in the comparison group from 2006 to 2016 (as the closest available match to the study cohort) by age group, sex, and type of infection. The denominators were the sum of mid-year population estimates ( Office for National Statistics, 2017 ) in the service provider's catchment boroughs, and the numerators were the numbers of hospital admissions. We applied these rates to the time-at-risk within each age and sex group in the study cohort, accounting for patients ageing during follow-up and not counting time while patients were admitted to hospital. The standardised admission ratio (SAR) was the observed admissions divided by the expected admissions (i.e. indirect standardisation). We also stratified these results by sex. We estimated the cost of each hospital admission using the NHS 2014/15 national reference costs ( Department of Health and Social Care, 2015 ), in which hospitals report spend according to diagnoses, clinical procedures and the duration of admissions. The cost of each admission is calculated using a combination of the diagnosis codes, procedure codes, and length of admission. To contextualise these costs, we estimated the annual cost of hospital treatment for bacterial infections among all people who inject drugs in London, by applying the admission rates in our cohort to an existing capture-recapture population estimate of 11,351 people who inject drugs in London in 2011/12 ( Hay et al., 2014 ), and using the mean cost of treatment for each diagnosis from our cohort. We used a Monte-Carlo method to estimate statistical uncertainty around this estimate, with details of this method provided in Supplementary Information. To compare the duration of hospital admission for patients who inject drugs and the general population, we drew a random sample of admissions from the comparison group, stratified by age group, sex, primary diagnosis, and year, at a ratio of 1:1. Analysis was conducted using R version 3.5.3 ( R Core Team, 2019 ). 2.1 Data source We used data from the Clinical Records Interactive Search (CRIS) resource at the South London and Maudsley NHS Foundation Trust Biomedical Research Centre. This is a research repository of anonymised data derived from the electronic health record system of a mental healthcare provider in South London, England ( Perera et al., 2016 ). The study population was 2335 patients aged 18–64 entering community-based substance use treatment between 1 January 2006 and 31 March 2017, with reported use of heroin and drug injection. Patients were linked using NHS number, date of birth, sex and postcode to inpatient hospital admissions data from the national Hospital Episode Statistics for England database, and to mortality data from the UK Office for National Statistics. Linkage was conducted by NHS Digital, a public sector statistical agency. The end of follow-up was the participant's 65th birthday, death, or 31 March 2017. Some patients have long periods of engagement with the drug treatment service, while other only attend one appointment, but data linkage was available for all patients regardless of their engagement with the service. We treated admissions within two days of discharge after a previous admission as a single admission. We also accessed hospital admission data for all residents in the healthcare provider's catchment area of the London Boroughs of Croydon, Lambeth, Lewisham and Southwark (the 'comparison group'). 2.2 Outcome measures We defined severe bacterial infections as hospital admission with a primary cause of cutaneous abscess (ICD-10 code L02), cellulitis (L03), phlebitis or thrombophlebitis (I80), sepsis or septicaemia (A40, A41), endocarditis (I30.1, I39, I33.0, 140.0, I41.0), septic arthritis or osteomyelitis (M86, M00, M463, M46.5), and necrotising fasciitis (M72.6). We also counted all-cause hospital admissions. 2.3 Participant characteristics Data were derived from routinely collected information. Drugs used by patients and 'route of administration' (i.e. whether the patient reports injecting) were from the treatment service's National Drug Treatment Monitoring System (NDTMS) data set. This is a standardised patient assessment conducted periodically by drug treatment services in England ( Marsden et al., 2009 ; Public Health England, 2018a ). We identified that patients injected heroin if this was recorded on any NDTMS record during follow-up. In most of these cases, heroin injection was recorded on the earliest record, but we also included patients where it was recorded later, since some patients do not initially disclose injection. Date of birth and sex were taken from the healthcare provider's central patient database. For descriptive purposes, we reported: (a) drugs other than heroin were listed for at least 10 % of participants, (b) whether homelessness or unstable housing was listed in patient databases, and (c) whether serious mental illness, defined as a diagnosis or bipolar disorder or schizophrenia, was listed in patient databases. 2.4 Statistical analysis To calculate an expected number of admissions, we first calculated admission rates in the comparison group from 2006 to 2016 (as the closest available match to the study cohort) by age group, sex, and type of infection. The denominators were the sum of mid-year population estimates ( Office for National Statistics, 2017 ) in the service provider's catchment boroughs, and the numerators were the numbers of hospital admissions. We applied these rates to the time-at-risk within each age and sex group in the study cohort, accounting for patients ageing during follow-up and not counting time while patients were admitted to hospital. The standardised admission ratio (SAR) was the observed admissions divided by the expected admissions (i.e. indirect standardisation). We also stratified these results by sex. We estimated the cost of each hospital admission using the NHS 2014/15 national reference costs ( Department of Health and Social Care, 2015 ), in which hospitals report spend according to diagnoses, clinical procedures and the duration of admissions. The cost of each admission is calculated using a combination of the diagnosis codes, procedure codes, and length of admission. To contextualise these costs, we estimated the annual cost of hospital treatment for bacterial infections among all people who inject drugs in London, by applying the admission rates in our cohort to an existing capture-recapture population estimate of 11,351 people who inject drugs in London in 2011/12 ( Hay et al., 2014 ), and using the mean cost of treatment for each diagnosis from our cohort. We used a Monte-Carlo method to estimate statistical uncertainty around this estimate, with details of this method provided in Supplementary Information. To compare the duration of hospital admission for patients who inject drugs and the general population, we drew a random sample of admissions from the comparison group, stratified by age group, sex, primary diagnosis, and year, at a ratio of 1:1. Analysis was conducted using R version 3.5.3 ( R Core Team, 2019 ). 3 Results The cohort included 2335 patients with a total follow-up time of 16,242 years (median 8.0, range 0–11.2). The mean age at baseline was 36.3 years (sd 8.4) and 1727 (74 %) were male, which is similar to the profile of people entering opioid treatment nationally ( Public Health England and Department of Health, 2017 ) ( Table 1 ). 352 patients died (15 %) during follow-up, of whom <10 had an underlying cause of a bacterial infection. Table 1 Cohort characteristics at baseline. Table 1 Variable Level Number (%) Age at index 18−24 217 (9) 25−34 842 (36) 35−44 921 (39) 45−54 314 (13) 55−64 41 (2) Mean (sd) 36.3 (8.4) Age during follow-up Mean (sd) 39.9 (8.4) Sex Male 1727 (74) Female 608 (26) Ethnicity White 2065 (88) Black 119 (5) Mixed 66 (3) Asian 28 (1) Other 57 (2) Other drugs Crack cocaine 1950 (84) Alcohol 1198 (51) Cannabis 513 (22) Benzodiazepines 386 (17) Unstable housing 1410 (60) Severe mental health problems 366 (16) Total 2335 (100) Of an initial cohort of 2469 patients, 134 (5 %) were not linked to NHS hospital data and were excluded from analysis. Excluded patients did not differ in terms of sex (p = 0.49), but were slightly younger at baseline (mean age 34.2 vs. 36.4 years; p = 0.005). 3.1 Hospitalisation Patients were hospitalised 9315 times, of which 1180 (13 %) were primarily caused by a bacterial infection. The incidence density was 73 hospitalisations per 1000 person-years (95 % CI 69–77). The rate of bacterial infections was high throughout follow-up, with no evidence of a change in incidence over time for either men or women ( Fig. 1 ). Fig. 1 Rate of hospital admission for severe bacterial infection in a cohort of 2335 people who inject heroin in South London, England, by time after first treatment episode (error bars show 95% confidence intervals). Fig. 1 Compared to the general population, the study cohort was 50.0 (95 % CI 47.2−52.9) times more likely to be admitted to hospital for treatment of a bacterial infection. Hospitalisation rates were substantially raised for each type of bacterial infection ( Table 2 ). The rate of all-cause admission was also raised, but much less so, at 3.7 times the general population. Table 2 Hospital admissions and costs of treatment for severe bacterial infections in a cohort of 2335 people who inject heroin in South London, England, with 16,434 years of follow-up. Table 2 Primary diagnosis Observed admissions Expected admissions SAR (95 % CI) Mean cost, £ (sd) Median cost, £ (IQR) Abscess 487 9.3 52.4 (47.8−57.2) 4307 (3035) 3898 (2,660−5,296) Cellulitis 282 7.8 36.0 (31.9−40.4) 3579 (2503) 2731 (1,880−4,432) Phlebitis and thrombophlebitis 233 2.0 115.2 (100.9−131.0) 3261 (2969) 2277 (1,772−4,136) Septicaemia and bacteraemia 56 2.3 24.3 (18.3−31.5) 8687 (5060) 9250 (5,221−9,763) Osteomyelitis and septic arthritis 42 0.2 174.4 (125.7−235.8) 14,134 (52,843) 5694 (3,980−7,129) Endocarditis 82 1.9 43.7 (34.8−54.3) 12,963 (7765) 11,951 (6,893−15,197) Necrotising Fasciitis 9 <0.1 599.4 (274.1−1,137.9) 10,815 (7159) 11,926 (4,274−14,839) All bacterial infections * 1180 23.6 50.0 (47.2−52.9) 4980 (12,431) 3022 (2,148−5,296) All-cause 9274 2467.5 3.8 (3.7−3.8) ** ** SAR = Standardised admission ratio. * The total number of bacterial infections is less than the sum of each individual diagnosis because some admissions have two primary diagnoses (resulting from the process of merging hospital admissions that were within two days of each other). ** We did not calculate the cost of admissions with primary diagnosis unrelated to bacterial infection. The rate of hospitalisation among women was 1.50 (95 % CI 1.32–1.69) that of men, with similar or higher rates across diagnoses ( Fig. 2 ). Although women also had a higher rate of all-cause hospital admission than men, this was in proportion to differences between women and men in the general population. All-cause SARs for men and women were therefore similar, at 3.7 and 4.0 respectively. In contrast, women who inject heroin had higher rates of admission for bacterial infections than men, and the differences were disproportionate to underlying differences between men and women in the general population. The SARs for bacterial infections were therefore higher for women than for men (see Supplementary Information). Fig. 2 Rate of hospital admission for severe bacterial infection in a cohort of 2335 people who inject heroin, by primary diagnosis. (error bars show 95% confidence intervals). Fig. 2 Compared to hospital inpatients from the general population with the same age, sex, primary cause of hospital admission and year of admission, patients who inject heroin had a longer duration of admission (mean 7.4 days) and were more likely to self-discharge (13 % vs. 1 %). Comparisons of the duration, admission method and discharge methods for people who inject heroin and the general population are provided in Supplementary Information. 3.2 Cost of treatment The mean cost per admission was £4980 (sd. £12,431), with higher costs for invasive infections. The cost per admission was heavily right-skewed (common in healthcare cost data) and means were higher than medians. Modelling suggested 869 admissions for treatment of bacterial infections per year among 11,351 people who inject drugs in London, with a total cost of £4.5 million (95 % CI £3.7-£5.4 million), based on 2014/15 prices. 3.1 Hospitalisation Patients were hospitalised 9315 times, of which 1180 (13 %) were primarily caused by a bacterial infection. The incidence density was 73 hospitalisations per 1000 person-years (95 % CI 69–77). The rate of bacterial infections was high throughout follow-up, with no evidence of a change in incidence over time for either men or women ( Fig. 1 ). Fig. 1 Rate of hospital admission for severe bacterial infection in a cohort of 2335 people who inject heroin in South London, England, by time after first treatment episode (error bars show 95% confidence intervals). Fig. 1 Compared to the general population, the study cohort was 50.0 (95 % CI 47.2−52.9) times more likely to be admitted to hospital for treatment of a bacterial infection. Hospitalisation rates were substantially raised for each type of bacterial infection ( Table 2 ). The rate of all-cause admission was also raised, but much less so, at 3.7 times the general population. Table 2 Hospital admissions and costs of treatment for severe bacterial infections in a cohort of 2335 people who inject heroin in South London, England, with 16,434 years of follow-up. Table 2 Primary diagnosis Observed admissions Expected admissions SAR (95 % CI) Mean cost, £ (sd) Median cost, £ (IQR) Abscess 487 9.3 52.4 (47.8−57.2) 4307 (3035) 3898 (2,660−5,296) Cellulitis 282 7.8 36.0 (31.9−40.4) 3579 (2503) 2731 (1,880−4,432) Phlebitis and thrombophlebitis 233 2.0 115.2 (100.9−131.0) 3261 (2969) 2277 (1,772−4,136) Septicaemia and bacteraemia 56 2.3 24.3 (18.3−31.5) 8687 (5060) 9250 (5,221−9,763) Osteomyelitis and septic arthritis 42 0.2 174.4 (125.7−235.8) 14,134 (52,843) 5694 (3,980−7,129) Endocarditis 82 1.9 43.7 (34.8−54.3) 12,963 (7765) 11,951 (6,893−15,197) Necrotising Fasciitis 9 <0.1 599.4 (274.1−1,137.9) 10,815 (7159) 11,926 (4,274−14,839) All bacterial infections * 1180 23.6 50.0 (47.2−52.9) 4980 (12,431) 3022 (2,148−5,296) All-cause 9274 2467.5 3.8 (3.7−3.8) ** ** SAR = Standardised admission ratio. * The total number of bacterial infections is less than the sum of each individual diagnosis because some admissions have two primary diagnoses (resulting from the process of merging hospital admissions that were within two days of each other). ** We did not calculate the cost of admissions with primary diagnosis unrelated to bacterial infection. The rate of hospitalisation among women was 1.50 (95 % CI 1.32–1.69) that of men, with similar or higher rates across diagnoses ( Fig. 2 ). Although women also had a higher rate of all-cause hospital admission than men, this was in proportion to differences between women and men in the general population. All-cause SARs for men and women were therefore similar, at 3.7 and 4.0 respectively. In contrast, women who inject heroin had higher rates of admission for bacterial infections than men, and the differences were disproportionate to underlying differences between men and women in the general population. The SARs for bacterial infections were therefore higher for women than for men (see Supplementary Information). Fig. 2 Rate of hospital admission for severe bacterial infection in a cohort of 2335 people who inject heroin, by primary diagnosis. (error bars show 95% confidence intervals). Fig. 2 Compared to hospital inpatients from the general population with the same age, sex, primary cause of hospital admission and year of admission, patients who inject heroin had a longer duration of admission (mean 7.4 days) and were more likely to self-discharge (13 % vs. 1 %). Comparisons of the duration, admission method and discharge methods for people who inject heroin and the general population are provided in Supplementary Information. 3.2 Cost of treatment The mean cost per admission was £4980 (sd. £12,431), with higher costs for invasive infections. The cost per admission was heavily right-skewed (common in healthcare cost data) and means were higher than medians. Modelling suggested 869 admissions for treatment of bacterial infections per year among 11,351 people who inject drugs in London, with a total cost of £4.5 million (95 % CI £3.7-£5.4 million), based on 2014/15 prices. 4 Discussion In this cohort of people who inject heroin in South London, bacterial infections were a major cause of morbidity but not mortality. Our results show a 50-fold increased risk of severe bacterial infections when compared to the general population, with high risk persisting for several years after starting treatment. Hospital treatment of bacterial infections in people who inject drugs can be complex and expensive. Clinicians sometimes retain these patients in hospital for longer to ensure antibiotic courses are completed. Patients may leave hospital against medical advice if opiate substitution is unavailable ( Summers et al., 2018 ) and often have poor continuing care, leading to readmission and antimicrobial-resistant infections. In our sample, 13 % of admissions ended in discharge against medical advice, compared to 1 % of admissions in the comparison group. Our results suggest that the cost of hospitalisations for these conditions is £4.5 million per year in London (2014/15 prices), which is substantial considering that the total expenditure on drug misuse treatment in London in 2017/18 was £64 million ( Ministry of Housing, Communities and Local Government, 2018). A previous study estimated costs of hospital treatment for bacterial infection among people who inject drugs in England using the mean cost of treatment in the general population, at £944 to £1566 per admission (2004/05 prices) ( Hope et al., 2008 ). Our data show substantially higher costs among people who inject drugs, at £4980 per admission (2014/15 prices), which is similar to a study of 128 episodes of bacterial infection in people who inject drugs at one London hospital ( Marks et al., 2013 ). The incidence rate of severe bacterial infections (73/1000 person-years) was higher than rates observed in cohorts of people who inject opiates in Sweden (24/1000) and Canada (61/1000) ( Dahlman et al., 2018 ; Lloyd-Smith et al., 2010 ) and lower than a rate observed in Switzerland (86/1000) ( Bassetti et al., 2002 ). The variation is likely due to differing demographics, injecting behaviours, types of heroin, local services, and the period when the study was conducted. The higher risk associated with female sex is consistent across these studies. Many studies have shown that women form a minority among people who inject drugs, with higher risk of acquiring infections. As well as higher risk of bacterial infections, some studies have shown than women who inject drugs have higher risk of acquiring blood-borne viral infections than men, with the difference at least partly due to injecting-related risks such as sharing syringes, injection by partners, and younger age at first injection ( Doherty et al., 2000 ; Tracy et al., 2014 ). Women who inject drugs may face greater self-stigma and social stigma relating to drug injection and injection-related injuries ( Iversen et al., 2015 ), and may therefore avoid health services. Policy recommendations to reduce HIV risk among women who inject drugs have argued that the most successful interventions focus on contextual factors such as women's intimate relationships, housing, employment, and childcare arrangements, rather injecting behaviours ( Pinkham et al., 2012 ), and this may also be true for bacterial infections. A large proportion of our sample (60 %) had experienced homelessness or housing problems. The association between housing problems and bacterial infections among people who inject drugs has been shown in previous studies ( Dahlman et al., 2018 ; Hope et al., 2008 ). Sleeping in homeless shelters and other temporary accommodation may be associated with increased bacterial colonisation ( Leibler et al., 2019 ). Homeless people are also more likely to inject in public places, which is associated with rushing the procedure, not cleaning skin before injecting, a lack of clean water for preparing drug solutions, and not having a clean surface to assemble the drugs ( Small et al., 2007 ), which increase the risk of infection. Interventions to improve housing in this population may reduce the risk of bacterial infections, as well as improving many other health and social outcomes. Among people using accommodation with shared bathroom facilities, improved shower hygiene may decrease the risk of colonisation ( Leibler et al., 2019 ). Qualitative research has found multiple barriers to healthcare among people who inject drugs. People may delay treatment due to normalisation of pain, fear of stigma in services, and concern about inadequate opioid substitution and pain control when admitted to hospital ( Neale et al., 2007 ; Summers et al., 2018 ). Hospitals sometimes employ drug and alcohol liaison workers a to improve accessibility for people who use drugs, and this may lead to more effective treatment ( Reeve et al., 2016 ). Early treatment may also be encouraged by specialist community clinics that provide antibiotics and wound care. These services are sometimes commissioned as part of community drug services, but may have become less available in England as funding for addictions services has reduced ( Advisory Council on the Misuse of Drugs, 2017 ). In addition to interventions focusing on gender, housing, and health service accessibility, there are several effective interventions that reduce injecting-related risk. These include interventions that reduce the need for injecting, such as opiate substitution; and interventions that improve injecting safety and hygiene, such as safe injecting facilities and provision of sterile injecting equipment ( Dunleavy et al., 2017 ). We observed continued high risk of severe bacterial ten years after initiation of treatment. This reflects the long-term nature of heroin dependence and highlights the need for continued health assessments and inspection of injecting sites. 4.1 Strengths and limitation Recruiting and retaining people who use drugs in traditional cohort studies can be challenging, and many studies are small and suffer from loss-to-follow-up. Strengths of our study include the large sample size, long follow-up, and complete data on hospitalisations and deaths. The linked hospital records were available from all NHS hospitals in England (rather than only local hospitals), which is important because people who use drugs are mobile and may use health services in other parts of the country. In our study, 55 % of admissions occurred outside of the four local London boroughs. Our sample is drawn from a community drug treatment service, and therefore excludes people who have never sought treatment. In England, an estimated three-quarters of people who use illicit opiates have had at least one episode of treatment and half are currently engaged with treatment ( Public Health England, 2018b ). Those who have never engaged with treatment may include both higher risk patients who are not accessing harm reduction services, and lower risk patients who have lower need for services. Given the high proportion of the population who have used drug treatment services, our results are likely to be a reasonable estimate of the rate of infection among people who inject heroin in London. We focused on severe infections that require inpatient treatment. Community surveys suggest that soft tissue infections are very common in people who inject drugs ( Larney et al., 2017 ), and many self-treat ( Monteiro et al., 2020 ) or wait for symptoms to resolve. Additionally, some patients will have sought treatment for less severe infections from general practitioners, but we did not have access to primary care data for this study. Consequently, this study only captures the most severe infections and provides a lower bound for healthcare utilisation. Although our data included potential risk factors for bacterial infections, such as housing status and mental health problems, we did not seek to analyse their association with the risk of bacterial infections. Previous studies have reported risk factors for bacterial infections in this population, and our data lacked detailed information relating to injecting risk, such as duration and frequency of injecting. This study highlights that cohorts based on electronic health records such as ours have strengths when reporting of incidence and costs, but may be limited when analysing risk factors or 'risk environments', which requires a detailed understanding of participants' context. 4.1 Strengths and limitation Recruiting and retaining people who use drugs in traditional cohort studies can be challenging, and many studies are small and suffer from loss-to-follow-up. Strengths of our study include the large sample size, long follow-up, and complete data on hospitalisations and deaths. The linked hospital records were available from all NHS hospitals in England (rather than only local hospitals), which is important because people who use drugs are mobile and may use health services in other parts of the country. In our study, 55 % of admissions occurred outside of the four local London boroughs. Our sample is drawn from a community drug treatment service, and therefore excludes people who have never sought treatment. In England, an estimated three-quarters of people who use illicit opiates have had at least one episode of treatment and half are currently engaged with treatment ( Public Health England, 2018b ). Those who have never engaged with treatment may include both higher risk patients who are not accessing harm reduction services, and lower risk patients who have lower need for services. Given the high proportion of the population who have used drug treatment services, our results are likely to be a reasonable estimate of the rate of infection among people who inject heroin in London. We focused on severe infections that require inpatient treatment. Community surveys suggest that soft tissue infections are very common in people who inject drugs ( Larney et al., 2017 ), and many self-treat ( Monteiro et al., 2020 ) or wait for symptoms to resolve. Additionally, some patients will have sought treatment for less severe infections from general practitioners, but we did not have access to primary care data for this study. Consequently, this study only captures the most severe infections and provides a lower bound for healthcare utilisation. Although our data included potential risk factors for bacterial infections, such as housing status and mental health problems, we did not seek to analyse their association with the risk of bacterial infections. Previous studies have reported risk factors for bacterial infections in this population, and our data lacked detailed information relating to injecting risk, such as duration and frequency of injecting. This study highlights that cohorts based on electronic health records such as ours have strengths when reporting of incidence and costs, but may be limited when analysing risk factors or 'risk environments', which requires a detailed understanding of participants' context. 5 Conclusions People who inject heroin have extreme and long-term risk of severe bacterial infections. Women are at higher risk than men. Data sharing Researchers with appropriate registrations and permissions can access data from CRIS. Contact the NIHR Maudsley Biomedical Research Centre for further information. https://www.maudsleybrc.nihr.ac.uk/ Ethical approval The dataset was approved as an anonymised data set for secondary data analyses by the Oxfordshire Research Ethics Committee C (reference number: 08/H0606/71+5). This analysis was approved by the South London and Maudsley NHS Foundation Trust Biomedical Research Centre CRIS Oversight Committee (reference number: 17−073). Funding DL and MH are funded by the 10.13039/501100012618 NIHR [Doctoral Research Fellowship DRF-2018-11-ST2-016, CDF-2016-09-014]. This work was supported by a grant from the 10.13039/100010269 Wellcome Trust [109823/Z/15/Z] to KIM. JS is supported by the NIHR Biomedical Research Centre for Mental Health at South London and Maudsley NHS Foundation Trust and King's College London; and is an NIHR Senior Investigator. The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health and Social Care. Role of the funding source The funders had no role in the study design, the collection, analysis and interpretation of data, in the writing of the report and in the decision to submit the article for publication. This manuscript represents independent research part funded by the National Institute for Health Research (NIHR) Biomedical Research Centre at South London and Maudsley NHS Foundation Trust and King's College London. Contribution statement DL and KIM conceived the study and drafted the analysis plan. All authors revised and approved the analysis plan. AJ and MP created the analytical dataset. DL conducted data analysis and drafted the manuscript. All authors critically revised the manuscript. All authors have read and approved the final manuscript. Declaration of Competing Interest MK is employed by the South London and Maudsley NHS Trust (SLaM) as clinical lead for Lambeth Addictions Consortium, which provides treatment for the patients included in this study. JS is a clinician and researcher and has worked extensively on clinical trials and wider research. JS's employer (King's College London) receives, unconnected to this specific study but connected to his wider work, project grant support and/or honoraria and/or consultancy payments from government agencies, charitable sources and also from pharmaceutical companies related to funding for clinical trials and research studies (for fuller information see www.kcl.ac.uk/ioppn/depts/addictions/people/hod.aspx ). Appendix A Supplementary data The following is Supplementary data to this article:

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3703949/

The Yersinia virulence effector YopM binds caspase-1 to arrest inflammasome assembly and processing

Summary Inflammasome assembly activates caspase-1 and initiates the inflammatory cell death program pyroptosis, which is protective against numerous pathogens. Consequently, several pathogens, including the plague causing bacterium Yersinia pestis , avoid activating this pathway to enhance their virulence. However, bacterial molecules that directly modulate the inflammasome have yet to be identified. Examining the contribution of Yersinia type III secretion effectors to caspase-1 activation, we identified the leucine-rich repeat effector YopM as a potent antagonist of both caspase-1 activity and activation. YopM directly binds caspase-1, which both inhibits caspase-1 activity and sequesters it to block formation of the mature inflammasome. Caspase-1 activation antagonizes Yersinia survival in vivo and consequently YopM inhibition of caspase-1 is required for Yersinia pathogenesis. Thus, a bacterium obstructs pyroptosis utilizing a direct mechanism of caspase-1 inhibition which is distinct from known viral or host inhibitors.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2437906/

Tumour endoproteases: the cutting edge of cancer drug delivery?

Despite progression in anticancer drug development and improvements in the clinical utilization of therapies, current treatment regimes are still dependent upon the use of systemic antiproliferative cytotoxic agents. Although these agents are unquestionably potent, their efficacy is limited by toxicity towards 'normal' cells and a lack of tumour selective targeting, resulting in a therapeutic index which is modest at best. Consequently, the development of more tumour selective cancer treatments, with better discrimination between tumour and normal cells is unequivocally an important goal for cancer drug discovery. One such strategy is to exploit the tumour phenotype as a mechanism for tumour-selective delivery of potent therapeutics. An exciting approach in this area is to develop anticancer therapeutics as prodrugs, which are non-toxic until activated by enzymes localized specifically in the tumour. Enzymes suitable for tumour-activated prodrug development must have increased activity in the tumour relative to non-diseased tissue and an ability to activate the prodrug to its active form. One class of enzyme satisfying these criteria are the tumour endoproteases, particularly the serine- and metallo-proteases. These proteolytic enzymes are essential for tumour angiogenesis, invasion and metastasis, the major defining features of malignancy. This review describes the concept behind development of tumour-endoprotease activated prodrugs and discusses the various studies to date that have demonstrated the huge potential of this approach for improvement of cancer therapy.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2430184/

The efficient packaging of Venezuelan equine encephalitis virus-specific RNAs into viral particles is determined by nsP1–3 synthesis

Alphaviruses are regarded as attractive systems for expression of heterologous genes and development of recombinant vaccines. Venezuelan equine encephalitis virus (VEE)-based vectors are particularly promising because of their specificity to lymphoid tissues and strong resistance to interferon. To improve understanding of the VEE genome packaging and optimize application of this virus as a vector, we analyzed in more detail the mechanism of packaging of the VEE-specific RNAs. The presence of the RNAs in the VEE particles during serial passaging in tissue culture was found to depend not only on the presence of packaging signal(s), but also on the ability of these RNAs to express in cis nsP1, nsP2 and nsP3 in the form of a P123 precursor. Packaging of VEE genomes into infectious virions was also found to be more efficient compared to that of Sindbis virus, in spite of lower levels of RNA replication and structural protein production.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9694496/

Bilateral Optic Neuritis after COVID-19 Vaccination: A Case Report

Background: Neuro-ophthalmic manifestations after vaccines are rare, with optic neuritis (ON) being the most common presentation. Patients with vaccine-related ON are similar to those with idiopathic ON. The temporal relationship between vaccination against and the occurrence of ON is vital. Here, we report a case of bilateral ON after the administration of the ChAdOx1-S nCoV-19 SARS-CoV-2 vaccine. Case: A 49-year-old healthy Asian female presented with sudden onset of bilateral blurred vision within 2 days. She complained of photophobia and extraocular pain upon movement over 3 days. Upon examination, her best corrected visual acuity (BCVA) was 20/30 in the right eye and 20/200 in the left eye. Anterior segment findings were unremarkable, with normal intraocular pressure. Fundoscopic examination revealed bilateral disc edema with vessel engorgement. Visual field examination revealed profound visual field defect in both eyes. She denied any trauma, use of new medication or medical history. She had received the ChAdOx1 nCoV-19 SARS-CoV-2 vaccine 14 days prior. Under suspicion of vaccine-related optic neuritis, she was given intravenous methylprednisolone 1 gm/day for 3 days, shifting to oral prednisolone under gradual tapering for 2 weeks. Conclusions: Typically presenting with sudden-onset visual decline and extraocular pain during movement, acute ON is generally idiopathic. Bilateral ON is rare, but quick identification is important because it can potentially lead to permanent loss of vision if left untreated. Vaccination-induced ON is even rarer but not difficult to treat. However, such patients require further evaluation and long-term follow-up because they may be prone to other neurological disorders in the future. 1. Introduction Since emerging in November 2019, COVID-19, or Coronavirus Disease 19, has become a priority healthcare issue. Induced by the SARS-CoV-2 virus, it presents a diverse variety of mainly respiratory symptoms, which can range from the asymptomatic to serious, life-threatening conditions, collectively referred to as severe acute respiratory syndrome. To date, it has claimed over 6 million deaths worldwide. As the COVID-19 pandemic continues, the development of vaccines against this virus has been crucial to prevent morbidity and mortality. To date, more than six vaccines have been approved for COVID-19 treatment (BNT162b2, mRNA-1273, ChAdOx1-S, BBV152 (Covaxin), Ad26.COV2.S, and Sputnik V (Gam-COVID-Vac)). More than 12.5 billion people worldwide have received these vaccines. Optic neuritis, an inflammatory optic nerve disorder, is relatively infrequent with a prevalence of only 0.1%. Its primary clinical characteristic is the rapid (1–2 days) onset of loss of vision, and it is aggravated by heat or exercise and is frequently accompanied by eye pain and/or headache [ 1 , 2 ]. Common symptoms and signs include reduced visual acuity, ocular pain during eye movement, reduced brightness perception, and altered color vision. While not common, the development of post-vaccination optic neuritis has occasionally been reported [ 3 , 4 , 5 , 6 , 7 , 8 , 9 ]. Since the widespread administration of vaccines against SARS-CoV-2, several potential adverse effects and diseases caused by the vaccine have been reported. In March 2021, there were several reports of patients experiencing thrombosis events with thrombocytopenia after AZ vaccination [ 10 ]. Cases of neurological or neuro-ophthalmological manifestations after COVID-19 vaccinations were also observed and have been published. One case of neuromyelitis optical spectrum disorder (NMOSD) after coronavirus mRNA-1273 vaccination has been reported [ 11 ]. Another case involving simultaneous optic neuritis and thyroiditis after SARS-CoV-2 vaccination has been reported [ 12 ]. A few adverse neuro-ophthalmic case report series related to COVID-19 vaccines have also recently emerged [ 8 , 13 , 14 , 15 ]. Most cases of post-vaccine optic neuritis are unilateral; however, here, we report a rare case of bilateral optic neuritis after ChAdOx1-S vaccination. 2. Case Presentation A 49-year-old Asian female visited our emergency department due to acute vision loss within 2 days. She had no prior medical history and denied any recent trauma. She had bilateral, extraocular pain while rotating her eyes. Examination revealed that the best-corrected visual acuity (BCVA) in the right eye was 20/30 and in the left eye was 20/200. Intraocular pressure OU was normal, and the anterior chamber was unremarkable. The swinging light test showed a positive relative afferent pupillary defect in the left eye. The Ishihara color vision test revealed 15/15 OU. Fundoscopy showed bilateral disc edema with the left eye more prominent than the right eye ( Figure 1 A). She had had her first inoculation with the ChAdOx1 COVID-19 vaccine around 2 weeks prior, after which the acute vision loss occurred. Under suspicion of post-vaccine ON, we performed emergent orbital MRI. T2-weighed MRI revealed significant enhancement of the bilateral optic nerve, more obvious on the left side than on the right side ( Figure 1 B). No parenchymal lesions were observed. Fluorescein angiography showed bilateral disc leakage in the late phase ( Figure 1 C). The patient's blood tests, including serum glucose, complete blood count, and kidney and liver function test returned normal. CRP and ESR were 2.2 mg/L and 26 mm/h, respectively. Immunological assessment covering ANA, anti-DNA, ANCA, anti-Ro, anti-La, and antiphospholipid antibodies returned negative, as did the serum anti-aquaporin-4 (AQP4) antibody test. Unfortunately, a serum anti-myelin-oligodendrocyte-glycoprotein (MOG) antibody test was not commercially available. The PCR test for SARS-CoV-2 based on nasopharyngeal swab was negative. The patient refused lumbar puncture at that time, and therefore, no CSF-related data are available. The patient was given high-dosage IV steroids (methylprednisolone 1 g/day for 3 days) with subsequent improvement in visual acuity from day 2, followed by oral prednisolone (1 mg/kg), tapering over 2 weeks. Seven weeks later, the patient's best-corrected visual acuity in both eyes was 20/20 and the papillitis was resolved. The timeline of the clinical course is shown in Figure 2 . 3. Discussion and Conclusions We report a case of bilateral ON, occurring two weeks after the patient's first dose of the ChAdOx1-S vaccine. The case involved the sudden onset of vision decline in both eyes with the typical presentation of ON: extraocular pain during eye movement. However, color vision remained normal; in this instance, the case differs from typical idiopathic ON due to intact color vision. This patient had no medical history, which confirmed the temporal relationship between the vaccine and ON. Most cases of vaccine-related ON occur 2 to 3 weeks after vaccination. Some patients' vision decline is subtle, while others' is dramatic, depending on how much of the optic nerve is involved. ON is a known but rare adverse side effect of vaccination, the exact mechanism of which is not clear. Possibly, the vaccine's activation of the host immune system leads to T-cell activation, which then damages the optic nerve's myelin sheath [ 16 ]. According to previous literature, the incidence of optic neuritis is 2.6–3.9 per 10,000 adult persons [ 17 ] and 1.79–2.46 per 100,000 persons in the pediatric population [ 18 ]. The incidence of vaccine-related optic neuritis has not been reported because neuro-ophthalmic presentations related to vaccination are extremely rare [ 19 ]. In the case of a relatively new vaccine, such as the COVID-19 vaccine, any reports of adverse effects after vaccination should be collected to comprehensively assess side effects. Case reports and case series related to post-vaccine optic neuritis have been published recently because more physicians have become aware of this possible side effect [ 8 , 9 , 14 , 20 , 21 , 22 , 23 , 24 , 25 ]. However, since the causative relationship between vaccination and idiopathic optic neuritis is difficult to delineate, in the case of our patient, we were clinically left with only the temporal association since she also had no medical history, and the most common period for vaccine-related immune reactions is two to three weeks after inoculation [ 26 ]. In Taiwan, five vaccines have been issued: ChAdOx1-S (AstraZeneca, AZ), BNT162b2 (BNT), mRNA-1273 (Moderna), Nuvaxoid (Novavax), and MVC-COV1901(Medigen). Since the first outbreak in May 2021, a large population has been vaccinated with the ChAdOx1-S (AstraZeneca) vaccine because this was the only vaccine our government could offer at that time. As more vaccines were used and lower-aged groups were vaccinated, more information on post-vaccine adverse effects was reported. In a report dated 19 June 2021, the Taiwan Central Epidemic Command Center (CECC) revealed that a total of 1,374,956 AstraZeneca shots had been administered in Taiwan up to 18 June, of which 269,056 were given to the elderly, aged 75 years and older. To date, the Vaccine Adverse Event Reporting System (VAERS) has registered 49 reports of death following ChAdOx1-S vaccination. At the time of writing this manuscript, the detailed VAERS report could be accessed through the website of the Taiwan Center for Disease Control ( https://www.cdc.gov.tw/en/Disease/SubIndex/ (accessed on 28 September 2022)) ( Table 1 ). Recently, two cases of acute zonal occult outer retinopathy (AZOOR) and bilateral arteritic anterior ischemic optic neuropathy (AAION) after COVID-19 m-RNA vaccination were reported [ 27 ]. Four cases of post-vaccine ON were reported in a case series, three of which were unilateral ON and the remaining one bilateral ON. However, these four cases regained their vision after prompt treatment [ 14 ]. ON may result from autoimmune, infectious, or inflammatory disorders. ON is one of the most commonly seen neurological adverse effects after vaccines, including measles, influenza, hepatitis A/B, pneumococcal vaccine, human papilloma virus, and rabies vaccines [ 3 , 4 , 5 , 6 , 28 , 29 , 30 , 31 , 32 , 33 , 34 , 35 , 36 , 37 , 38 , 39 ]. The mechanisms for these inflammatory/autoimmune events are mostly unknown, but some pathways are presumed. First, molecular mimicking, based on COVID-19-related autoimmune reactions to viral proteins resulting from mRNA translation, which induce immune interactions with human cells. Second, some vaccine adjuvants may activate the NLRP3 inflammasome, leading to inflammation and immunity [ 40 ]. Our case had bilateral ON temporally associated with ChAdOx1-S (AstraZeneca) vaccination. The patient responded well to corticosteroid therapy. Her ON and visual acuity recovered soon after the treatment. However, given the demyelinating nature of ON, other demyelination diseases, such as NMOSD, MOG-ON, or MS, might occur in the future. Patients with post-vaccine ON should be educated to become aware of any dangerous signs/symptoms of demyelination disorders for the patient still has the possibility to develop any of the above kinds of demyelination disorders [ 41 , 42 ].

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4857645/

Evaluation of a Tabletop Emergency Preparedness Exercise for Pharmacy Students

Objective. To describe the implementation and effect of an emergency preparedness laboratory activity on student knowledge, willingness to participate in emergency preparedness training, current level of preparedness, and the importance of a pharmacist's role in disaster response. Design. Second-year pharmacy students in the infectious disease module participated in a laboratory activity based on a basic disaster response tabletop exercise format. Three case-based scenarios involving infectious diseases were created by participating faculty members. Assessment. Surveys before and after the laboratory were used to assess the activity's effect on student knowledge, willingness to participate in emergency preparedness training, current level of preparedness, and the importance of a pharmacist's role in disaster response. In addition, the postsurvey assessed student perceptions of the activity's success at accomplishing faculty-specified outcomes from Appendix B of the Accreditation Council for Pharmacy Education's (ACPE) Standards. Conclusion. Implementation of an emergency response laboratory activity may improve overall students' knowledge of, confidence in, and understanding of their role as pharmacists in an emergency response, while incorporating a variety of skills and knowledge outcomes.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10654918/

Unconventional structure and mechanisms for membrane interaction and translocation of the NF-κB-targeting toxin AIP56

Bacterial AB toxins are secreted key virulence factors that are internalized by target cells through receptor-mediated endocytosis, translocating their enzymatic domain to the cytosol from endosomes (short-trip) or the endoplasmic reticulum (long-trip). To accomplish this, bacterial AB toxins evolved a multidomain structure organized into either a single polypeptide chain or non-covalently associated polypeptide chains. The prototypical short-trip single-chain toxin is characterized by a receptor-binding domain that confers cellular specificity and a translocation domain responsible for pore formation whereby the catalytic domain translocates to the cytosol in an endosomal acidification-dependent way. In this work, the determination of the three-dimensional structure of AIP56 shows that, instead of a two-domain organization suggested by previous studies, AIP56 has three-domains: a non-LEE encoded effector C (NleC)-like catalytic domain associated with a small middle domain that contains the linker-peptide, followed by the receptor-binding domain. In contrast to prototypical single-chain AB toxins, AIP56 does not comprise a typical structurally complex translocation domain; instead, the elements involved in translocation are scattered across its domains. Thus, the catalytic domain contains a helical hairpin that serves as a molecular switch for triggering the conformational changes necessary for membrane insertion only upon endosomal acidification, whereas the middle and receptor-binding domains are required for pore formation. Introduction Bacterial toxins are crucial virulence factors that lead to death or dysfunction of target cells, greatly contributing to the pathology in the host organism. Many toxins act on intracellular targets and are either injected into the cytosol by specialized protein secretion systems 1 or contain the components that allow them to reach the cytosol autonomously, as is the case with AB toxins 2 , 3 . Indeed, AB toxins have a modular multidomain structure typically organized in two distinct components: (i) component A, displaying enzymatic a ctivity and targeting a crucial eukaryotic cytosolic factor; and (ii) component B, capable of b inding to receptor(s) and conferring cellular specificity. The components can be encoded by one gene, originating a single-chain toxin, or by independent genes, resulting in multicomponent toxins with variable stoichiometry (AB 2 , AB 5 , A 2 B 5 , AB 7/8 ) 3 , 4 . AB toxins are internalized by the target cell through receptor-mediated endocytosis and their catalytic domains reach the cytosol of the cell either directly from the endosomal compartment upon acidic pH-triggered unfolding (short-trip toxins) or from the endoplasmic reticulum (ER), after retrograde transport from the endosomal compartment to the Golgi and from there to the ER (long-trip toxins) 3 – 5 . In short-trip toxins, the B component usually includes a receptor-binding domain/region and a translocation domain/region, the latter responsible for the formation of a pore through which the catalytic component passes into the cytosol 3 , 4 . Of specific interest to this work are the short-trip single-chain toxins, whose prototypical domain organization is exemplified by diphtheria toxin (DT), with a catalytic domain-containing A-component (DTa) and a B component formed by a translocation domain and a receptor-binding domain 3 , 4 , 6 – 8 . The catalytic and translocation domains are normally linked by a disulfide bridge between cysteine residues located at the ends of a short peptide, creating a loop called linker peptide or simply linker 3 , 4 . These toxins are usually secreted in an inactive form, often becoming active upon proteolytic nicking at the linker by a bacterial or host protease. Proteolytic nicking divides the toxin into components A (light chain) and B (heavy chain), which are held together by the disulfide bridge 3 , 4 . After translocation, the reducing environment of the cytosol disrupts the disulfide bond, releasing the catalytic domain. Some toxins, such as Cytotoxic Necrotizing Factors (CNFs) 9 , 10 , in which the disulfide bond is absent, contain a linker peptide between the A and B components that only becomes amenable to proteolytic cleavage after endosomal acidic pH-induced unfolding of the toxin. Moreover, there are other variations to the prototypical organization, as well as to the size and structural composition of the domains in some single-chain toxins, as exemplified by the large clostridial toxins, whose translocation and receptor binding is mediated by the same domain 11 – 15 . Apoptosis‐inducing protein of 56 kDa (AIP56) is a major virulence factor of Photobacterium damselae subsp. piscicida ( Phdp ) 16 , a Gram-negative bacterium that infects warm water marine fish with high economic importance for aquaculture 17 . AIP56 is a short-trip single-chain AB toxin 18 , 19 secreted by the type II secretion system of Phdp 20 . It has zinc-metalloprotease activity towards NF-kB p65, and targets fish phagocytes leading to their elimination by post-apoptotic secondary necrosis 16 , 18 , 21 , 22 . Ex-vivo experiments have shown that mouse macrophages also undergo NF-kB p65 inactivation and apoptosis following intoxication with AIP56 19 , suggesting that the toxin recognizes a phylogenetically conserved, but yet unidentified, receptor(s). It has been previously shown that AIP56 is endocytosed by a clathrin- and dynamin-dependent process and undergoes acidic pH-dependent translocation from the endosomes to the cytosol, through a process assisted by Hsp90 and cyclophilin A/D 19 , 23 . In previous studies 18 , amino acid sequence and limited proteolysis analyses suggested a two-domain organization for AIP56: (i) an A domain homologous to NleC (non-LEE encoded effector C), a type III effector with enzymatic activity towards NF-kB p65 that is injected into the cytosol by the type III secretion system of several enteric bacteria associated with human diseases 24 – 26 ; and (ii) a receptor-binding B domain, homologous to Protein D from bacteriophage APSE2 ( Acyrthosiphon pisum secondary endosymbiont 2) 27 – 30 . The two domains of AIP56 would be connected by an unusually long linker peptide (35 amino acids), the extremities of which are connected by a disulfide bridge required for intoxication 31 . However, contrary to most single-chain toxins 3 , 4 , AIP56 does not require proteolytic cleavage (nicking) to become catalytically active. Importantly, nicking of its linker peptide abrogates toxicity, but not toxin internalization, suggesting that an intact linker is required for the translocation step 18 . Genes encoding AIP56-like toxins are also found in the genomes of Candidatus Symbiopectobacterium, Arsenophonus nasoniae , Shewanella psycrophila and of several Vibrio species, the latter including strains isolated from humans 32 . In addition, several genes encoding putative toxins containing a domain similar to either the catalytic domain or the receptor-binding domain of AIP56 (AIP56-related toxins) are also present in several prokaryotes and eukaryotes genomes 33 . In this work, the three-dimensional structure of AIP56 was determined using X-ray crystallography and the elements involved in acidic pH-induced conformational changes, membrane interaction and translocation were characterized. Results Three-dimensional structure of AIP56 The three-dimensional structure of AIP56 was determined at 2.5 à resolution (data collection and refinement statistics are summarized in Supplementary Table 1 ). The crystallographic asymmetric unit contains four copies of AIP56 (chains A to D; root mean square deviation [rmsd] of 0.3–0.4 à over 437–452 aligned Cα atoms), but the extent of the inter-protomer contacts is compatible with a monomeric organization, in agreement with small-angle X-ray scattering (SAXS) data (see below). The final crystallographic model of AIP56 (Fig. 1a ) does not include residues 290-295 (ND1), 362-385 (ND2), 422-429 (ND3) and 454–455 (ND4), which are poorly defined in the electron density maps. The missing regions (ND1-4) were added with Modeller 34 based on the best model (out of 5) generated with AlphaFold2_advanced 35 , 36 (Supplementary Fig. 1a ). Fig. 1 Three-dimensional structure of AIP56. a Cartoon representation of the AIP56 monomer (Chain A), with domains colored wheat (catalytic domain), blue (middle domain) or green (receptor-binding domain). The active site residues are shown as red sticks, the disulfide bond (C262 and C298) as magenta sticks and the zinc ion as a black sphere. The regions not defined (ND) in the structure are indicated by arrows or represented by a dashed line. The N- (Nter) and C-termini (C-ter) are labeled. b SAXS profile showing adjustment of the curve (red line) calculated from the structural model of AIP56 obtained by the Modeller program (Supplementary Fig. 1a ), as calculated in CRYSOL, to the experimental scattering curve (gray dots) of AIP56. c SAXS profile showing better fitting of the SREFLEX model (red line). The bottom panels in (b) and (c) show the residual plots for the respective fits, with the residuals defined as Δ/σ = [Iexp(q) – Icalc(q)]/σexp(q). χ 2 scores were calculated in CRYSOL. d Complete SREFLEX simulated movement after the refinement stage. Vectors (red arrows) are drawn connecting the equivalent residues from the crystal structure (colored as in a) to the SREFLEX model (gray). e Cartoon representation of the superposed catalytic domains of AIP56 (wheat) and NleC (gray; PDB: 4Q3J ). The modified Ψ-loop β-sheet motif (β1–β3) of AIP56 is colored blue. The N- (Nter) and C-termini (C-ter) and secondary structure elements are labeled. The inset shows a close-up of the active sites, including the zinc ion (Zn; black sphere) and the coordinating residues from AIP56 (red) and NleC (gray). f The linker peptide contacts both the catalytic and the receptor-binding domains. Top, cartoon representation highlighting the contact regions; Bottom, C-alpha trace of the contact region with contacts within a range of 4 à is represented. Top and Bottom representations colored as in ( a ). g The catalytic and receptor-binding domains contact directly through helix α2 of the catalytic domain. Top and Bottom representations colored as in ( a ). The three-dimensional structure of AIP56 (Fig. 1a and Supplementary Fig. 1a ) shows a three-domain organization: (i) a N-terminal NleC-like catalytic domain (N1-F255) with a conserved zinc-binding motif (H165ExxH); (ii) a middle domain (G256-E307), with a disulfide bond (C262 and C298) flanking the 35 amino acid long linker peptide (S263 to E297), packed against the catalytic domain; and (iii) a C-terminal domain (P308-N497) previously shown to be involved in receptor binding 18 . SAXS analysis of AIP56 (Fig. 1b, c, d , Supplementary Table 2 and Supplementary Fig. 1b ) revealed a reasonable fit between the experimental scattering data and the theoretical scattering profile calculated with CRYSOL 37 for the structural model of monomeric AIP56 obtained with Modeller ( χ 2 = 5.05) (Fig. 1b , upper panel), which is considerably worse when using the crystal structure of AIP56 without the missing linkers (ND1-4, χ 2 = 39). The monomeric nature of AIP56 in solution is confirmed by the remarkable agreement between the sequence-derived molecular mass and experimental estimates (Supplementary Table 2 and Supplementary Fig. 1b ). The corresponding residual plot showed non-random features that appeared to worsen at a q range below 0.2 à −1 (Fig. 1b , bottom panel). Using normal mode analysis (NMA) with SREFLEX 38 it was possible to model the interdomain conformational flexibility of AIP56, therefore improving the overall fit with the experimental scattering data ( χ 2 = 1.20) and a much improved residual plot (Fig. 1c, d ). The experimental AIP56 structure was compared with the structures of 19 putative AIP56-like toxins (Supplementary Table 3 ) predicted by AlphaFold2_advanced 35 , 36 . All structures are highly similar, with major differences observed only in the middle domains, particularly in the linker region (Supplementary Fig. 1c ). Moreover, despite the high aminoacidic (Supplementary Fig. 2b ) and structural (Supplementary Fig. 1c ) conservation of the catalytic and receptor-binding domains, the AlphaFold-generated structures for A. nasoniae toxins resulted in distinct relative positions of those domains. Whether this results from inaccurate predictions of the middle domains by AlphaFold2 or corresponds to the real positions of the catalytic and receptor-binding domains in A. nasoniae toxins remains to be investigated. While BLAST searches retrieve many genes encoding domains homologous to the catalytic and/or receptor-binding domains of AIP56, genes encoding domains homologous to just the middle domain are not found. Together with structural data and primary structure analysis (Supplementary Fig. 2 ), this suggests that AIP56 and AIP56-like toxins arose from the fusion of a gene homologous to nleC 24 , 25 , encoding the catalytic domain, with a gene homologous to the one encoding Protein D from bacteriophage APSE2 27 – 29 , which codes for the receptor-binding domain 18 , with the middle domain likely evolving from a gene/intergenic region of the recipient genome, where those genes were incorporated. Alternatively, the middle domain evolved also as an independent gene, but homologous genes are not yet available in the databases or are difficult to detect due to high sequence divergence. Detailed analysis of AIP56 structure The catalytic domain of AIP56 has an overall Zincin fold similar to that of NleC (PDB entry 4Q3J ; rmsd of 1.4 à for 257 aligned Cα atoms), with three β-strands forming a modified Ψ-loop β-sheet motif (β1-β3) overlaid on 9 α-helices (Fig. 1e ) 24 . The active-site zinc ion of AIP56 is coordinated by H165 and H169 from helix α6 and by D176 located at the loop between helices α6 and α7. In two of the AIP56 monomers of the asymmetric unit, Y207 from the loop between helices α7 and α8 is the fourth zinc ligand, similar to what is observed in NleC 24 . The conserved E166 is ideally placed to promote the nucleophilic attack. As described before for NleC 24 , the dimensions and shape of the active-site cleft of AIP56, as well as the pattern of glutamate and aspartate side chains along the ridges of the cleft, mimic the DNA backbone phosphates to which NF-kB p65 binds 39 – 41 (Supplementary Fig. 3a ). In NleC, four of the acidic residues in the cleft ridges - E115, E118, D139 and E150 - were shown to be important for efficient proteolysis 24 . Of these, E115 and E150, which are conserved in AIP56 (E96 and E131, respectively) (Supplementary Fig. 3a ), are the most relevant 24 . This is probably also the case for AIP56 given that neither E118 nor D139 from NleC are conserved in AIP56 (the corresponding residues are N99 and N120, respectively) but, despite this, AIP56 cleaves human p65 in vitro more efficiently than NleC (Supplementary Fig. 3b ). The middle domain of AIP56 is composed of one α-helix (α10), two 3 10 helices (η5 and η6) and two short antiparallel β-strands (β4-β5) (Fig. 1a and Supplementary Fig. 1a ). Two cysteine residues, one in helix α10 and the other in strand β5, form a disulfide bond previously shown to be important for AIP56 toxicity 18 , 31 . These two cysteines flank the partially structured linker peptide that comprises helix η5 and strand β4 plus a connecting 15 residue-long unstructured loop (P270-A285) that contains an aspartate-rich patch at its tip (D274HDDD278) (Supplementary Fig. 2 ). The linker peptide establishes contacts with both the preceding catalytic and the succeeding receptor-binding domain (Fig. 1f ). AIP56's receptor-binding domain 18 , comprises a central antiparallel seven-stranded β-sheet (β7–β13) sandwiched by helices α11, α12, α15 and α16 on one side, and helices α13, α14 and α17 on the other (Fig. 1a and Supplementary Fig. 1a ). Two additional short β-strands (β6 and β14) form a small β-sheet perpendicular to the large one. Despite its simple architecture, the receptor-binding domain of AIP56 has an unusually high content (84 out of 190) of aromatic (11 tyrosine, 8 tryptophan and 14 phenylalanine) and hydrophobic (11 alanine, 9 isoleucine, 15 leucine and 16 glycine) residues. Although the central twisted antiparallel β-sheet of the AIP56's receptor-binding domain is similar to that of other unrelated proteins (Supplementary Fig. 3c ), the 7 α-helices that surround the central β-sheet have a unique organization. Although this domain shows some flexibility with respect to the catalytic and middle domains (Fig. 1d ), it establishes a series of contacts with helix α2 of the catalytic domain that prevent large structural movements (Fig. 1g ). Pore formation requires both the middle and receptor-binding domains Many short-trip single-chain toxins exhibit a three-domain organization, where the middle domain is dedicated to pore formation and facilitates the translocation of the catalytic domain into the cytosol 2 , 3 . This raises the question of whether the middle domain of AIP56, despite its structural simplicity, could mediate pore formation. The structural organization of AIP56, together with previous results showing that a chimera comprising β-lactamase (Bla) fused to the middle and receptor-binding domains of AIP56 (Bla L19-W286 AIP56 L258-N497 ) (Fig. 2a ) was able to deliver Bla into the cytosol of mouse bone marrow-derived macrophages (mBMDM) 23 support this concept. To further test this, several truncated versions of AP56 were used in experiments with black lipid bilayers (Fig. 2a ), namely: the catalytic domain alone (AIP56 N1-G256 ), the catalytic domain with the middle domain (AIP56 N1-E307 ), the middle domain with the receptor-binding domain (AIP56 L258-N497 ) and the receptor-binding domain alone (AIP56 T299-N497 ). Of these, only the construct containing both the middle and receptor-binding domains (AIP56 L258-N497 ) displayed membrane-interacting activity after acidification (pH 4.8–5.0) of the cis-side of the membrane, although the observed activity was different and much lower than that obtained with intact AIP56 (Fig. 2b ). In agreement, the Bla L19-W286 AIP56 L258-N497 chimera interacted with black lipid membranes, whereas a chimera with the AIP56 receptor-binding domain replaced by the DT receptor-binding (DTR) domain (AIP56 N1-E307 DTR S406-S560 ) did not display membrane activity (Fig. 2b ) and was unable to deliver the AIP56 catalytic domain into the cytosol of U-2 OS cells (Supplementary Fig. 4a, b ). Altogether, the black lipid bilayer experiments suggest that the interaction of AIP56 with the membrane requires both the middle and receptor-binding domains. Fig. 2 Pore formation requires both the middle and receptor-binding domains. a Schematic representation of AIP56 variants and chimeric proteins. AIP56 domains are colored as in Fig. 1a ; β-lactamase (Bla) and diphtheria receptor-binding (DTR) domain, gray. b Only AIP56 L258-N497 and Bla L19-W286 AIP56 L258-N497 interacted with artificial black lipid membranes. Single-channel record of DiPhPC/n‐decane membranes after addition of the indicated proteins to the cis-side of the black lipid bilayer at a final concentration indicated in the figure. Measurements were performed with 50 mV (AIP56 L258-N497 ) or 150 mV (Bla L19-W286 AIP56 L258-N497 ) at room temperature. Membrane activity was induced by acidification (pH 4.8–5.0; arrows) of the aqueous phase at the cis‐side of the chamber with exception for Bla L19-W286 AIP56 L258-N497 , which formed stable pores at pH 6. The average single-channel conductance was about 16 pS for 110 steps. Each result shown is representative of at least three ( n = 3) independent measurements. c Pharmacological inhibition of vacuolar ATPase pump does not affect cytosolic delivery of Bla by Bla L19-W286 AIP56 L258-N497 . The cleaved/uncleaved CCF4-AM ratios were determined by quantifying the indicated number of microscopic fields per condition. Representative images used for quantification are in Supplementary Fig. 4d . Results shown represent one out of three ( n = 3) independent experiments. Statistical significance was tested by One-way ANOVA and p values for individual comparisons were calculated by Tukey's HSD test and indicated on top of the brackets. Data are presented as mean values ± SD. Actual p values from left-to-right: p  0.99, p = 0.005, p = 0.90, p = 0.81; for pH 4.5: p = 0.02, p > 0.99, p > 0.99, p > 0.99, p = 0.78, p = 0.36; ns non-significant. Open (without concanamycin A) or closed (with concanamycin A) symbols as well as color coding have been added to facilitate the reading of the experimental conditions, as specified below the graph. Samples were derived from the same experiment and the blots processed in parallel. NF-kB p65, nuclear factor kappa-light-chain-enhancer of activated B cells subunit p65; ConcA, concanamycin A (black); AIP56, purple, AIP56 D274S/D276-278S , red; AIP56 D274N/D276-278N , green. c AIP56 D274S/D276-278S and AIP56 D274N/D276-278N retained the ability to interact with black lipid bilayers. Proteins were used at a final concentration of 14 nM. Membrane activity was induced by acidification (pH 4.8; red arrows) of the aqueous phase at the cis-side of the chamber. Each result shown is representative of at least three ( n = 3) independent measurements. Source data for ( b ) and ( c ) are provided in the Source data file. Three-dimensional structure of AIP56 The three-dimensional structure of AIP56 was determined at 2.5 à resolution (data collection and refinement statistics are summarized in Supplementary Table 1 ). The crystallographic asymmetric unit contains four copies of AIP56 (chains A to D; root mean square deviation [rmsd] of 0.3–0.4 à over 437–452 aligned Cα atoms), but the extent of the inter-protomer contacts is compatible with a monomeric organization, in agreement with small-angle X-ray scattering (SAXS) data (see below). The final crystallographic model of AIP56 (Fig. 1a ) does not include residues 290-295 (ND1), 362-385 (ND2), 422-429 (ND3) and 454–455 (ND4), which are poorly defined in the electron density maps. The missing regions (ND1-4) were added with Modeller 34 based on the best model (out of 5) generated with AlphaFold2_advanced 35 , 36 (Supplementary Fig. 1a ). Fig. 1 Three-dimensional structure of AIP56. a Cartoon representation of the AIP56 monomer (Chain A), with domains colored wheat (catalytic domain), blue (middle domain) or green (receptor-binding domain). The active site residues are shown as red sticks, the disulfide bond (C262 and C298) as magenta sticks and the zinc ion as a black sphere. The regions not defined (ND) in the structure are indicated by arrows or represented by a dashed line. The N- (Nter) and C-termini (C-ter) are labeled. b SAXS profile showing adjustment of the curve (red line) calculated from the structural model of AIP56 obtained by the Modeller program (Supplementary Fig. 1a ), as calculated in CRYSOL, to the experimental scattering curve (gray dots) of AIP56. c SAXS profile showing better fitting of the SREFLEX model (red line). The bottom panels in (b) and (c) show the residual plots for the respective fits, with the residuals defined as Δ/σ = [Iexp(q) – Icalc(q)]/σexp(q). χ 2 scores were calculated in CRYSOL. d Complete SREFLEX simulated movement after the refinement stage. Vectors (red arrows) are drawn connecting the equivalent residues from the crystal structure (colored as in a) to the SREFLEX model (gray). e Cartoon representation of the superposed catalytic domains of AIP56 (wheat) and NleC (gray; PDB: 4Q3J ). The modified Ψ-loop β-sheet motif (β1–β3) of AIP56 is colored blue. The N- (Nter) and C-termini (C-ter) and secondary structure elements are labeled. The inset shows a close-up of the active sites, including the zinc ion (Zn; black sphere) and the coordinating residues from AIP56 (red) and NleC (gray). f The linker peptide contacts both the catalytic and the receptor-binding domains. Top, cartoon representation highlighting the contact regions; Bottom, C-alpha trace of the contact region with contacts within a range of 4 à is represented. Top and Bottom representations colored as in ( a ). g The catalytic and receptor-binding domains contact directly through helix α2 of the catalytic domain. Top and Bottom representations colored as in ( a ). The three-dimensional structure of AIP56 (Fig. 1a and Supplementary Fig. 1a ) shows a three-domain organization: (i) a N-terminal NleC-like catalytic domain (N1-F255) with a conserved zinc-binding motif (H165ExxH); (ii) a middle domain (G256-E307), with a disulfide bond (C262 and C298) flanking the 35 amino acid long linker peptide (S263 to E297), packed against the catalytic domain; and (iii) a C-terminal domain (P308-N497) previously shown to be involved in receptor binding 18 . SAXS analysis of AIP56 (Fig. 1b, c, d , Supplementary Table 2 and Supplementary Fig. 1b ) revealed a reasonable fit between the experimental scattering data and the theoretical scattering profile calculated with CRYSOL 37 for the structural model of monomeric AIP56 obtained with Modeller ( χ 2 = 5.05) (Fig. 1b , upper panel), which is considerably worse when using the crystal structure of AIP56 without the missing linkers (ND1-4, χ 2 = 39). The monomeric nature of AIP56 in solution is confirmed by the remarkable agreement between the sequence-derived molecular mass and experimental estimates (Supplementary Table 2 and Supplementary Fig. 1b ). The corresponding residual plot showed non-random features that appeared to worsen at a q range below 0.2 à −1 (Fig. 1b , bottom panel). Using normal mode analysis (NMA) with SREFLEX 38 it was possible to model the interdomain conformational flexibility of AIP56, therefore improving the overall fit with the experimental scattering data ( χ 2 = 1.20) and a much improved residual plot (Fig. 1c, d ). The experimental AIP56 structure was compared with the structures of 19 putative AIP56-like toxins (Supplementary Table 3 ) predicted by AlphaFold2_advanced 35 , 36 . All structures are highly similar, with major differences observed only in the middle domains, particularly in the linker region (Supplementary Fig. 1c ). Moreover, despite the high aminoacidic (Supplementary Fig. 2b ) and structural (Supplementary Fig. 1c ) conservation of the catalytic and receptor-binding domains, the AlphaFold-generated structures for A. nasoniae toxins resulted in distinct relative positions of those domains. Whether this results from inaccurate predictions of the middle domains by AlphaFold2 or corresponds to the real positions of the catalytic and receptor-binding domains in A. nasoniae toxins remains to be investigated. While BLAST searches retrieve many genes encoding domains homologous to the catalytic and/or receptor-binding domains of AIP56, genes encoding domains homologous to just the middle domain are not found. Together with structural data and primary structure analysis (Supplementary Fig. 2 ), this suggests that AIP56 and AIP56-like toxins arose from the fusion of a gene homologous to nleC 24 , 25 , encoding the catalytic domain, with a gene homologous to the one encoding Protein D from bacteriophage APSE2 27 – 29 , which codes for the receptor-binding domain 18 , with the middle domain likely evolving from a gene/intergenic region of the recipient genome, where those genes were incorporated. Alternatively, the middle domain evolved also as an independent gene, but homologous genes are not yet available in the databases or are difficult to detect due to high sequence divergence. Detailed analysis of AIP56 structure The catalytic domain of AIP56 has an overall Zincin fold similar to that of NleC (PDB entry 4Q3J ; rmsd of 1.4 à for 257 aligned Cα atoms), with three β-strands forming a modified Ψ-loop β-sheet motif (β1-β3) overlaid on 9 α-helices (Fig. 1e ) 24 . The active-site zinc ion of AIP56 is coordinated by H165 and H169 from helix α6 and by D176 located at the loop between helices α6 and α7. In two of the AIP56 monomers of the asymmetric unit, Y207 from the loop between helices α7 and α8 is the fourth zinc ligand, similar to what is observed in NleC 24 . The conserved E166 is ideally placed to promote the nucleophilic attack. As described before for NleC 24 , the dimensions and shape of the active-site cleft of AIP56, as well as the pattern of glutamate and aspartate side chains along the ridges of the cleft, mimic the DNA backbone phosphates to which NF-kB p65 binds 39 – 41 (Supplementary Fig. 3a ). In NleC, four of the acidic residues in the cleft ridges - E115, E118, D139 and E150 - were shown to be important for efficient proteolysis 24 . Of these, E115 and E150, which are conserved in AIP56 (E96 and E131, respectively) (Supplementary Fig. 3a ), are the most relevant 24 . This is probably also the case for AIP56 given that neither E118 nor D139 from NleC are conserved in AIP56 (the corresponding residues are N99 and N120, respectively) but, despite this, AIP56 cleaves human p65 in vitro more efficiently than NleC (Supplementary Fig. 3b ). The middle domain of AIP56 is composed of one α-helix (α10), two 3 10 helices (η5 and η6) and two short antiparallel β-strands (β4-β5) (Fig. 1a and Supplementary Fig. 1a ). Two cysteine residues, one in helix α10 and the other in strand β5, form a disulfide bond previously shown to be important for AIP56 toxicity 18 , 31 . These two cysteines flank the partially structured linker peptide that comprises helix η5 and strand β4 plus a connecting 15 residue-long unstructured loop (P270-A285) that contains an aspartate-rich patch at its tip (D274HDDD278) (Supplementary Fig. 2 ). The linker peptide establishes contacts with both the preceding catalytic and the succeeding receptor-binding domain (Fig. 1f ). AIP56's receptor-binding domain 18 , comprises a central antiparallel seven-stranded β-sheet (β7–β13) sandwiched by helices α11, α12, α15 and α16 on one side, and helices α13, α14 and α17 on the other (Fig. 1a and Supplementary Fig. 1a ). Two additional short β-strands (β6 and β14) form a small β-sheet perpendicular to the large one. Despite its simple architecture, the receptor-binding domain of AIP56 has an unusually high content (84 out of 190) of aromatic (11 tyrosine, 8 tryptophan and 14 phenylalanine) and hydrophobic (11 alanine, 9 isoleucine, 15 leucine and 16 glycine) residues. Although the central twisted antiparallel β-sheet of the AIP56's receptor-binding domain is similar to that of other unrelated proteins (Supplementary Fig. 3c ), the 7 α-helices that surround the central β-sheet have a unique organization. Although this domain shows some flexibility with respect to the catalytic and middle domains (Fig. 1d ), it establishes a series of contacts with helix α2 of the catalytic domain that prevent large structural movements (Fig. 1g ). Pore formation requires both the middle and receptor-binding domains Many short-trip single-chain toxins exhibit a three-domain organization, where the middle domain is dedicated to pore formation and facilitates the translocation of the catalytic domain into the cytosol 2 , 3 . This raises the question of whether the middle domain of AIP56, despite its structural simplicity, could mediate pore formation. The structural organization of AIP56, together with previous results showing that a chimera comprising β-lactamase (Bla) fused to the middle and receptor-binding domains of AIP56 (Bla L19-W286 AIP56 L258-N497 ) (Fig. 2a ) was able to deliver Bla into the cytosol of mouse bone marrow-derived macrophages (mBMDM) 23 support this concept. To further test this, several truncated versions of AP56 were used in experiments with black lipid bilayers (Fig. 2a ), namely: the catalytic domain alone (AIP56 N1-G256 ), the catalytic domain with the middle domain (AIP56 N1-E307 ), the middle domain with the receptor-binding domain (AIP56 L258-N497 ) and the receptor-binding domain alone (AIP56 T299-N497 ). Of these, only the construct containing both the middle and receptor-binding domains (AIP56 L258-N497 ) displayed membrane-interacting activity after acidification (pH 4.8–5.0) of the cis-side of the membrane, although the observed activity was different and much lower than that obtained with intact AIP56 (Fig. 2b ). In agreement, the Bla L19-W286 AIP56 L258-N497 chimera interacted with black lipid membranes, whereas a chimera with the AIP56 receptor-binding domain replaced by the DT receptor-binding (DTR) domain (AIP56 N1-E307 DTR S406-S560 ) did not display membrane activity (Fig. 2b ) and was unable to deliver the AIP56 catalytic domain into the cytosol of U-2 OS cells (Supplementary Fig. 4a, b ). Altogether, the black lipid bilayer experiments suggest that the interaction of AIP56 with the membrane requires both the middle and receptor-binding domains. Fig. 2 Pore formation requires both the middle and receptor-binding domains. a Schematic representation of AIP56 variants and chimeric proteins. AIP56 domains are colored as in Fig. 1a ; β-lactamase (Bla) and diphtheria receptor-binding (DTR) domain, gray. b Only AIP56 L258-N497 and Bla L19-W286 AIP56 L258-N497 interacted with artificial black lipid membranes. Single-channel record of DiPhPC/n‐decane membranes after addition of the indicated proteins to the cis-side of the black lipid bilayer at a final concentration indicated in the figure. Measurements were performed with 50 mV (AIP56 L258-N497 ) or 150 mV (Bla L19-W286 AIP56 L258-N497 ) at room temperature. Membrane activity was induced by acidification (pH 4.8–5.0; arrows) of the aqueous phase at the cis‐side of the chamber with exception for Bla L19-W286 AIP56 L258-N497 , which formed stable pores at pH 6. The average single-channel conductance was about 16 pS for 110 steps. Each result shown is representative of at least three ( n = 3) independent measurements. c Pharmacological inhibition of vacuolar ATPase pump does not affect cytosolic delivery of Bla by Bla L19-W286 AIP56 L258-N497 . The cleaved/uncleaved CCF4-AM ratios were determined by quantifying the indicated number of microscopic fields per condition. Representative images used for quantification are in Supplementary Fig. 4d . Results shown represent one out of three ( n = 3) independent experiments. Statistical significance was tested by One-way ANOVA and p values for individual comparisons were calculated by Tukey's HSD test and indicated on top of the brackets. Data are presented as mean values ± SD. Actual p values from left-to-right: p  0.99, p = 0.005, p = 0.90, p = 0.81; for pH 4.5: p = 0.02, p > 0.99, p > 0.99, p > 0.99, p = 0.78, p = 0.36; ns non-significant. Open (without concanamycin A) or closed (with concanamycin A) symbols as well as color coding have been added to facilitate the reading of the experimental conditions, as specified below the graph. Samples were derived from the same experiment and the blots processed in parallel. NF-kB p65, nuclear factor kappa-light-chain-enhancer of activated B cells subunit p65; ConcA, concanamycin A (black); AIP56, purple, AIP56 D274S/D276-278S , red; AIP56 D274N/D276-278N , green. c AIP56 D274S/D276-278S and AIP56 D274N/D276-278N retained the ability to interact with black lipid bilayers. Proteins were used at a final concentration of 14 nM. Membrane activity was induced by acidification (pH 4.8; red arrows) of the aqueous phase at the cis-side of the chamber. Each result shown is representative of at least three ( n = 3) independent measurements. Source data for ( b ) and ( c ) are provided in the Source data file. Discussion AIP56 is the only AB-type toxin characterized to date that targets NF-kB 18 . It has a general intoxication mechanism 19 similar to other short-trip single-chain AB toxins that after receptor-mediated endocytosis undergo acidic pH-dependent conformational changes resulting in the exposure of hydrophobic regions, followed by membrane interaction with pore formation and translocation of the catalytic domain from endosomes into the cytosol 3 , 4 . The mechanism of translocation is well described for several of those toxins including DT 7 , 46 – 48 , 50 , 56 , 73 , clostridial neurotoxins 78 – 82 , Pasteurella multocida toxin 49 and CNFs 9 , 10 , 83 , all containing a dedicated middle translocation domain 9 , 10 , 45 – 50 or, as in large clostridial toxins, a common translocation and receptor-binding domain 14 . However, AIP56 lacks a domain functionally and structurally equivalent to the translocation domain of those toxins. Instead, the small size and simple structure of the AIP56 middle domain apparently resulted in the spread of the structural elements involved in the AIP56 translocation process across all of its domains. Indeed, in AIP56, the pH-sensing residues that control the conformational changes required for triggering membrane interaction and translocation seem to be located at the carboxyl-terminal portion of the catalytic domain, in a helical hairpin that is not involved in pore formation. Thus, it is proposed that the pH-sensing residues prevent pore formation until the low pH conditions found in the endosome induce unfolding of the catalytic domain for translocation through the pore. Accordingly, the Bla L19-W286 AIP56 L258-N497 chimera (without hairpin) could not only form stable pores in the artificial lipid bilayers at a higher pH than AIP56 but also translocate Bla to the cytosol even when endosomal acidification was prevented by Concanamycin A. The results presented here also show that the middle and receptor-binding domains are both required for pore formation since: (i) interaction with black lipid bilayers was only observed with constructs that have both domains (AIP56 L258-N497 and Bla L19-W286 AIP56 L258-N497 ); (ii) the middle and receptor-binding domains of AIP56 were sufficient to deliver Bla into the cytosol (Bla L19-W286 AIP56 L258-N497 ); and (iii) disruption of the linker region within the middle domain 18 or replacing the AIP56 receptor-binding domain by the DT receptor-binding domain (AIP56 N1-E307 DTR S406-S560 ) abrogates cell toxicity, as monitored by p65 cleavage. The specific role each of these domains plays in pore formation requires further study. However, it is plausible that a region(s) within the receptor-binding domain promotes toxin oligomerization and the middle domain forms the transmembrane channel. Indeed, the small size and lack of structural complexity of the middle domain accords with the channel-forming regions that in β-PFTs 84 – 88 and anthrax PA 51 , 52 refold to a β-hairpin that participates in the formation of a β-barrel, being particularly similar to the aerolysin insertion loop (prestem loop or tongue) 86 , 88 (Supplementary Fig. 8a ). While in β-PFTs 84 – 88 and PA 51 , 52 the conformational alterations of the insertion regions is structurally restricted, with release and refolding to β-hairpin triggered by proteolytic- and/or pH-dependent activation, the linker region of AIP56 appears to be relatively flexible at physiological pH. Given the observed resistance of both the catalytic and receptor-binding domains of AIP56 to proteolytic cleavage, it can be proposed that after binding to its receptor(s) and endocytosis, acidification will trigger a structural alteration of the pH-sensing hairpin leading to increased overall flexibility of the toxin and reshaping the linker region into a β-hairpin which will be part of a β-barrel channel formed upon toxin oligomerization. In this scenario, although the disulfide bridge is required for the intoxication process of AIP56 31 , it is not clear whether it plays a role in pore formation or is simply required for maintaining the catalytic and receptor-binding domains close together. It should be noted that the involvement of a disulfide bridge on the release of the catalytic domain in the cytosolic reducing environment is not mandatory, as supported by the fact that CNFs lack a disulfide-bridged linker-peptide loop, but still contains a linker peptide between the catalytic and translocation domains that is cleaved after endosomal acidic pH-induced unfolding of the toxin 9 , 10 . In fact, in the short-trip single-chain toxins that contain a disulfide-bridged linker-peptide loop that requires cleavage, the linker peptide is short and located between the catalytic and translocation domains 3 , 4 , whereas in AIP56 the linker peptide is much longer, making up almost entirely the middle domain. Furthermore, its cleavage abrogates translocation but not toxin internalization 18 , supporting that it will be involved in the translocation process and may not be cleaved. Interestingly, the predicted structure of AIP56-related toxins exhibits a three-domain organization, including a known (e.g., CdtB-like domain) or predicted (e.g., presence of a zinc-metalloprotease binding motif) enzymatic domain, a structurally complex middle domain and a domain homologous to the receptor-binding domain of AIP56 (Supplementary Fig. 8b ). This three-domain organization resembles that of many short-trip single-chain toxins, such as DT 7 , clostridial neurotoxins 78 – 82 , P. multocida toxin 49 and Pseudomonas aeruginosa exotoxin A 89 , wherein their middle pore-forming/translocation domain is flanked by the catalytic and receptor-binding domains, which suggests that the middle domain in AIP56-related toxins and AIP56/AIP56-like toxins may be responsible for the formation of the transmembrane pore. Finally, it is possible that upon translocating to the cytosol, the catalytic domain interacts with Hsp90 through the region of its pH-sensing hairpin 19 , 23 . In summary, the structural and functional data described in this work provide new insights on AIP56 translocation mechanism and may contribute to develop prophylaxis and treatments based on AIP56. Future work may elucidate whether the conclusions drawn for AIP56 can be extended to the increasing number of AIP56-like and -related toxins. Methods Ethics statement This study was carried out in accordance with European and Portuguese legislation for the use of animals for scientific purposes (Directive 2010/63/EU; Decreto-Lei 113/2013). The work was approved by the ORBEA (Animal Welfare and Ethics Body) of i3S and was licensed by Direcção-Geral de Alimentação e Veterinária (DGAV), the Portuguese authority for animal protection (ref. 004933). Mice C57BL/6J mice were bred and housed (filter top cages, eurostandard type II, corn cob bedding material and nesting paper and plastic tube rolls for enrichment) at the animal facility of the Instituto de Investigação e Inovação em Saúde with 12 h light/12 h dark light cycle, 20–24 °C and 45–65% humidity. The mice were fed sterilized food (2014S, Envigo) and water ad libitum and were euthanized by CO 2 inhalation followed by cervical dislocation. Reagents and antibodies PMSF (P7626), pronase E from Streptomyces griseus (P5147), diphtheria toxin (DT) from Corynebacterium diphtheriae (D0564), Proteinase K from Engyodontium album (P2308), α-chymotrysin from bovine pancreas type II (C4129), 8-Anilino-1-naphthalenesulfonic acid (ANS, A1028) and concanamycin A (ConcA, C9705) were purchased from Sigma Aldrich. HBSS, HEPES, l -glutamine, sodium pyruvate and 17-(dimethylaminoethylamino)−17-demethoxygeldanamycin (17-DMAG, ant-dgl-5) were purchased from Invitrogen. DMEM (10938-025), heat-inactivated fetal bovine serum (FBS, 10500064) and penicillin/streptomycin (P/S, 15140122), all Gibco, and LiveBLAzer™ FRET-B/G Loading Kit with CCF4-AM (K1095) were purchased from Life Technologies. CCF4-AM is a lipophilic, esterified form of the CCF4 substrate, which is a Fluorescence Resonance Energy Transfer (FRET) substrate that consists of a cephalosporin core linking 7-hydroxycoumarin to fluorescein. Isopropyl β-D-1-thiogalactopyranoside (IPTG, MB026) was purchased from NZYTech. The anti-human NF-κB p65 C-terminal domain (clone c-20) rabbit polyclonal antibody (sc-372, dilution 1:3000) was from Santa Cruz Biotechnology and the anti-V5 (R960-25, dilution 1:5000) mouse monoclonal antibody was purchased from Invitrogen. Goat anti-IgG rabbit alkaline phosphatase conjugated secondary antibody (A9919, dilution 1:10000) and goat anti-IgG mouse alkaline phosphatase conjugated secondary antibody (A2429, dilution 1:10000) were purchased from Sigma Aldrich. Cells Mouse bone marrow-derived macrophages (mBMDM) were derived from bone marrow of femurs and tibias from 4–8 week-old C57BL/6J male mice, as previously described 90 . Briefly, femurs and tibias were flushed with 10 mL of Hanks' balanced salt solution (HBSS) and the cells recovered by centrifugation and resuspended in supplemented DMEM (DMEM with 10 mM glutamine, 10 mM HEPES, 1 mM sodium pyruvate, 10% (v/v) FBS and 1% penicillin/streptomycin) plus 10% (v/v) of L929 cell conditioned medium (LCCM) as a source of macrophage colony stimulating factor 91 . Fibroblasts were removed by culturing the cells overnight, at 37 °C in a humidified chamber in a 7% (v/v) CO 2 atmosphere, on a cell culture dish. The non-adherent cells were collected, centrifuged, resuspended in supplemented DMEM plus 10% (v/v) LCCM at a concentration of 5 × 10 5 cells mL −1 , plated in 24-well cell culture plates at a density of 5 × 10 5 cells/well and incubated as above. Three days later, 10% (v/v) LCCM was added and on the 7th day, the medium was renewed. Cells were used at day 10. To obtain LCCM, L929 cells were grown in 75 cm 2 filtered cap flasks in supplemented DMEM until reaching 100% confluence. Cells were then diluted 1:100 in fresh supplemented DMEM and incubated for 10 days at 37 °C, 7% (v/v) CO 2 . The supernatant was collected, pooled, centrifuged at 750 × g for 10 min and filtered. LCCM was aliquoted and stored at −20 °C until used. DNA Constructs for recombinant protein production Constructs encoding full-length AIP56 cloned into the NcoI/XhoI restriction sites of pET28a (Novagen) in frame with a C-terminal 6His-tag (pET28AIP56H+) or with a V5 plus 6xHis-tag (pET28AIP56V5H+) are described in ref. 16 and ref. 19 , respectively. The coding regions of AIP56 N1-G256 , AIP56 N1-E307 , AIP56 L258-N497 and AIP56 T299-N497 were PCR-amplified using pET28AIP56H+ as template and cloned into the NcoI/XhoI restriction sites of pET28a in frame with a 6xHis-tag at the C-terminus (AIP56 N1-G256 , AIP56 N1-E307 and AIP56 T299-N497 ) or N-terminus (AIP56 L258-N497 ). To obtain the constructs for expressing Bla-AIP56 chimeric proteins, the sequence encoding Bla L19-W286 was amplified from plasmid p327 (gifted by Dr. Dimitri Panagiotis Papatheodorou) and cloned into pET28a NcoI/SacI restriction sites, yielding plasmid pET28Bla L19-W286 . DNA sequences encoding AIP56 P210-N497 or AIP56 L258-N497 were then amplified using pET28AIP56H+ as template and ligated into the SacI/XhoI restriction sites of pET28Bla L19-W286 in frame with a C-terminal 6xHis-tag. To generate chimera AIP56 N1-E307 DTR S406-S560 , AIP56 N1-E307 encoding sequence was amplified from pET28AIP56H+ and cloned into pET28 NcoI/SacI restriction sites, yielding plasmid pET28AIP56 N1-E307 . DTR S406-S560 coding sequence (accession number WP_072564851.1 ) was then amplified from pET-22b DT 51E/148K (gifted by Dr. John Collier; Addgene plasmid # 11081 ; RRID:Addgene_11081) and ligated into the SacI/XhoI restriction sites of pET28AIP56 N1-E307 in frame with a C-terminal 6xHis-tag. Constructs coding for AIP56 E214K , AIP56 E218K , AIP56 H222K , AIP56 H231K , AIP56 E234K , AIP56 E214K/E218K/H222K , AIP56 H231K/E234K , AIP56 D274S/D276-278S and AIP56 D274N/D276-278N with or without V5-tag were generated with the QuickChange Site‐Directed Mutagenesis Kit (Stratagene, 200518) following the manufacturer's instructions, using pET28AIP56V5H+ 19 and pET28AIP56H+ 16 as template, respectively. Plasmids and primers used in this study are listed in Supplementary Table 5 . Production of recombinant proteins Full-length His-tagged AIP56 (AIP56), full-length V5 plus His-tagged AIP56 (AIP56-V5), AIP56 N1-G256 , AIP56 N1-E307 DTR S406-S560 and AIP56 mutant variants were expressed in Escherichia coli BL21(DE3), AIP56 N1-E307 in E. coli BL21-CodonPlus (DE3) and AIP56 T299-N497 in E. coli SoluBL21 (DE3). Bla-AIP56 chimeras and AIP56 L258-N497 were expressed in E. coli Rosetta (DE3). Transformed E. coli cells were cultured at 37 °C in 1 L of Luria Bertani (LB) broth with shaking (200 rpm), except for AIP56 N1-E307 DTR S406-S560 for which 6 L were used. At OD 600 ~ 0.6, 0.5 mM IPTG was added and protein expression carried out at 17 °C for 4 h in the case of Bla-AIP56 chimeras or for 20 h for all other proteins. Cells were harvested by centrifugation and resuspended in 30 mL of Buffer: (i) 50 mM Bis-Tris pH 6.5, 300 mM NaCl for AIP56, AIP56 N1-G256 , AIP56 T299-N497 and AIP56 mutant variants; (ii) 50 mM Tris pH 8.0, 300 mM NaCl for AIP56 L258-N497 , AIP56 N1-E307 and AIP56 N1-E307 DTR S406-S560 ; or (iii) 20 mM Tris pH 8.0, 200 mM NaCl, 5% (v/v) glycerol for Bla-AIP56 chimeras. In all cases, cell lysis was performed by sonication. After centrifugation, the recombinant proteins were purified from the supernatant using nickel‐affinity chromatography (Ni‐NTA agarose, ABT) followed by size-exclusion chromatography (Superose 12 10/300 GL, GE Healthcare), except in the cases of AIP56 N1-E307 and AIP56 N1-E307DTR in which only nickel‐affinity chromatography was performed. Recombinant proteins were analyzed by SDS-PAGE and purities determined by densitometry of Coomassie Blue-stained gels (Image Lab Software, BioRad). Protein batches used in this work were ≥90% pure. Determination of protein concentration The concentrations of recombinant proteins were determined by measuring absorbance at 280 nm using a NanoDrop 1000 and/or NanoDrop One (Thermo Fisher Scientific) considering the extinction coefficient and the molecular weight calculated with the ProtParam tool ( http://www.expasy.org/tools/protparam.html ). Crystallization Initial crystallization hits for AIP56 were identified by high-throughput screening at the HTX Lab of the EMBL Grenoble Outstation (Grenoble, France). AIP56 crystals were obtained by mixing equal volumes of protein (7 mg mL −1 ) and crystallization solution (400 mM sodium acetate, 100 mM Tris pH 8.5, 15% (w/v) PEG 4 K, 10 mM taurine), which was optimized using an additive screen kit (HR2-428; Hampton Research). Prior to data collection, crystals were successively transferred into the crystallization solution supplemented with 15% and 30% (v/v) glycerol, followed by flash-cooling in liquid nitrogen. Structure determination Several X-ray diffraction datasets were collected from AIP56 crystals at Synchrotron SOLEIL (Saint-Aubin, France; beamlines Proxima-1 and Proxima-2) and at the European Synchrotron Radiation Facility (ESRF, Grenoble, France; beamlines ID23, ID29 and ID30). The best dataset was collected on beamline ID29 on a Pilatus 6 M (Dectris) detector (2700 images, 0.1° rotation, 0.02 s exposure, wavelength 0.972 à ), indexed and integrated with XDS 92 and analyzed using the STARANISO server as diffraction was highly anisotropic 93 . Statistics for data processing are summarized in Supplementary Table 1 . The structure of AIP56 was solved by molecular replacement with PHASER 94 using an AIP56 model generated by the artificial intelligence program AlphaFold2_advanced 35 , 36 . Several cycles of model building and refinement were performed using COOT, Phenix.refine and BUSTER 95 – 99 . All illustrations of macromolecular models were produced with PyMOL 100 . The final refined coordinates and structure factors were deposited at the Protein Data Bank under PDB entry 7ZPF and the corresponding diffraction images at the SBGrid Data Bank (10.15785/SBGRID/911). Small-angle X-ray scattering (SAXS) data acquisition, processing and analysis X-ray scattering data were collected at the SWING beamline of the SOLEIL Synchrotron (Saint-Aubin, France). Measurements were performed using a HPLC Bio Sec3 (Agilent) size-exclusion column, online with the SAXS measuring cell, a 1.5 mm diameter quartz capillary contained in an evacuated vessel. The sample-to-detector (Dectris Eiger 4 M) distance was set to 2000 mm and the wavelength λ to 1.0 à , allowing useful data collection over the momentum transfer range of 0.005 à −1 < q < 0.5 à −1 ( q = 4π sin( θ )/ λ ). SAXS data were collected directly after elution of the AIP56 protein through the HPLC column equilibrated in 50 mM HEPES buffer pH 7.5, 500 mM NaCl. Fifty µl of protein sample were injected at 15 °C and two different initial concentrations, 13.4 and 3.2 mg mL −1 , in order to obtain a curve with a good statistic and devoid of interparticle effects. The flow rate was 0.2 mL min −1 , frame duration was 0.99 sec and the dead time between frames was 0.01 sec. Scattering of the elution buffer before void volume was recorded and used as buffer scattering for subtraction from all protein patterns. The scattered intensities were displayed on an absolute scale using the scattering by water. Data were first analyzed using Foxtrot and then using the US-SOMO HPLC module 101 . This program provided for each SAXS frame the values of the scattering intensity I(0) and of the radius of gyration Rg by applying the Guinier analysis together with a calculation of the approximate molar mass using the Rambo and Tainer approach 102 . Identical frames under the main elution peak were selected using Cormap 103 and averaged for further analysis. Both averaged curves corresponding to the two initial concentrations were merged. Data was deposit in the Small-Angle Scattering Biological Data Bank under accession code SASDNW6 . Intoxication assays mBMDM monolayers seeded in 24-well plates (5 × 10 5 cells/well) were intoxicated by continuous incubation with 174 nM AIP56 or AIP56 variants, diluted in culture medium (supplemented DMEM + 10% (v/v) LCCM) at 37 °C for 4 h. The NF-κB p65 level was evaluated by western blotting after 2 h incubation with the toxin and was used as readout for arrival of AIP56 into the cytosol. Fluorescence Resonance Energy Transfer (FRET) based assay mBMDM cells cultured at a density of 1.5 × 10 5 cells/well in 8-well plates (ibidi) were pre-treated with concanamycin A (ConcA; 10 nM), 17-DMAG (20 µM) or their vehicles (DMSO for ConcA and water for 17-DMAG) in supplemented DMEM without FBS for 1 h at 37 °C. Stock solutions (250 µM ConcA in DMSO and 8.1 mM 17-DMAG in water) were diluted with culture medium prior to the experiment. Next, the medium was replaced by fresh medium containing the inhibitors and Bla L19-W286 AIP56 P210-N497 or Bla L19-W286 AIP56 L258-N497 was added at a final concentration of 525 nM. Untreated cells and cells treated with the chimeras' vehicle were used as controls. The cells were incubated for 15 min at 37 °C, washed twice with PBS and loaded with CCF4-AM (LiveBLAzer™ FRET-B/G Loading Kit, Life Technologies; K1095), according to the manufacturer's instructions, in HBSS supplemented with 10 mM glutamine, 10 mM HEPES, 1 mM sodium pyruvate and 10% (v/v) FBS for 30 min at room temperature (RT). Subsequently, mBMDM were washed twice with supplemented HBSS and fixed at RT for 15 min in 4% (w/v) paraformaldehyde in Dulbecco's Phosphate Buffered Saline (DPBS). Cells were then observed with a CFI PL APO LAMBDA 40X/0.95 objective on a Nikon Eclipse Ti-E microscope (Nikon). The samples were illuminated at 395 nm by a SpectraX light engine (Lumencor) using a quad dichroic filter 310DA/FI/TR/CY5-A and emission filters 450/50 and 525/50 (Semrock). Images were acquired with an EMCCD camera iXon ULTRA 888 (Andor Technologies). In each experiment, a determined number (as provided in the respective figure) of microscopic fields per condition were acquired and 3 independent experiments were performed. Ratiometric analysis of the microscopic field images acquired for emission wavelengths (Em) 447 nm and 520 nm, corresponding to the ratio of cleaved CCF4-AM/intact CCF4-AM, was made using a custom-made ImageJ macro ( https://github.com/PaulaSampaio/ALM-i3S_macros ) on Fiji software45 (ImageJ version 1.51n, NIH, USA) 104 . In parallel to every FRET-based assay, the activity of ConcA or 17-DMAG was confirmed by evaluating the inhibitory effect of the drugs on AIP56 intoxication, as previously described 23 . For this, mBMDM were pre-treated with ConcA (10 nM), 17-DMAG (20 µM) or their respective vehicles, in supplemented DMEM without FBS for 1 h at 37 °C prior to incubation with 174 nM AIP56 for 2 h while maintaining the inhibitory conditions. Mock-treated cells, cells treated only with the toxins and cells treated only with inhibitors were used as controls. The cleavage NF-κB p65 was evaluated by western blotting and was used as readout of AIP56 toxicity. SDS-PAGE and western blotting SDS-PAGE was performed using the Laemmli discontinuous buffer system 105 . Prior to loading, the samples were boiled for 5 min in SDS-PAGE sample buffer (50 mM Tris-HCl pH 8.8, 2% (v/v) SDS, 0.05% (v/v) bromophenol blue, 10% (v/v) glycerol, 2 mM EDTA, 100 mM DTT). For western blotting, the proteins were transferred onto Amersham TM Protran 0.45 µm nitrocellulose membranes (GE Healthcare Life Science). The protein loading on the membranes was controlled by Ponceau S staining. The membranes were blocked for 1 h at RT with 5% (w/v) skimmed milk in Tris-buffered saline (TBS) containing 0.1% (v/v) Tween 20 (TBS-T) followed by incubation for 1 h at RT with the primary antibodies diluted in blocking buffer. Immunoreactive bands were detected with alkaline phosphatase conjugated secondary antibodies and NBT/BCIP (Promega). Blots shown correspond to representative results of at least 2 independent experiments. Uncropped scans of all blots are provided in the Source Data file. The quantification of the blots was performed by densitometry using the Fiji software 104 . The results are expressed as the density of the p65 band relative to the density of the Ponceau S of the same lane. Each graph combines the results of at least two independent experiments. Interaction with black lipid bilayers Analysis of the interaction of AIP56, AIP56 N1-G256 , AIP56 N1-E307 , AIP56 L258-N497 , AIP56 T299-N497 , AIP56 E214K , AIP56 E218K , AIP56 H222K , AIP56 H231K , AIP56 E234K , AIP56 D274S/D276-278S , AIP56 D274N/D276-278N , Bla L19-W286 AIP56 P210-N497 , Bla L19-W286 AIP56 L258-N497 and AIP56 N1-E307 DTR S406-S560 with black lipid bilayers was performed as described previously 19 , 31 , 106 . Briefly, membranes were formed between two aqueous compartments of a Teflon chamber from DiPhPC/n‐decane (diphytanoyl phosphatidylcholine [Avanti Polar Lipids, Alabaster, AL] in n‐decane) in a 150 mM KCl solution buffered with 10 mM HEPES to pH 7.0 or 10 mM MES to pH 6.0. After the membrane had turned black, purified protein samples mixed 1:1 with cholesterol suspension in water were added to the cis compartment of the chamber while stirring. Single-channel recordings of the membrane current in the presence of recombinant protein were performed with 50 mV or 150 mV at RT. With exception of the Bla-AIP56 chimeras, all tested samples did not show any membrane activity under these conditions for at least 10 min, showing that the samples were essentially free of bacterial contaminants that could form channels (porins). For Bla chimeras, to inhibit eventual porin activity, 50 μM of norfloxacin (OmpF) and 50 μM of ceftobiprole (OmpC) were added in the cis-side of the chamber. When indicated, membrane activity was induced by acidification (pH 4.8–5.0) of the aqueous phase at the cis‐side of the chamber. The membrane current was measured with a pair of Ag/AgCl electrodes with salt bridges switched in series with a voltage source and a highly sensitive current amplifier (Keithley 427). The amplified signal was recorded by a strip chart recorder. Each experiment was repeated at least three times. Limited proteolysis AIP56 or AIP56 H231K/E234K (5.36 μM) were incubated with Proteinase K (8.65 nM) on ice in 20 mM Tris pH 8.0, 200 mM NaCl or 20 mM Bis-Tris pH 6.5 or pH 5.0 with 200 mM NaCl. Aliquots were removed at different incubation times and the reaction stopped by adding SDS-PAGE sample buffer followed by heating at 95 °C for 5 min. Digests were analyzed by SDS-PAGE. pH‐induced translocation assay Translocation across the cell membrane in response to acidification was assessed using a protocol first described for diphtheria toxin 107 , 108 and used before for wild type AIP56 19 , 31 . Briefly, mBMDM were incubated for 30 min on ice with 174 nM AIP56 or AIP56 variants, in the absence or presence of 10 nM ConcA (that was maintained during the entire assay to inhibit normal toxin uptake). After removing the supernatant, the cells were incubated for 1 h at 37 °C with buffer at pH 4.5 or 7.0 followed by incubation in culture medium at pH 7.4 for 2 h. Different pH values were obtained by adding H 3 PO 4 to a solution containing 0.5 mM MgCl 2 , 0.9 mM CaCl 2 , 2.7 mM KCl, 1.5 mM KH 2 PO 4 , 3.2 mM Na 2 HPO 4 and 137 mM NaCl. In all experiments, mock-treated cells were used as controls. NF-kB p65 cleavage was analyzed by western blotting. Fluorescence assay with 8-Anilino-1-naphthalenesulfonic acid (ANS) Binding of ANS to the recombinant proteins was measured by following the increase of the fluorescence signal (from 400 to 550 nm) on a Horiba Fluoromax-4 spectrofluorimeter (excitation at 380 nm). V5 plus His-tagged proteins (1.5 µM) were titrated into a quartz cuvette with 75 µM ANS. The experiments were conducted at 25 °C at different pH (100 mM ammonium acetate, 150 mM NaCl for pH 5.0 and 5.5; 100 mM MOPS, 150 mM NaCl for pH 6.0, and 100 mM HEPES, 150 mM NaCl for pH 7.0). Bioinformatic tools The 10 amino acid T1-like motif in AIP56 was first identified in silico by Blasting ( https://blast.ncbi.nlm.nih.gov/Blast.cgi ) the known T1-like motif from diphtheria toxin, botulinum neurotoxins serotypes A, C and D, Clostridium difficile toxin B (TcdB), cytotoxic necrotizing factor 1 (CNF1), P. multocida toxin and anthrax edema and lethal factors (Supplementary Table 1 ) against AIP56 sequence. The identified AIP56 T1-like motif was then analyzed by the Multiple Expectation maximization for Motif Elucidation MEME tool ( https://meme-suite.org/meme/tools/meme ) 109 . The helical wheel (Supplementary Fig. 5b ) was drawn using Galaxy ( https://cpt.tamu.edu/galaxy-pub ) 110 . Surface charge distribution of the catalytic domain of AIP56 (Supplementary Fig. 5c ) at pH 7.4, 6.0 and 5.0 was calculated using PDB2PQR and APBS 111 based on pK a values predicted by PROPKA3.1 112 . Statistical analysis In each legend, the size of the experiment is indicated. Normality of the data were assessed through the Shapiro–Wilk test. Except for Fig. 4b , one‐way ANOVA was used to assess the differences between conditions. Regarding Figs. 2c and 5b , multiple comparisons were evaluated using the Tukey's HSD (honest significant difference) test. Since the data in Fig. 4b did not pass the normality test, differences between conditions were assessed by Kruskal–Wallis nonparametric test and the adjusted p values for individual comparisons were obtained by Bonferroni correction. Statistical significance was set to p < 0.05. All statistical analysis was performed with a confidence interval of 95%. Statistical analysis was performed using the IBM SPSS Statistics (v25) software (Armonk, NY). All graphs were constructed using GraphPad Prism 8 (GraphPad Software, San Diego, CA). Reporting summary Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article. Ethics statement This study was carried out in accordance with European and Portuguese legislation for the use of animals for scientific purposes (Directive 2010/63/EU; Decreto-Lei 113/2013). The work was approved by the ORBEA (Animal Welfare and Ethics Body) of i3S and was licensed by Direcção-Geral de Alimentação e Veterinária (DGAV), the Portuguese authority for animal protection (ref. 004933). Mice C57BL/6J mice were bred and housed (filter top cages, eurostandard type II, corn cob bedding material and nesting paper and plastic tube rolls for enrichment) at the animal facility of the Instituto de Investigação e Inovação em Saúde with 12 h light/12 h dark light cycle, 20–24 °C and 45–65% humidity. The mice were fed sterilized food (2014S, Envigo) and water ad libitum and were euthanized by CO 2 inhalation followed by cervical dislocation. Reagents and antibodies PMSF (P7626), pronase E from Streptomyces griseus (P5147), diphtheria toxin (DT) from Corynebacterium diphtheriae (D0564), Proteinase K from Engyodontium album (P2308), α-chymotrysin from bovine pancreas type II (C4129), 8-Anilino-1-naphthalenesulfonic acid (ANS, A1028) and concanamycin A (ConcA, C9705) were purchased from Sigma Aldrich. HBSS, HEPES, l -glutamine, sodium pyruvate and 17-(dimethylaminoethylamino)−17-demethoxygeldanamycin (17-DMAG, ant-dgl-5) were purchased from Invitrogen. DMEM (10938-025), heat-inactivated fetal bovine serum (FBS, 10500064) and penicillin/streptomycin (P/S, 15140122), all Gibco, and LiveBLAzer™ FRET-B/G Loading Kit with CCF4-AM (K1095) were purchased from Life Technologies. CCF4-AM is a lipophilic, esterified form of the CCF4 substrate, which is a Fluorescence Resonance Energy Transfer (FRET) substrate that consists of a cephalosporin core linking 7-hydroxycoumarin to fluorescein. Isopropyl β-D-1-thiogalactopyranoside (IPTG, MB026) was purchased from NZYTech. The anti-human NF-κB p65 C-terminal domain (clone c-20) rabbit polyclonal antibody (sc-372, dilution 1:3000) was from Santa Cruz Biotechnology and the anti-V5 (R960-25, dilution 1:5000) mouse monoclonal antibody was purchased from Invitrogen. Goat anti-IgG rabbit alkaline phosphatase conjugated secondary antibody (A9919, dilution 1:10000) and goat anti-IgG mouse alkaline phosphatase conjugated secondary antibody (A2429, dilution 1:10000) were purchased from Sigma Aldrich. Cells Mouse bone marrow-derived macrophages (mBMDM) were derived from bone marrow of femurs and tibias from 4–8 week-old C57BL/6J male mice, as previously described 90 . Briefly, femurs and tibias were flushed with 10 mL of Hanks' balanced salt solution (HBSS) and the cells recovered by centrifugation and resuspended in supplemented DMEM (DMEM with 10 mM glutamine, 10 mM HEPES, 1 mM sodium pyruvate, 10% (v/v) FBS and 1% penicillin/streptomycin) plus 10% (v/v) of L929 cell conditioned medium (LCCM) as a source of macrophage colony stimulating factor 91 . Fibroblasts were removed by culturing the cells overnight, at 37 °C in a humidified chamber in a 7% (v/v) CO 2 atmosphere, on a cell culture dish. The non-adherent cells were collected, centrifuged, resuspended in supplemented DMEM plus 10% (v/v) LCCM at a concentration of 5 × 10 5 cells mL −1 , plated in 24-well cell culture plates at a density of 5 × 10 5 cells/well and incubated as above. Three days later, 10% (v/v) LCCM was added and on the 7th day, the medium was renewed. Cells were used at day 10. To obtain LCCM, L929 cells were grown in 75 cm 2 filtered cap flasks in supplemented DMEM until reaching 100% confluence. Cells were then diluted 1:100 in fresh supplemented DMEM and incubated for 10 days at 37 °C, 7% (v/v) CO 2 . The supernatant was collected, pooled, centrifuged at 750 × g for 10 min and filtered. LCCM was aliquoted and stored at −20 °C until used. DNA Constructs for recombinant protein production Constructs encoding full-length AIP56 cloned into the NcoI/XhoI restriction sites of pET28a (Novagen) in frame with a C-terminal 6His-tag (pET28AIP56H+) or with a V5 plus 6xHis-tag (pET28AIP56V5H+) are described in ref. 16 and ref. 19 , respectively. The coding regions of AIP56 N1-G256 , AIP56 N1-E307 , AIP56 L258-N497 and AIP56 T299-N497 were PCR-amplified using pET28AIP56H+ as template and cloned into the NcoI/XhoI restriction sites of pET28a in frame with a 6xHis-tag at the C-terminus (AIP56 N1-G256 , AIP56 N1-E307 and AIP56 T299-N497 ) or N-terminus (AIP56 L258-N497 ). To obtain the constructs for expressing Bla-AIP56 chimeric proteins, the sequence encoding Bla L19-W286 was amplified from plasmid p327 (gifted by Dr. Dimitri Panagiotis Papatheodorou) and cloned into pET28a NcoI/SacI restriction sites, yielding plasmid pET28Bla L19-W286 . DNA sequences encoding AIP56 P210-N497 or AIP56 L258-N497 were then amplified using pET28AIP56H+ as template and ligated into the SacI/XhoI restriction sites of pET28Bla L19-W286 in frame with a C-terminal 6xHis-tag. To generate chimera AIP56 N1-E307 DTR S406-S560 , AIP56 N1-E307 encoding sequence was amplified from pET28AIP56H+ and cloned into pET28 NcoI/SacI restriction sites, yielding plasmid pET28AIP56 N1-E307 . DTR S406-S560 coding sequence (accession number WP_072564851.1 ) was then amplified from pET-22b DT 51E/148K (gifted by Dr. John Collier; Addgene plasmid # 11081 ; RRID:Addgene_11081) and ligated into the SacI/XhoI restriction sites of pET28AIP56 N1-E307 in frame with a C-terminal 6xHis-tag. Constructs coding for AIP56 E214K , AIP56 E218K , AIP56 H222K , AIP56 H231K , AIP56 E234K , AIP56 E214K/E218K/H222K , AIP56 H231K/E234K , AIP56 D274S/D276-278S and AIP56 D274N/D276-278N with or without V5-tag were generated with the QuickChange Site‐Directed Mutagenesis Kit (Stratagene, 200518) following the manufacturer's instructions, using pET28AIP56V5H+ 19 and pET28AIP56H+ 16 as template, respectively. Plasmids and primers used in this study are listed in Supplementary Table 5 . Production of recombinant proteins Full-length His-tagged AIP56 (AIP56), full-length V5 plus His-tagged AIP56 (AIP56-V5), AIP56 N1-G256 , AIP56 N1-E307 DTR S406-S560 and AIP56 mutant variants were expressed in Escherichia coli BL21(DE3), AIP56 N1-E307 in E. coli BL21-CodonPlus (DE3) and AIP56 T299-N497 in E. coli SoluBL21 (DE3). Bla-AIP56 chimeras and AIP56 L258-N497 were expressed in E. coli Rosetta (DE3). Transformed E. coli cells were cultured at 37 °C in 1 L of Luria Bertani (LB) broth with shaking (200 rpm), except for AIP56 N1-E307 DTR S406-S560 for which 6 L were used. At OD 600 ~ 0.6, 0.5 mM IPTG was added and protein expression carried out at 17 °C for 4 h in the case of Bla-AIP56 chimeras or for 20 h for all other proteins. Cells were harvested by centrifugation and resuspended in 30 mL of Buffer: (i) 50 mM Bis-Tris pH 6.5, 300 mM NaCl for AIP56, AIP56 N1-G256 , AIP56 T299-N497 and AIP56 mutant variants; (ii) 50 mM Tris pH 8.0, 300 mM NaCl for AIP56 L258-N497 , AIP56 N1-E307 and AIP56 N1-E307 DTR S406-S560 ; or (iii) 20 mM Tris pH 8.0, 200 mM NaCl, 5% (v/v) glycerol for Bla-AIP56 chimeras. In all cases, cell lysis was performed by sonication. After centrifugation, the recombinant proteins were purified from the supernatant using nickel‐affinity chromatography (Ni‐NTA agarose, ABT) followed by size-exclusion chromatography (Superose 12 10/300 GL, GE Healthcare), except in the cases of AIP56 N1-E307 and AIP56 N1-E307DTR in which only nickel‐affinity chromatography was performed. Recombinant proteins were analyzed by SDS-PAGE and purities determined by densitometry of Coomassie Blue-stained gels (Image Lab Software, BioRad). Protein batches used in this work were ≥90% pure. Determination of protein concentration The concentrations of recombinant proteins were determined by measuring absorbance at 280 nm using a NanoDrop 1000 and/or NanoDrop One (Thermo Fisher Scientific) considering the extinction coefficient and the molecular weight calculated with the ProtParam tool ( http://www.expasy.org/tools/protparam.html ). Crystallization Initial crystallization hits for AIP56 were identified by high-throughput screening at the HTX Lab of the EMBL Grenoble Outstation (Grenoble, France). AIP56 crystals were obtained by mixing equal volumes of protein (7 mg mL −1 ) and crystallization solution (400 mM sodium acetate, 100 mM Tris pH 8.5, 15% (w/v) PEG 4 K, 10 mM taurine), which was optimized using an additive screen kit (HR2-428; Hampton Research). Prior to data collection, crystals were successively transferred into the crystallization solution supplemented with 15% and 30% (v/v) glycerol, followed by flash-cooling in liquid nitrogen. Structure determination Several X-ray diffraction datasets were collected from AIP56 crystals at Synchrotron SOLEIL (Saint-Aubin, France; beamlines Proxima-1 and Proxima-2) and at the European Synchrotron Radiation Facility (ESRF, Grenoble, France; beamlines ID23, ID29 and ID30). The best dataset was collected on beamline ID29 on a Pilatus 6 M (Dectris) detector (2700 images, 0.1° rotation, 0.02 s exposure, wavelength 0.972 à ), indexed and integrated with XDS 92 and analyzed using the STARANISO server as diffraction was highly anisotropic 93 . Statistics for data processing are summarized in Supplementary Table 1 . The structure of AIP56 was solved by molecular replacement with PHASER 94 using an AIP56 model generated by the artificial intelligence program AlphaFold2_advanced 35 , 36 . Several cycles of model building and refinement were performed using COOT, Phenix.refine and BUSTER 95 – 99 . All illustrations of macromolecular models were produced with PyMOL 100 . The final refined coordinates and structure factors were deposited at the Protein Data Bank under PDB entry 7ZPF and the corresponding diffraction images at the SBGrid Data Bank (10.15785/SBGRID/911). Small-angle X-ray scattering (SAXS) data acquisition, processing and analysis X-ray scattering data were collected at the SWING beamline of the SOLEIL Synchrotron (Saint-Aubin, France). Measurements were performed using a HPLC Bio Sec3 (Agilent) size-exclusion column, online with the SAXS measuring cell, a 1.5 mm diameter quartz capillary contained in an evacuated vessel. The sample-to-detector (Dectris Eiger 4 M) distance was set to 2000 mm and the wavelength λ to 1.0 à , allowing useful data collection over the momentum transfer range of 0.005 à −1 < q < 0.5 à −1 ( q = 4π sin( θ )/ λ ). SAXS data were collected directly after elution of the AIP56 protein through the HPLC column equilibrated in 50 mM HEPES buffer pH 7.5, 500 mM NaCl. Fifty µl of protein sample were injected at 15 °C and two different initial concentrations, 13.4 and 3.2 mg mL −1 , in order to obtain a curve with a good statistic and devoid of interparticle effects. The flow rate was 0.2 mL min −1 , frame duration was 0.99 sec and the dead time between frames was 0.01 sec. Scattering of the elution buffer before void volume was recorded and used as buffer scattering for subtraction from all protein patterns. The scattered intensities were displayed on an absolute scale using the scattering by water. Data were first analyzed using Foxtrot and then using the US-SOMO HPLC module 101 . This program provided for each SAXS frame the values of the scattering intensity I(0) and of the radius of gyration Rg by applying the Guinier analysis together with a calculation of the approximate molar mass using the Rambo and Tainer approach 102 . Identical frames under the main elution peak were selected using Cormap 103 and averaged for further analysis. Both averaged curves corresponding to the two initial concentrations were merged. Data was deposit in the Small-Angle Scattering Biological Data Bank under accession code SASDNW6 . Intoxication assays mBMDM monolayers seeded in 24-well plates (5 × 10 5 cells/well) were intoxicated by continuous incubation with 174 nM AIP56 or AIP56 variants, diluted in culture medium (supplemented DMEM + 10% (v/v) LCCM) at 37 °C for 4 h. The NF-κB p65 level was evaluated by western blotting after 2 h incubation with the toxin and was used as readout for arrival of AIP56 into the cytosol. Fluorescence Resonance Energy Transfer (FRET) based assay mBMDM cells cultured at a density of 1.5 × 10 5 cells/well in 8-well plates (ibidi) were pre-treated with concanamycin A (ConcA; 10 nM), 17-DMAG (20 µM) or their vehicles (DMSO for ConcA and water for 17-DMAG) in supplemented DMEM without FBS for 1 h at 37 °C. Stock solutions (250 µM ConcA in DMSO and 8.1 mM 17-DMAG in water) were diluted with culture medium prior to the experiment. Next, the medium was replaced by fresh medium containing the inhibitors and Bla L19-W286 AIP56 P210-N497 or Bla L19-W286 AIP56 L258-N497 was added at a final concentration of 525 nM. Untreated cells and cells treated with the chimeras' vehicle were used as controls. The cells were incubated for 15 min at 37 °C, washed twice with PBS and loaded with CCF4-AM (LiveBLAzer™ FRET-B/G Loading Kit, Life Technologies; K1095), according to the manufacturer's instructions, in HBSS supplemented with 10 mM glutamine, 10 mM HEPES, 1 mM sodium pyruvate and 10% (v/v) FBS for 30 min at room temperature (RT). Subsequently, mBMDM were washed twice with supplemented HBSS and fixed at RT for 15 min in 4% (w/v) paraformaldehyde in Dulbecco's Phosphate Buffered Saline (DPBS). Cells were then observed with a CFI PL APO LAMBDA 40X/0.95 objective on a Nikon Eclipse Ti-E microscope (Nikon). The samples were illuminated at 395 nm by a SpectraX light engine (Lumencor) using a quad dichroic filter 310DA/FI/TR/CY5-A and emission filters 450/50 and 525/50 (Semrock). Images were acquired with an EMCCD camera iXon ULTRA 888 (Andor Technologies). In each experiment, a determined number (as provided in the respective figure) of microscopic fields per condition were acquired and 3 independent experiments were performed. Ratiometric analysis of the microscopic field images acquired for emission wavelengths (Em) 447 nm and 520 nm, corresponding to the ratio of cleaved CCF4-AM/intact CCF4-AM, was made using a custom-made ImageJ macro ( https://github.com/PaulaSampaio/ALM-i3S_macros ) on Fiji software45 (ImageJ version 1.51n, NIH, USA) 104 . In parallel to every FRET-based assay, the activity of ConcA or 17-DMAG was confirmed by evaluating the inhibitory effect of the drugs on AIP56 intoxication, as previously described 23 . For this, mBMDM were pre-treated with ConcA (10 nM), 17-DMAG (20 µM) or their respective vehicles, in supplemented DMEM without FBS for 1 h at 37 °C prior to incubation with 174 nM AIP56 for 2 h while maintaining the inhibitory conditions. Mock-treated cells, cells treated only with the toxins and cells treated only with inhibitors were used as controls. The cleavage NF-κB p65 was evaluated by western blotting and was used as readout of AIP56 toxicity. SDS-PAGE and western blotting SDS-PAGE was performed using the Laemmli discontinuous buffer system 105 . Prior to loading, the samples were boiled for 5 min in SDS-PAGE sample buffer (50 mM Tris-HCl pH 8.8, 2% (v/v) SDS, 0.05% (v/v) bromophenol blue, 10% (v/v) glycerol, 2 mM EDTA, 100 mM DTT). For western blotting, the proteins were transferred onto Amersham TM Protran 0.45 µm nitrocellulose membranes (GE Healthcare Life Science). The protein loading on the membranes was controlled by Ponceau S staining. The membranes were blocked for 1 h at RT with 5% (w/v) skimmed milk in Tris-buffered saline (TBS) containing 0.1% (v/v) Tween 20 (TBS-T) followed by incubation for 1 h at RT with the primary antibodies diluted in blocking buffer. Immunoreactive bands were detected with alkaline phosphatase conjugated secondary antibodies and NBT/BCIP (Promega). Blots shown correspond to representative results of at least 2 independent experiments. Uncropped scans of all blots are provided in the Source Data file. The quantification of the blots was performed by densitometry using the Fiji software 104 . The results are expressed as the density of the p65 band relative to the density of the Ponceau S of the same lane. Each graph combines the results of at least two independent experiments. Interaction with black lipid bilayers Analysis of the interaction of AIP56, AIP56 N1-G256 , AIP56 N1-E307 , AIP56 L258-N497 , AIP56 T299-N497 , AIP56 E214K , AIP56 E218K , AIP56 H222K , AIP56 H231K , AIP56 E234K , AIP56 D274S/D276-278S , AIP56 D274N/D276-278N , Bla L19-W286 AIP56 P210-N497 , Bla L19-W286 AIP56 L258-N497 and AIP56 N1-E307 DTR S406-S560 with black lipid bilayers was performed as described previously 19 , 31 , 106 . Briefly, membranes were formed between two aqueous compartments of a Teflon chamber from DiPhPC/n‐decane (diphytanoyl phosphatidylcholine [Avanti Polar Lipids, Alabaster, AL] in n‐decane) in a 150 mM KCl solution buffered with 10 mM HEPES to pH 7.0 or 10 mM MES to pH 6.0. After the membrane had turned black, purified protein samples mixed 1:1 with cholesterol suspension in water were added to the cis compartment of the chamber while stirring. Single-channel recordings of the membrane current in the presence of recombinant protein were performed with 50 mV or 150 mV at RT. With exception of the Bla-AIP56 chimeras, all tested samples did not show any membrane activity under these conditions for at least 10 min, showing that the samples were essentially free of bacterial contaminants that could form channels (porins). For Bla chimeras, to inhibit eventual porin activity, 50 μM of norfloxacin (OmpF) and 50 μM of ceftobiprole (OmpC) were added in the cis-side of the chamber. When indicated, membrane activity was induced by acidification (pH 4.8–5.0) of the aqueous phase at the cis‐side of the chamber. The membrane current was measured with a pair of Ag/AgCl electrodes with salt bridges switched in series with a voltage source and a highly sensitive current amplifier (Keithley 427). The amplified signal was recorded by a strip chart recorder. Each experiment was repeated at least three times. Limited proteolysis AIP56 or AIP56 H231K/E234K (5.36 μM) were incubated with Proteinase K (8.65 nM) on ice in 20 mM Tris pH 8.0, 200 mM NaCl or 20 mM Bis-Tris pH 6.5 or pH 5.0 with 200 mM NaCl. Aliquots were removed at different incubation times and the reaction stopped by adding SDS-PAGE sample buffer followed by heating at 95 °C for 5 min. Digests were analyzed by SDS-PAGE. pH‐induced translocation assay Translocation across the cell membrane in response to acidification was assessed using a protocol first described for diphtheria toxin 107 , 108 and used before for wild type AIP56 19 , 31 . Briefly, mBMDM were incubated for 30 min on ice with 174 nM AIP56 or AIP56 variants, in the absence or presence of 10 nM ConcA (that was maintained during the entire assay to inhibit normal toxin uptake). After removing the supernatant, the cells were incubated for 1 h at 37 °C with buffer at pH 4.5 or 7.0 followed by incubation in culture medium at pH 7.4 for 2 h. Different pH values were obtained by adding H 3 PO 4 to a solution containing 0.5 mM MgCl 2 , 0.9 mM CaCl 2 , 2.7 mM KCl, 1.5 mM KH 2 PO 4 , 3.2 mM Na 2 HPO 4 and 137 mM NaCl. In all experiments, mock-treated cells were used as controls. NF-kB p65 cleavage was analyzed by western blotting. Fluorescence assay with 8-Anilino-1-naphthalenesulfonic acid (ANS) Binding of ANS to the recombinant proteins was measured by following the increase of the fluorescence signal (from 400 to 550 nm) on a Horiba Fluoromax-4 spectrofluorimeter (excitation at 380 nm). V5 plus His-tagged proteins (1.5 µM) were titrated into a quartz cuvette with 75 µM ANS. The experiments were conducted at 25 °C at different pH (100 mM ammonium acetate, 150 mM NaCl for pH 5.0 and 5.5; 100 mM MOPS, 150 mM NaCl for pH 6.0, and 100 mM HEPES, 150 mM NaCl for pH 7.0). Bioinformatic tools The 10 amino acid T1-like motif in AIP56 was first identified in silico by Blasting ( https://blast.ncbi.nlm.nih.gov/Blast.cgi ) the known T1-like motif from diphtheria toxin, botulinum neurotoxins serotypes A, C and D, Clostridium difficile toxin B (TcdB), cytotoxic necrotizing factor 1 (CNF1), P. multocida toxin and anthrax edema and lethal factors (Supplementary Table 1 ) against AIP56 sequence. The identified AIP56 T1-like motif was then analyzed by the Multiple Expectation maximization for Motif Elucidation MEME tool ( https://meme-suite.org/meme/tools/meme ) 109 . The helical wheel (Supplementary Fig. 5b ) was drawn using Galaxy ( https://cpt.tamu.edu/galaxy-pub ) 110 . Surface charge distribution of the catalytic domain of AIP56 (Supplementary Fig. 5c ) at pH 7.4, 6.0 and 5.0 was calculated using PDB2PQR and APBS 111 based on pK a values predicted by PROPKA3.1 112 . Statistical analysis In each legend, the size of the experiment is indicated. Normality of the data were assessed through the Shapiro–Wilk test. Except for Fig. 4b , one‐way ANOVA was used to assess the differences between conditions. Regarding Figs. 2c and 5b , multiple comparisons were evaluated using the Tukey's HSD (honest significant difference) test. Since the data in Fig. 4b did not pass the normality test, differences between conditions were assessed by Kruskal–Wallis nonparametric test and the adjusted p values for individual comparisons were obtained by Bonferroni correction. Statistical significance was set to p < 0.05. All statistical analysis was performed with a confidence interval of 95%. Statistical analysis was performed using the IBM SPSS Statistics (v25) software (Armonk, NY). All graphs were constructed using GraphPad Prism 8 (GraphPad Software, San Diego, CA). Reporting summary Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article. Supplementary information Supplementary Information Peer Review File Reporting Summary Source data Source Data Supplementary information Supplementary Information Peer Review File Reporting Summary Source data Source Data Supplementary information The online version contains supplementary material available at 10.1038/s41467-023-43054-z. Peer review information Nature Communications thanks Melissa Graewert and the other, anonymous, reviewers for their contribution to the peer review of this work. A peer review file is available.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8675849/

State Feedback and Synergetic controllers for tuberculosis in infected population

Abstract Tuberculosis (TB) is a contagious disease which can easily be disseminated in a society. A five state Susceptible, exposed, infected, recovered and resistant (SEIRs) epidemiological mathematical model of TB has been considered along with two non‐linear controllers: State Feedback (SFB) and Synergetic controllers have been designed for the control and prevention of the TB in a population. Using the proposed controllers, the infected individuals have been reduced/controlled via treatment, and susceptible individuals have been prevented from the disease via vaccination. A mathematical analysis has been carried out to prove the asymptotic stability of proposed controllers by invoking the Lyapunov control theory. Simulation results using MATLAB/Simulink manifest that the non‐linear controllers show fast convergence of the system states to their respective desired levels. Comparison shows that proposed SFB controller performs better than Synergetic controller in terms of convergence time, steady state error and oscillations. 1 INTRODUCTION Tuberculosis (TB), having two major types, MDR tuberculosis and XDR tuberculosis, is caused by a bacterium called mycobacterium . It is one of the top 10 causes of death across the world [ 1 ], which affects mainly the human lungs apart from other parts like brain, bones, kidney and spine. It is a transferable disease that can spread over a population. Figure 1 shows that when an infected individual exhales, mycobacterium is transferred to the air that can affect the healthy individuals in the surrounding. The other reasons of TB are bad living conditions, malnourishment, smoking, and so forth. FIGURE 1 Transmission of tuberculosis in population [ 2 ] Microbacterium Tuberculosis (Mtb) patients who show resistant to anti‐TB drugs, Isoniazed and Rifampicin, are termed as MDR patients and who show resistant to any injectable anti‐TB drug are termed as XDR. These types of patients do not respond to 6 or 9 months' treatment. They may take about 2 years of treatment with high toxic anti‐TB medicines to fully recover. These types of Mtb patients are real threat to control and prevent the TB from its spread [ 2 ]. Individuals who are HIV positive and infected from TB have 20%–40% more chances to develop active TB which is a leading cause of death HIV positive patients [ 2 ]. According to the WHO annual report 2019, globally 1.7 billion people are infected with Mtb and approximately 10 million people suffer from TB every year. About 50–500 people per million population are infected across the world. The male‐female ratio of TB is 2:1. It can affect anybody but is more dangerous for the adults. Developing countries are highly burdened from TB because of poverty, bad living conditions, unavailability of treatment facilities and malnourishment. The spread of TB can be curtailed by timely diagnosis, treatment, improvement of the health facilities and introducing health conscious activities in the society [ 1 ]. Early diagnosis of the disease decreases both social and medical impacts. Surveillance of TB can be done by using Google trends [ 3 ] and by observing its counter medication [ 4 ]. Spread of TB can also be controlled by giving health education to the society. Effect of different health education methods on secondary and primary school students in northern province of Jiangsu has been discussed in [ 5 ]. Bio‐mathematics has played a very important role in the development of the mathematical models of epidemic diseases including TB. In previous research works, several mathematical models of TB have been developed. First of all the stability of the model is examined, then a preventive control in the form of vaccination and treatment of infected class is established and an objective function is defined by researchers. Later, an optimal control law for prevention and control of the infectious TB was defined [ 6 , 7 ]. Discrete TB model with two different infectious compartments has been discussed in [ 8 , 9 ]. Stability analysis, bifurcation of TB model and complex system modelling of TB in Nigeria have been discussed in [ 10 , 11 ]. Computer modelling of sensitive type mycobacterium TB and modelling using regression analysis have also been studied [ 12 , 13 ]. Influence of multiple reinfections in TB dynamics is discussed in [ 14 ]. Stability analysis has been discussed for five states TB model in [ 15 ]. Optimal control technique has been applied on four and five states non‐linear mathematical models of tuberculosis using Pontryagin's maximum principle [ 16 ]. A dynamic behaviour of four states of TB transmission and an optimal controller for its treatment has been discussed in [ 17 ]. Synergetic control technique takes into account a macro‐variable whose number depends on the number of inputs. It contains the errors of the states which we want to track [ 18 ]. It has been applied for tracking of infected cells during anti‐viral therapy [ 19 ], to control the growth of cancer cells [ 20 ], on systems of non‐linear equations [ 21 ], magnetic levitation system [ 22 ], to minimize HIV concentration in blood plasma [ 23 ] and for stabilizing the medium voltage microgrids [ 24 ]. In State Feedback controller design, the output tracks the desired reference signal asymptotically if the reference signal and its derivative are bounded []. In this research work, an updated mathematical model of the TB has been considered to design two non‐linear controllers for preventing the spread of TB and reducing the infected individuals by giving them treatment and vaccination on time. These non‐linear controllers are Synergetic and State Feedback (SFB) controllers which have been designed for the treatment and vaccination of infected population.Schematic diagram for the proposed close loop control system has been shown in Figure 2 . The rest of the article is organized as follows; Section 2 describes the non‐linear mathematical model of TB considered for this research. Section 2.2 details the problem statement, and Section 3 describes the design of the proposed non‐linear controllers. Simulation results have been presented in Section 4 , where the comparison of the proposed controllers has been made and finally the article has been concluded in Section 4 . FIGURE 2 Schematic diagram for close loop control system 2 NON‐LINEAR TUBERCULOSIS MODEL TB is an airborne disease that can spread from person to person via aerosolization to the individuals and through air. 2.1 SEIR tuberculosis model There are a number of mathematical models for the transmission of TB. SIR model [28] incorporates the three state variables: susceptible, infected and recovered. Latest model of the transmission of the Mtb is a five state SEIR model [29] which describes the transmission of the Mtb in a human host taking into account the effect of the MDR and XDR without making the model complicated. In this model, human population is classified into five classes; Infected ( x 1 ), Susceptible ( x 2 ), Exposed ( x 3 ), Recovered ( x 4 ) and Resistant ( x 5 ). Size of the human population N can be written as (1) N = x 1 + x 2 + x 3 + x 4 + x 5 where the recruitment to the susceptible class is taken by birth rate ( λ ). Size of each class varies due to natural death rate and rates at which the individuals become susceptible, exposed, infected, recovered and resistant. The model is given below. (2) x · 1 = ϵ x 2 − ( μ + α + γ + σ ) x 1 − u 1 (3) x · 2 = β x 2 x 3 − ( μ + ϵ ) x 2 (4) x · 3 = λ N − β x 1 x 3 − μ x 3 + ρ x 4 − u 2 (5) x · 4 = γ x 1 − ( μ + ρ ) x 4 + δ x 5 (6) x · 5 = σ x 4 ( μ + α 1 + δ ) x 5 Parameters that are used in the model are described as recruitment by births ( λ ), force of infection ( β ), human death rate ( μ ), active TB disease induction rate ( α ), MDR TB disease induction rate ( α 1 ), humans recovery rate from MDR TB ( δ ), rate at which exposed become infectious ( ϵ ), rate at which infected becomes resistant ( σ ) and rate at which recovered becomes susceptible ( ρ ). 2.2 Problem statement There are number of optimal control strategies for the prevention and control of TB, but it still is one of the leading causes of the death worldwide. There is no non‐linear controller purposed for the prevention and control of TB so far in the literature. This model is nonlinear due to presence of terms x 2 x 3 and x 1 x 3 in Equations ( 3 ) and ( 4 ), respectively. Therefor designing a non‐linear controller would be a good option to cater for the spread of TB, as non‐linear controllers usually show better convergence, lesser steady state error and negligible oscillations and undershoots/overshoots. 2.1 SEIR tuberculosis model There are a number of mathematical models for the transmission of TB. SIR model [28] incorporates the three state variables: susceptible, infected and recovered. Latest model of the transmission of the Mtb is a five state SEIR model [29] which describes the transmission of the Mtb in a human host taking into account the effect of the MDR and XDR without making the model complicated. In this model, human population is classified into five classes; Infected ( x 1 ), Susceptible ( x 2 ), Exposed ( x 3 ), Recovered ( x 4 ) and Resistant ( x 5 ). Size of the human population N can be written as (1) N = x 1 + x 2 + x 3 + x 4 + x 5 where the recruitment to the susceptible class is taken by birth rate ( λ ). Size of each class varies due to natural death rate and rates at which the individuals become susceptible, exposed, infected, recovered and resistant. The model is given below. (2) x · 1 = ϵ x 2 − ( μ + α + γ + σ ) x 1 − u 1 (3) x · 2 = β x 2 x 3 − ( μ + ϵ ) x 2 (4) x · 3 = λ N − β x 1 x 3 − μ x 3 + ρ x 4 − u 2 (5) x · 4 = γ x 1 − ( μ + ρ ) x 4 + δ x 5 (6) x · 5 = σ x 4 ( μ + α 1 + δ ) x 5 Parameters that are used in the model are described as recruitment by births ( λ ), force of infection ( β ), human death rate ( μ ), active TB disease induction rate ( α ), MDR TB disease induction rate ( α 1 ), humans recovery rate from MDR TB ( δ ), rate at which exposed become infectious ( ϵ ), rate at which infected becomes resistant ( σ ) and rate at which recovered becomes susceptible ( ρ ). 2.2 Problem statement There are number of optimal control strategies for the prevention and control of TB, but it still is one of the leading causes of the death worldwide. There is no non‐linear controller purposed for the prevention and control of TB so far in the literature. This model is nonlinear due to presence of terms x 2 x 3 and x 1 x 3 in Equations ( 3 ) and ( 4 ), respectively. Therefor designing a non‐linear controller would be a good option to cater for the spread of TB, as non‐linear controllers usually show better convergence, lesser steady state error and negligible oscillations and undershoots/overshoots. 3 NON‐LINEAR CONTROLLERS DESIGN FOR SEIR TB MODEL We have considered SEIR TB model given by Equations ( (2) , (3) , (4) , (5) , (6) )–( (2) , (3) , (4) , (5) , (6) ) in order to design the controllers. Two non‐linear controllers, Synergetic and State Feedback controller, are to be designed for treatment and vaccination of infected population. The control inputs u 1 and u 2 give the number of infected and susceptible individuals for the treatment and vaccination respectively. 3.1 Synergetic controller design Synergetic controller is to be designed for the system to track some state of the system to its desired level. Synergetic control technique will be used to design the control input u 1 and u 2 . We have taken two macro‐variables, since the number of input variables are two, defined as (7) σ 1 = c 1 e 1 + c 2 e 3 + c 4 e 4 + c 5 e 5 and (8) σ 2 = c 2 e 2 + c 3 e 3 + c 4 e 4 + c 5 e 5 The error of each state is defined below which is the difference between actual value and reference value of that state. (9) e 1 = x 1 − x 1 r e f e 2 = x 2 − x 2 r e f e 3 = x 3 − x 3 r e f e 4 = x 4 − x 4 r e f e 5 = x 5 − x 5 r e f All the states would track the desired value if the errors converge to zero respectively. Taking the time derivative of Equation ( 10 ), we have (10) e · 1 = x · 1 − x · 1 r e f e · 2 = x · 2 − x · 2 r e f e · 3 = x · 3 − x · 3 r e f e · 4 = x · 4 − x · 4 r e f e · 5 = x · 5 − x · 5 r e f Since reference value of each state is constant, so their time derivatives will be zero, we get (11) e · 1 = x · 1 e · 2 = x · 2 e · 3 = x · 3 e · 4 = x · 4 e · 5 = x · 5 Taking time derivative of Equations ( 7 ) and ( 8 ), gives (12) σ · 1 = c 1 e · 1 + c 2 e · 2 + c 4 e · 4 + c 5 e · 5 (13) = c 1 x · 1 + c 2 x · 2 + c 2 x · 2 + c 5 x · 5 (14) σ · 2 = c 2 e · 2 + c 3 e · 3 + c 4 e · 4 + c 5 e · 5 (15) = c 2 x · 2 + c 3 x · 3 + c 4 x · 4 + c 5 x · 5 The macro‐variables σ 1 and σ 2 are supposed to satisfy the dynamic evaluation presented by the following equation (16) T σ · + σ = 0 where T represents the convergence rate of σ 1 and is a positive constant. Putting down the values of σ 1 and σ 1 · from Equations ( 7 ) and ( 13 ) respectively and solving for u 1 , we get (17) u 1 = 1 T c 1 − T c 2 x 2 · − T c 4 x 4 · − T c 5 x 5 · − σ 1 + T ϵ x 2 − T ( μ + α + γ + σ ) x 1 . Now, putting down the value of σ 2 and σ 2 · from Equations ( 8 ) and ( 15 ) respectively and solving for u 2 , we get (18) u 2 = 1 r T c 3 x 4 − T c 3 λ N − β x 1 x 3 T c 3 − μ x 3 T c 3 − σ − T c 4 x 4 · − T c 5 x 5 · . The control input u 1 and u 2 in Equations ( 17 ) and ( 18 ) are the required controls obtained through the Synergetic control technique which gives the number of infected and susceptible individuals to be treated and vaccinated, respectively. To prove asymptotic stability of Equation ( 16 ), we consider the Lyapunov candidate function as (19) V 3 = 1 2 σ 2 Taking the time derivative of Equation ( 19 ), we get (20) V · 3 = σ σ · Putting down the value of σ · from Equation ( 16 ), we get (21) V · 3 = − σ 2 T = − 2 T 1 2 σ 2 Using Equation ( 19 ), we can write (22) V · 3 = − 2 T V 3 When t is zero V 3 becomes equal to V o which is its initial value. (23) V 3 = V o e − 2 T t . Hence the dynamical system is exponentially stable using Lyapunov theory. 3.2 State Feedback controller design In order to design the State Feedback controller, we take x 1 as output of the system that is (24) y 1 = x 1 = e 1 + x 1 r e f Taking the time derivative of Equation ( 24 ), we have (25) y 1 · = x · 1 The state x 1 will track the desired value if the error e 1 will converge to zero. Therefore, taking time derivative of Equation ( 24 ) and putting down the value of x 1 · from Equation ( 2 ), we can write (26) e · 1 = ϵ x 2 − ( μ + α + γ + σ ) x 1 − u 1 Error e 1 will converge to zero if Lyapunov candidate function of error e 1 given by Equation ( 26 ) is negative definite. For this purpose, we keep (27) ϵ x 2 − ( μ + α + γ + σ ) x 1 − u 1 = − F 1 e 1 where F 1 is positive constant. Equation ( 26 ) becomes (28) e · 1 = − F 1 e 1 Solving Equation ( 27 ) for u 1 , we have (29) u 1 = ( μ + α + γ + σ ) x 1 − ϵ x 2 + F 1 e 1 The control input u 1 in Equation ( 29 ) is the required one obtained through State Feedback control technique which gives the number of infected individuals to be treated. In similar way, we can design u 2 by choosing x 3 as output (30) y 2 = x 3 Taking time derivative of the Equation ( 30 ), we have (31) y 2 · = x 3 · Taking time derivative of e 3 given by Equation ( 5 ), we have (32) e · 3 = x · 3 Using the value of x · 3 given by Equation ( 3 ), we have (33) e · 3 = λ N − β x 3 x 1 − μ x 3 + ρ x 4 − u 2 . Error e 3 given by Equation ( 5 ) will converge to zero if Lyapunov candidate function of error e 3 given by Equation ( 33 ) is negative definite. For this purpose, we keep (34) λ N − β x 3 x 1 − μ x 3 + ρ x 4 − u 2 = − F 2 e 3 Equation ( 33 ) becomes (35) e · 3 = − F 2 e 3 Solving Equation ( 34 ) for u 2 , we have (36) u 2 = λ N − β x 3 x 1 − μ x 3 + ρ x 4 + F 3 e 3 The control input u 2 in Equation ( 36 ) is the required control obtained through the State Feedback control technique which gives the number of susceptible individuals to be vaccinated. 3.1 Synergetic controller design Synergetic controller is to be designed for the system to track some state of the system to its desired level. Synergetic control technique will be used to design the control input u 1 and u 2 . We have taken two macro‐variables, since the number of input variables are two, defined as (7) σ 1 = c 1 e 1 + c 2 e 3 + c 4 e 4 + c 5 e 5 and (8) σ 2 = c 2 e 2 + c 3 e 3 + c 4 e 4 + c 5 e 5 The error of each state is defined below which is the difference between actual value and reference value of that state. (9) e 1 = x 1 − x 1 r e f e 2 = x 2 − x 2 r e f e 3 = x 3 − x 3 r e f e 4 = x 4 − x 4 r e f e 5 = x 5 − x 5 r e f All the states would track the desired value if the errors converge to zero respectively. Taking the time derivative of Equation ( 10 ), we have (10) e · 1 = x · 1 − x · 1 r e f e · 2 = x · 2 − x · 2 r e f e · 3 = x · 3 − x · 3 r e f e · 4 = x · 4 − x · 4 r e f e · 5 = x · 5 − x · 5 r e f Since reference value of each state is constant, so their time derivatives will be zero, we get (11) e · 1 = x · 1 e · 2 = x · 2 e · 3 = x · 3 e · 4 = x · 4 e · 5 = x · 5 Taking time derivative of Equations ( 7 ) and ( 8 ), gives (12) σ · 1 = c 1 e · 1 + c 2 e · 2 + c 4 e · 4 + c 5 e · 5 (13) = c 1 x · 1 + c 2 x · 2 + c 2 x · 2 + c 5 x · 5 (14) σ · 2 = c 2 e · 2 + c 3 e · 3 + c 4 e · 4 + c 5 e · 5 (15) = c 2 x · 2 + c 3 x · 3 + c 4 x · 4 + c 5 x · 5 The macro‐variables σ 1 and σ 2 are supposed to satisfy the dynamic evaluation presented by the following equation (16) T σ · + σ = 0 where T represents the convergence rate of σ 1 and is a positive constant. Putting down the values of σ 1 and σ 1 · from Equations ( 7 ) and ( 13 ) respectively and solving for u 1 , we get (17) u 1 = 1 T c 1 − T c 2 x 2 · − T c 4 x 4 · − T c 5 x 5 · − σ 1 + T ϵ x 2 − T ( μ + α + γ + σ ) x 1 . Now, putting down the value of σ 2 and σ 2 · from Equations ( 8 ) and ( 15 ) respectively and solving for u 2 , we get (18) u 2 = 1 r T c 3 x 4 − T c 3 λ N − β x 1 x 3 T c 3 − μ x 3 T c 3 − σ − T c 4 x 4 · − T c 5 x 5 · . The control input u 1 and u 2 in Equations ( 17 ) and ( 18 ) are the required controls obtained through the Synergetic control technique which gives the number of infected and susceptible individuals to be treated and vaccinated, respectively. To prove asymptotic stability of Equation ( 16 ), we consider the Lyapunov candidate function as (19) V 3 = 1 2 σ 2 Taking the time derivative of Equation ( 19 ), we get (20) V · 3 = σ σ · Putting down the value of σ · from Equation ( 16 ), we get (21) V · 3 = − σ 2 T = − 2 T 1 2 σ 2 Using Equation ( 19 ), we can write (22) V · 3 = − 2 T V 3 When t is zero V 3 becomes equal to V o which is its initial value. (23) V 3 = V o e − 2 T t . Hence the dynamical system is exponentially stable using Lyapunov theory. 3.2 State Feedback controller design In order to design the State Feedback controller, we take x 1 as output of the system that is (24) y 1 = x 1 = e 1 + x 1 r e f Taking the time derivative of Equation ( 24 ), we have (25) y 1 · = x · 1 The state x 1 will track the desired value if the error e 1 will converge to zero. Therefore, taking time derivative of Equation ( 24 ) and putting down the value of x 1 · from Equation ( 2 ), we can write (26) e · 1 = ϵ x 2 − ( μ + α + γ + σ ) x 1 − u 1 Error e 1 will converge to zero if Lyapunov candidate function of error e 1 given by Equation ( 26 ) is negative definite. For this purpose, we keep (27) ϵ x 2 − ( μ + α + γ + σ ) x 1 − u 1 = − F 1 e 1 where F 1 is positive constant. Equation ( 26 ) becomes (28) e · 1 = − F 1 e 1 Solving Equation ( 27 ) for u 1 , we have (29) u 1 = ( μ + α + γ + σ ) x 1 − ϵ x 2 + F 1 e 1 The control input u 1 in Equation ( 29 ) is the required one obtained through State Feedback control technique which gives the number of infected individuals to be treated. In similar way, we can design u 2 by choosing x 3 as output (30) y 2 = x 3 Taking time derivative of the Equation ( 30 ), we have (31) y 2 · = x 3 · Taking time derivative of e 3 given by Equation ( 5 ), we have (32) e · 3 = x · 3 Using the value of x · 3 given by Equation ( 3 ), we have (33) e · 3 = λ N − β x 3 x 1 − μ x 3 + ρ x 4 − u 2 . Error e 3 given by Equation ( 5 ) will converge to zero if Lyapunov candidate function of error e 3 given by Equation ( 33 ) is negative definite. For this purpose, we keep (34) λ N − β x 3 x 1 − μ x 3 + ρ x 4 − u 2 = − F 2 e 3 Equation ( 33 ) becomes (35) e · 3 = − F 2 e 3 Solving Equation ( 34 ) for u 2 , we have (36) u 2 = λ N − β x 3 x 1 − μ x 3 + ρ x 4 + F 3 e 3 The control input u 2 in Equation ( 36 ) is the required control obtained through the State Feedback control technique which gives the number of susceptible individuals to be vaccinated. 4 SIMULATION RESULTS In these simulations, we have considered the SIER model given by the Equations (2) , (3) , (4) , (5) , (6) and simulated the proposed controllers given by the Equations ( 17 ), ( 18 ), ( 29 ) and ( 36 ) in MATLAB/Simulink. In the graphs, time (year) is taken along the x ‐axis and number of individuals to be treated/vaccinated are taken along y ‐axis. Reference value of infected and susceptible classes is taken to be zero. Description of the other parameters and their values used for these simulations are given by the Table 1 . TABLE 1 Parameters and their values Sr.# Parameters Values 1 Transmission rate ( β ) 0.35/year 2 Infection rate ( ϵ ) 0.25/year 3 Disease induced rate ( α ) 0.01/year 4 Recovery rate due to prompt disease( γ ) 0.5/year 5 Resistance disease induced death rate (α 1 ) 0.0575/year 6 Resistance rate to treatment ( σ ) 0.470104/year 7 Recovery rate after second line of treatment ( δ ) 0.1106456/year 8 Rate of loss of immunity ( I b ) 0.05/year 9 Natural mortality ( μ ) 0.019896/year 4.1 For infected class Responses of the proposed controllers for the infected class of people have been shown in the Figure 3 . It has been observed that convergence time for the SFB controller and Synergistic controller are 1.5 and 20 years respectively and there is no steady state error and oscillations shown by any of the proposed controller. FIGURE 3 Comparison of infected individuals 4.2 For susceptible class Responses of the proposed controllers for susceptible class have been shown in Figure 4 . The convergence time of the SFB controller and Synergetic controller are 3 years and 1 year, respectively. FIGURE 4 Comparison of susceptible individuals 4.3 Control signal u 1 and u 2 of proposed controllers The two control signals from proposed controllers given by Equations ( 17 ) and ( 29 ) and Equations ( 18 ) and ( 36 ) are shown in Figures 5 and 6 , respectively. The control input u 1 is the signal for the treatment of the infected class which for SFB tracks infected class to zero after 1 year. The control input u 1 of the Synergetic controller tracks the infected class to zero after 1.5. The area under the curve u 1 gives the total number of infected individuals to be given treatment for the cure of TB. The control input u 2 is the vaccination of the susceptible class. As vaccination of TB is the continuous process, each control signal from proposed controller depicts continuous process of vaccination, but with different number of individuals to be vaccinated. The area under the curve u 2 gives the total number of susceptible individuals to be given vaccination.Comparison of two non‐linear controllers: SFB and Synergetic controllers for infected individuals ( x 1 ) and susceptible individuals ( x 3 ) in terms of convergence time, steady state error (SSE) is given in the Table 2 . FIGURE 5 Comparison of u 1 FIGURE 6 Comparison of u 2 TABLE 2 Example table Controller Convergence (Year) SSE ( x 1 , x 3 ) State Feedback 1 No, No Synergetic 15 No, No Controller's responses have also been checked for uncertainties or disturbances in the system. The disturbance can be in the form of increase or decrease in number of infected individuals due to migration to the infected population from nearby areas or due to the migration of infected individuals from infected population to other areas respectively. For simulation purpose, this type of disturbance is taken as Gaussian noise with mean 10 and variance 100 as shown in Figure 7 . FIGURE 7 White Gaussian noise Figures 8 and 9 exhibit responses of SFB and Synergetic controllers due to this disturbance. Both controllers respond well to the disturbance. When there is increase in the number of infected and susceptible individuals, the response of each controller says that higher number of infected and susceptible individuals would be given treatment and vaccination respectively. SFB controller is not much affected by the disturbance as it shows similar convergence time with negligible oscillations. Convergence time of Synergetic controller is also similar as before, but it takes more time to converge as compared to State Feedback controller. FIGURE 8 Comparison of infected individuals due to effect of disturbance FIGURE 9 Comparison of susceptible individuals due to effect of disturbance 4.1 For infected class Responses of the proposed controllers for the infected class of people have been shown in the Figure 3 . It has been observed that convergence time for the SFB controller and Synergistic controller are 1.5 and 20 years respectively and there is no steady state error and oscillations shown by any of the proposed controller. FIGURE 3 Comparison of infected individuals 4.2 For susceptible class Responses of the proposed controllers for susceptible class have been shown in Figure 4 . The convergence time of the SFB controller and Synergetic controller are 3 years and 1 year, respectively. FIGURE 4 Comparison of susceptible individuals 4.3 Control signal u 1 and u 2 of proposed controllers The two control signals from proposed controllers given by Equations ( 17 ) and ( 29 ) and Equations ( 18 ) and ( 36 ) are shown in Figures 5 and 6 , respectively. The control input u 1 is the signal for the treatment of the infected class which for SFB tracks infected class to zero after 1 year. The control input u 1 of the Synergetic controller tracks the infected class to zero after 1.5. The area under the curve u 1 gives the total number of infected individuals to be given treatment for the cure of TB. The control input u 2 is the vaccination of the susceptible class. As vaccination of TB is the continuous process, each control signal from proposed controller depicts continuous process of vaccination, but with different number of individuals to be vaccinated. The area under the curve u 2 gives the total number of susceptible individuals to be given vaccination.Comparison of two non‐linear controllers: SFB and Synergetic controllers for infected individuals ( x 1 ) and susceptible individuals ( x 3 ) in terms of convergence time, steady state error (SSE) is given in the Table 2 . FIGURE 5 Comparison of u 1 FIGURE 6 Comparison of u 2 TABLE 2 Example table Controller Convergence (Year) SSE ( x 1 , x 3 ) State Feedback 1 No, No Synergetic 15 No, No Controller's responses have also been checked for uncertainties or disturbances in the system. The disturbance can be in the form of increase or decrease in number of infected individuals due to migration to the infected population from nearby areas or due to the migration of infected individuals from infected population to other areas respectively. For simulation purpose, this type of disturbance is taken as Gaussian noise with mean 10 and variance 100 as shown in Figure 7 . FIGURE 7 White Gaussian noise Figures 8 and 9 exhibit responses of SFB and Synergetic controllers due to this disturbance. Both controllers respond well to the disturbance. When there is increase in the number of infected and susceptible individuals, the response of each controller says that higher number of infected and susceptible individuals would be given treatment and vaccination respectively. SFB controller is not much affected by the disturbance as it shows similar convergence time with negligible oscillations. Convergence time of Synergetic controller is also similar as before, but it takes more time to converge as compared to State Feedback controller. FIGURE 8 Comparison of infected individuals due to effect of disturbance FIGURE 9 Comparison of susceptible individuals due to effect of disturbance 5 CONCLUSION In this research work, a community‐based five state mathematical model of TB named as SEIR epidemiological, has been considered. This model is unique in the sense that it includes all the states including infected, susceptible, exposed, recovered and resistant classes. SFB and Synergetic controllers have been designed for the prevention and control of this viral disease. Asymptotic stability of the system has been proved using Lyapunov theory. The simulations for the proposed controllers have been performed in MATLAB/Simulink. From the graphs, it is clear that the SFB controller shows good behaviour in terms of the convergence time, steady state error and oscillations as compared to the proposed Synergetic controller.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2877167/

Thermodynamics of Multivalent Interactions: Influence of the Linker

This paper describes a thermodynamic analysis of multivalent interactions, with the goal of clarifying the influence of the linker on the enhancement in avidity due to multivalency. The use of multivalency represents a promising approach to inhibit undesired biological interactions, promote desired cellular responses, and control recognition events at surfaces. Several groups have synthesized multivalent ligands that are orders of magnitude more potent than the corresponding monovalent ligands. A better understanding of the theoretical basis for the large enhancements in avidity would help guide the design of more potent synthetic multivalent ligands. In particular, there has been significant controversy regarding the extent to which the loss of conformational entropy of the linker influences the enhancement in avidity due to multivalency. To help clarify this issue, we present the thermodynamic analysis of a heterodivalent ligand-receptor interaction. Our analysis helps reconcile seemingly competing theoretical analyses of multivalent binding. Our results indicate that the dependence of the free energy of multivalent binding on linker length can be weak even if there is a signficant decrease in the conformational entropy of the linker on binding. Our results are also consistent with studies demonstrating that the use of flexible linkers represents an effective strategy to design potent multivalent ligands.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3032408/

Anthrax toxin targeting of myeloid cells through the CMG2 receptor is essential for establishment of Bacillus anthracis infections in mice

SUMMARY Bacillus anthracis kills through a combination of bacterial infection and toxemia. Anthrax toxin working via the CMG2 receptor mediates lethality late in infection, but its roles early in infection remain unclear. We generated myeloid-lineage specific CMG2-deficient mice to examine the roles of macrophages, neutrophils, and other myeloid cells in anthrax pathogenesis. Macrophages and neutrophils isolated from these mice were resistant to anthrax toxin. However, the myeloid-specific CMG2-deficient mice remained fully sensitive to both anthrax lethal and edema toxins, demonstrating that targeting of myeloid cells is not responsible for anthrax toxin-induced lethality. Surprisingly, the myeloid-specific CMG2-deficient mice were completely resistant to B. anthracis infection. Neutrophil depletion experiments suggest that B. anthracis relies on anthrax toxin secretion to evade the scavenging functions of neutrophils to successfully establish infection. This work demonstrates that anthrax toxin uptake through CMG2 and the resulting impairment of myeloid cells specifically neutrophils, is essential to anthrax infection.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4215998/

Cloning of Insertion Site Flanking Sequence and Construction of Transfer DNA Insert Mutant Library in Stylosanthes Colletotrichum

Stylosanthes sp. is the most important forage legume in tropical areas worldwide. Stylosanthes anthracnose, which is mainly caused by Colletotrichum gloeosporioides , is a globally severe disease in stylo production. Little progress has been made in anthracnose molecular pathogenesis research. In this study, Agrobacterium tumefaciens -mediated transformation was used to transform Stylosanthes colletotrichum strain CH008. The major factors of the genetic transformation system of S. colletotrichum were optimized as follows: A. tumefaciens ' AGL-1 concentration (OD 600 ), 0.8; concentration of Colletotrichum conidium, 1×10 6 conidia/mL; acetosyringone concentration, 100 mmol/L; induction time, 6 h; co-culture temperature, 25°C; and co-culture time, 3 d. Thus, the transformation efficiency was increased to 300–400 transformants per 106 conidia. Based on the optimized system, a mutant library containing 4616 mutants was constructed, from which some mutants were randomly selected for analysis. Results show that the mutants were single copies that could be stably inherited. The growth rate, spore amount, spore germination rate, and appressorium formation rate in some mutants were significantly different from those in the wild-type strain. We then selected the most appropriate method for the preliminary screening and re-screening of each mutant's pathogenic defects. We selected 1230 transformants, and obtained 23 strains with pathogenic defects, namely, 18 strains with reduced pathogenicity and five strains with lost pathogenicity. Thermal asymmetric interlaced PCR was used to identify the transfer DNA (T-DNA) integration site in the mutant that was coded 2430, and a sequence of 476 bp was obtained. The flanking sequence of T-DNA was compared with the Colletotrichum genome by BLAST, and a sequence of 401 bp was found in Contig464 of the Colletotrichum genome. By predicting the function of the flanking sequence, we discovered that T-DNA insertion in the promoter region of the putative gene had 79% homology with the aspartate aminotransferase gene in Magnaporthe oryzae (XP_003719674.1). Introduction Stylosanthes guianensis , a diverse tropical and subtropical forage legume, is native to South America, Central America, and Africa. It is used for grazing cattle and raising livestock. Species of Stylosanthes are used for soil improvement through nitrogen fixation, reclaiming degraded wastelands, and water and soil conservation [1] – [3] . Introduction of Stylosanthes sp. to China from Australia, Africa, and South America began in the late 1960s and has continued to the present. Stylosanthes sp. is principally grown in Hainan and Guangdong Provinces as an annual crop for cut-and-carry forage, leaf meal, and hay [4] . Anthracnose of Stylosanthes , mainly caused by Colletotrichum gloeosporioides (Penz.) Penz. & Sacc., has been the most significant biotic factor limiting the production, persistence, and utilization of Stylosanthes in several countries [5] . The fungus initially infects leaves via an appressorium that develops from the germinating spore on the plant surface, followed by turgor-driven penetration of the cuticle. Fungal colonization on the leaf tissue follows and is associated with host cell necrosis, leading to a blight-like symptom and the formation of spore masses as acervuli [6] , [7] . Research in Australia, Colombia, Brazil, and China has identified two biotypes of C. gloeosporioides infecting Stylosanthes sp. [8] – [13] . Similarly, biotypes A and B and putative biotype C from Africa have been described [14] . The diversity among strains pathogenic on Stylosanthes and their relationship with other strains were analyzed at the molecular level using various markers, such as dsRNA [15] , RFLP [16] , [17] , RAPD [14] , [18] – [20] , and ITS [21] . The diversity among the pathogen population from Brazil, Colombia, China, and India is extensive [19] . Agrobacterium tumefaciens -mediated transformation (ATMT) has been used to identify mutants of C. gloeosporioides impaired in pathogenicity to gain more insight into the molecular mechanisms of C. gloeosporioides pathogenesis [22] . ATMT is a suitable and efficient technique for insertion mutagenesis, genetic mapping, and related research in filamentous fungi [23] . ATMT has been used to transform over 50 different fungal species since it was first reported [24] . The advantages of ATMT are as follows: first, A. tumefaciens directly transforms fungal spores, hyphae, or tissues without protoplast preparation; second, the integration of transfer DNA (T-DNA) into the chromosome is random and generally involves a single copy, which can easily isolate and identify the insertion locus; third, ATMT is competent for the transformation of high-molecular-weight exogenous DNA [25] , [26] . Agrobacterium -mediated T-DNA tagging has been developed as a powerful tool for both random and targeted gene disruption; it is increasingly being regarded as the system of choice for many fungi [24] . Agrobacterium -mediated T-DNA tagging is a high-throughput system for identifying and analyzing novel genes [27] – [29] , and the key for its success is the discovery of T-DNA-inserted mutants with altered phenotypes. Traditional control measures for Stylosanthes anthracnose mainly involve chemical prevention and agricultural measure control. The use of pesticide produces the most direct effect, but it may easily cause a series of problems, such as pollution of the ecological environment. Theoretically, the most cost-effective and efficient control method is the cultivation of disease-resistant S. guianensis varieties. However, this method requires intensive manpower, material resources, and time. A single disease-resistant variety of S. guianensis cannot overcome the diversity and variability of Stylosanthes Colletotrichum . The resistance of cultivated disease-resistant varieties can only keep about 5a because Colletotrichum suffers from variation easily. Therefore, exploration on pathogenesis and causes for variation in Stylosanthes colletotrichum at the levels of molecular biology and functional genomics is of scientific significance and application value to effectively cultivate new varieties of S. guianensis with long-term disease resistance, and formulate permanent strategies for the reasonable control of Stylosanthes anthracnose. Based on the constructed anthracnose genetic transformation system of S. guianensis , this study aimed to clone genes related to pathopoiesia of pathogenic bacteria. In this study, we utilized and inserted A. tumefaciens -mediated T-DNA, which contains an anti-chlorimuronethyl gene, into genes of Stylosanthes colletotrichum gloeosporiodes Penz strain CH008 with strong pathogenicity to generate insertion mutations, construct a library of mutants from Stylosanthes colletotrichum gloeosporiodes Penz strain CH008, and provide many mutation materials for future studies on functional genes analysis. Based on the selection of mutants related to pathogenicity in the mutant library, PCR and Southern blot were used for molecular verification. Flanking sequences of T-DNA insertion in virulence genes were obtained via thermal asymmetric interlaced PCR (TAIL-PCR). However, sequencing work on the whole genome sequence of S. colletotrichum remains unfinished. To predict information such as related gene functions of pathopoiesia, we compared the flanking sequences with the Colletotrichum gene libraries of known sequences, and used BLAST to analyze the homologous sequences. The results of this study could provide a basis for further investigations on functions of disrupted genes. Materials and Methods No specific permission was required for the sampling locations of this study. Moreover, ethical approval for this study was not required because we did not handle or collect animals involved in any animal welfare regulations, and no endangered or protected species were used as samples in the experiments. The study was conducted in Institute of Tropical Bioscience and Biotechnology, Chinese Academy of Tropical Agricultural Sciences (CATAS), Haikou, Hainan Province. Materials Bacterial strains and plasmid for tests The bacterial strain used was A. tumefaciens AGL-1. The bacterium included a pSULF.gfp plasmid, which uses pCAMBIAl300 as a framework and contains the ILV1 gene of chlorimuronethyl resistance marker and binary carriers of reporter gene GFP. The strain was constructed by the Sainsbury Laboratory of John Innes Center in Norwich, Britain [30] , and donated by Professor He Zhaozu of Hainan University. The recipient bacterium was collected from main flower and grass planting areas in China. C olletotrichum strain CH008 with strong pathogenicity was obtained via purification and single spore isolation [20] . Culture medium PDA and LB were prepared by conventional methods. Details about the preparation of minimal medium (MM), induction medium (IM), and selective medium can be found in the literature [31] , [32] . S. guianensis for tests Highly sensitive S. guianensis variant AFT3309 was collected from the Forage Germplasm Garden of CATAS. The collected leaves were seven-day-old ternate compound leaves. Each compound leaf contained three small leaves. One small leaf was connected to a wild-type strain with strong pathogenicity, whereas the other two were connected to a transformant strain. Methods Construction of T-DNA insertion mutant library of S. colletotrichum ATMT for S. colletotrichum was performed according to the methods of Hu XW [33] and Lin CH [34] . Determination of optimum working concentration of antibiotics (chlorimuronethyl). DCM plates with various concentrations of chlorimuronethyl (0, 5, 10, 15, 20, 30, 40, 50, and 60 µg/mL) were prepared. Approximately 1 µL of activated spore suspension of S. colletotrichum strain CH008 (concentration, 10 6 spores/mL) was dropped into the center of DCM containing various concentrations of chlorimuronethyl (each concentration gradient was investigated in triplicate), and cultivated at 28°C for 5 d to observe the growth of C. gloeosporiodes Penz and determine the working concentration. Effects of acceptor materials on transformation . We used six different levels (0.2, 0.4, 0.6, 0.8, 1.0, and 1.2) for the OD 600 value of A. tumefaciens , and four different levels (10 4 , 10 5 , 10 6 , and 10 7 spores/mL) of S. colletotrichum gloeosporiodes spore fluid concentration. The effects of 24 different combinations on the transformation efficiency were examined. Experiments were performed in triplicate. Effects of Agrobacterium induction time on transformation. The effects of seven different induction times of Agrobacterium (3, 5, 6, 7, 8, 10, and 12 h) on the transformation efficiency were determined. Experiments were performed in triplicate. Effects of acetosyringone (AS) concentrations on transformation. The effects of four different levels of AS concentration (0, 100, 150, and 200 µmol/L) on the conversion efficiency were determined. Experiments were performed in triplicate. Effects of co-culture time on transformation. The effects of seven different co-culture times (1, 2, 3, 4, 5, 6, and 7 d) on the conversion efficiency were determined. Experiments were performed in triplicate. Co-culture was conducted following the methods of Merer [35] . Effects of co-culture temperature on transformation during co-culture. The effects of five different temperatures (20°C, 22°C, 25°C, 26°C, and 28°C) on the conversion efficiency were determined. Experiments were performed in triplicate. Co-culture was performed according to the methods of Merer [35] . Verification of transformants. A. PCR Total DNA of S. colletotrichum gloeosporiodes Penz transformants was extracted according to the conventional hexadecyltrimethylammonium bromide (CTAB) method. According to the GFP gene sequence in the plasmid pSuLF·GFP, primer pairs were designed. The primers were as follows: GFP-F: 5′-TACTGCAGATGGTGAGCAAGGGCGAG-3′ GFP-R: 5′-CGGGATCCCTTGTACAGCTCGTCCATG-3′ For PCR reaction, a 20 µL system was used. The following reaction conditions were used: pre-degeneration at 94°C for 3 min, degeneration at 94°C for 45 s, annealing at 58°C for 45 s, extension at 72°C for 1 min, and final extension at 72°C for 10 min after 30 cycles. Samples were stored at 10°C. B. Fluorescence microscopy test Thirty transformants from the mutant library were randomly selected and cultivated on PDA with illumination at 28°C. Sterile water (1 mL) was dropped on the bacterial colony, and a pipette was used to carefully blow and mix the colony. Subsequently, 3 µL of the bacterial colony was placed on a clean glass slide. The slide was covered with a cover slip, and examined by confocal fluorescence microscopy. An excitation wavelength of 400–500 nm was used. Analysis of T-DNA insertion mutant library of S. colletotrichum Analysis of the copy number of T-DNA insertion mutants (Southern hybridization) Mycelia of Colletotrichum mutants were inoculated to PDA/CM fluid culture medium containing 200 µg/mL cephalosporin, 50 µg/mL tetracycline, and 10 µg/mL chlorimuronethyl. The medium was shaken at 28°C and 150 rpm for 7 d. Filter paper was used for filtration, and mycelial pellets were collected. Total DNA of S. colletotrichum gloeosporiodes transformants was extracted according to the conventional CTAB method. A detail protocol of Southern blot was performed following the specifications of DIG-High Prime DNA Labeling and Detection Starter Kit I. Analysis of the growth rate of mutants Purified untransformed CH008 was cultivated, and 30 transformants were randomly selected and cultivated on PDA plates for 5 d. The concentration of spore fluid was adjusted to 10 4 spores/mL. A pipette was used to extract 2 µL of spore fluid, which was inoculated onto a PDA plate. The plate was incubated at 28°C. After 4 d of cultivation, the colony diameter was measured once a day. The difference in colony diameter of two adjacent days was determined to represent the colony growth rate, and each strain was analyzed three times. The growth of colonies was observed for 9 d, and photos of their morphology were taken. Sporulation quantity and spore morphology of mutants Thirty transformants and untransformed CH008 were cultivated according to the methods specified above. After 6 d, sterile water was used to dilute spore fluid. A blood counting chamber was used for counting. Spore morphology was observed, and differences in spore morphology were recorded. Conidial germination and appressorium formation of mutants Fresh conidia suspension liquid was obtained using a water washing method. Its concentration was adjusted to 1.0×10 5 and 1.0×10 6 conidia/mL. Suspension liquid was dropped to a clean glass slide, blotted with a piece of bibulous filter paper, and incubated at 28°C. A total of 100 conidia were statistically analyzed, and the appressorium formation rate of germinated spores was recorded. Analyses were performed in triplicate. Samples were observed at 4, 6, 8, 10, and 12 h. Selection of pathogenicity-defective transformants of S. colletotrichum mutant library and analysis of flanking sequences at T-DNA insertion sites Preliminary screening and re-screening of pathogenic defects of mutants For preliminary screening, two inoculation methods, namely, spore fluid and mycelium cake, were used. In the spore fluid method, sterile water was used to wash Colletotrichum conidium cultivated on PDA for 5 d to prepare 5×10 5 CFU/mL suspension liquid. Conidium liquid (1 µL) was dropped on S. guianensis leaves with punctured parts and normal ones. Mycelium cakes at the edge of Colletotrichum bacterial colony were cultivated on PDA for 5 d (diameter, 5 mm). The surface of a mycelium cake, which contained hyphae, was attached to S. guianensis leaves with punctured parts and normal ones to observe morbidity. This experiment was performed in triplicate. The re-screening method was similar to preliminary screening, except that it was performed using potted and complete plants. According to the comparison of different inoculation methods, the most appropriate one was selected and applied for preliminary screening and re-screening of the pathogenic defects of each mutant. Cloning of flanking sequences at T-DNA insertion sites of mutant strains with pathogenic defects TAIL-PCR was used for amplification of the flanking sequences at T-DNA insertion sites. Random primers, composition of nested primers at left and right boundaries, and PCR procedures were based on the methods of Mullins [36] . After the PCR products accepted a 1.0% AGE test, cloning and sequencing were conducted. BLAST comparison was implemented for the obtained sequence. Materials Bacterial strains and plasmid for tests The bacterial strain used was A. tumefaciens AGL-1. The bacterium included a pSULF.gfp plasmid, which uses pCAMBIAl300 as a framework and contains the ILV1 gene of chlorimuronethyl resistance marker and binary carriers of reporter gene GFP. The strain was constructed by the Sainsbury Laboratory of John Innes Center in Norwich, Britain [30] , and donated by Professor He Zhaozu of Hainan University. The recipient bacterium was collected from main flower and grass planting areas in China. C olletotrichum strain CH008 with strong pathogenicity was obtained via purification and single spore isolation [20] . Culture medium PDA and LB were prepared by conventional methods. Details about the preparation of minimal medium (MM), induction medium (IM), and selective medium can be found in the literature [31] , [32] . S. guianensis for tests Highly sensitive S. guianensis variant AFT3309 was collected from the Forage Germplasm Garden of CATAS. The collected leaves were seven-day-old ternate compound leaves. Each compound leaf contained three small leaves. One small leaf was connected to a wild-type strain with strong pathogenicity, whereas the other two were connected to a transformant strain. Bacterial strains and plasmid for tests The bacterial strain used was A. tumefaciens AGL-1. The bacterium included a pSULF.gfp plasmid, which uses pCAMBIAl300 as a framework and contains the ILV1 gene of chlorimuronethyl resistance marker and binary carriers of reporter gene GFP. The strain was constructed by the Sainsbury Laboratory of John Innes Center in Norwich, Britain [30] , and donated by Professor He Zhaozu of Hainan University. The recipient bacterium was collected from main flower and grass planting areas in China. C olletotrichum strain CH008 with strong pathogenicity was obtained via purification and single spore isolation [20] . Culture medium PDA and LB were prepared by conventional methods. Details about the preparation of minimal medium (MM), induction medium (IM), and selective medium can be found in the literature [31] , [32] . S. guianensis for tests Highly sensitive S. guianensis variant AFT3309 was collected from the Forage Germplasm Garden of CATAS. The collected leaves were seven-day-old ternate compound leaves. Each compound leaf contained three small leaves. One small leaf was connected to a wild-type strain with strong pathogenicity, whereas the other two were connected to a transformant strain. Methods Construction of T-DNA insertion mutant library of S. colletotrichum ATMT for S. colletotrichum was performed according to the methods of Hu XW [33] and Lin CH [34] . Determination of optimum working concentration of antibiotics (chlorimuronethyl). DCM plates with various concentrations of chlorimuronethyl (0, 5, 10, 15, 20, 30, 40, 50, and 60 µg/mL) were prepared. Approximately 1 µL of activated spore suspension of S. colletotrichum strain CH008 (concentration, 10 6 spores/mL) was dropped into the center of DCM containing various concentrations of chlorimuronethyl (each concentration gradient was investigated in triplicate), and cultivated at 28°C for 5 d to observe the growth of C. gloeosporiodes Penz and determine the working concentration. Effects of acceptor materials on transformation . We used six different levels (0.2, 0.4, 0.6, 0.8, 1.0, and 1.2) for the OD 600 value of A. tumefaciens , and four different levels (10 4 , 10 5 , 10 6 , and 10 7 spores/mL) of S. colletotrichum gloeosporiodes spore fluid concentration. The effects of 24 different combinations on the transformation efficiency were examined. Experiments were performed in triplicate. Effects of Agrobacterium induction time on transformation. The effects of seven different induction times of Agrobacterium (3, 5, 6, 7, 8, 10, and 12 h) on the transformation efficiency were determined. Experiments were performed in triplicate. Effects of acetosyringone (AS) concentrations on transformation. The effects of four different levels of AS concentration (0, 100, 150, and 200 µmol/L) on the conversion efficiency were determined. Experiments were performed in triplicate. Effects of co-culture time on transformation. The effects of seven different co-culture times (1, 2, 3, 4, 5, 6, and 7 d) on the conversion efficiency were determined. Experiments were performed in triplicate. Co-culture was conducted following the methods of Merer [35] . Effects of co-culture temperature on transformation during co-culture. The effects of five different temperatures (20°C, 22°C, 25°C, 26°C, and 28°C) on the conversion efficiency were determined. Experiments were performed in triplicate. Co-culture was performed according to the methods of Merer [35] . Verification of transformants. A. PCR Total DNA of S. colletotrichum gloeosporiodes Penz transformants was extracted according to the conventional hexadecyltrimethylammonium bromide (CTAB) method. According to the GFP gene sequence in the plasmid pSuLF·GFP, primer pairs were designed. The primers were as follows: GFP-F: 5′-TACTGCAGATGGTGAGCAAGGGCGAG-3′ GFP-R: 5′-CGGGATCCCTTGTACAGCTCGTCCATG-3′ For PCR reaction, a 20 µL system was used. The following reaction conditions were used: pre-degeneration at 94°C for 3 min, degeneration at 94°C for 45 s, annealing at 58°C for 45 s, extension at 72°C for 1 min, and final extension at 72°C for 10 min after 30 cycles. Samples were stored at 10°C. B. Fluorescence microscopy test Thirty transformants from the mutant library were randomly selected and cultivated on PDA with illumination at 28°C. Sterile water (1 mL) was dropped on the bacterial colony, and a pipette was used to carefully blow and mix the colony. Subsequently, 3 µL of the bacterial colony was placed on a clean glass slide. The slide was covered with a cover slip, and examined by confocal fluorescence microscopy. An excitation wavelength of 400–500 nm was used. Construction of T-DNA insertion mutant library of S. colletotrichum ATMT for S. colletotrichum was performed according to the methods of Hu XW [33] and Lin CH [34] . Determination of optimum working concentration of antibiotics (chlorimuronethyl). DCM plates with various concentrations of chlorimuronethyl (0, 5, 10, 15, 20, 30, 40, 50, and 60 µg/mL) were prepared. Approximately 1 µL of activated spore suspension of S. colletotrichum strain CH008 (concentration, 10 6 spores/mL) was dropped into the center of DCM containing various concentrations of chlorimuronethyl (each concentration gradient was investigated in triplicate), and cultivated at 28°C for 5 d to observe the growth of C. gloeosporiodes Penz and determine the working concentration. Effects of acceptor materials on transformation . We used six different levels (0.2, 0.4, 0.6, 0.8, 1.0, and 1.2) for the OD 600 value of A. tumefaciens , and four different levels (10 4 , 10 5 , 10 6 , and 10 7 spores/mL) of S. colletotrichum gloeosporiodes spore fluid concentration. The effects of 24 different combinations on the transformation efficiency were examined. Experiments were performed in triplicate. Effects of Agrobacterium induction time on transformation. The effects of seven different induction times of Agrobacterium (3, 5, 6, 7, 8, 10, and 12 h) on the transformation efficiency were determined. Experiments were performed in triplicate. Effects of acetosyringone (AS) concentrations on transformation. The effects of four different levels of AS concentration (0, 100, 150, and 200 µmol/L) on the conversion efficiency were determined. Experiments were performed in triplicate. Effects of co-culture time on transformation. The effects of seven different co-culture times (1, 2, 3, 4, 5, 6, and 7 d) on the conversion efficiency were determined. Experiments were performed in triplicate. Co-culture was conducted following the methods of Merer [35] . Effects of co-culture temperature on transformation during co-culture. The effects of five different temperatures (20°C, 22°C, 25°C, 26°C, and 28°C) on the conversion efficiency were determined. Experiments were performed in triplicate. Co-culture was performed according to the methods of Merer [35] . Verification of transformants. A. PCR Total DNA of S. colletotrichum gloeosporiodes Penz transformants was extracted according to the conventional hexadecyltrimethylammonium bromide (CTAB) method. According to the GFP gene sequence in the plasmid pSuLF·GFP, primer pairs were designed. The primers were as follows: GFP-F: 5′-TACTGCAGATGGTGAGCAAGGGCGAG-3′ GFP-R: 5′-CGGGATCCCTTGTACAGCTCGTCCATG-3′ For PCR reaction, a 20 µL system was used. The following reaction conditions were used: pre-degeneration at 94°C for 3 min, degeneration at 94°C for 45 s, annealing at 58°C for 45 s, extension at 72°C for 1 min, and final extension at 72°C for 10 min after 30 cycles. Samples were stored at 10°C. B. Fluorescence microscopy test Thirty transformants from the mutant library were randomly selected and cultivated on PDA with illumination at 28°C. Sterile water (1 mL) was dropped on the bacterial colony, and a pipette was used to carefully blow and mix the colony. Subsequently, 3 µL of the bacterial colony was placed on a clean glass slide. The slide was covered with a cover slip, and examined by confocal fluorescence microscopy. An excitation wavelength of 400–500 nm was used. Analysis of T-DNA insertion mutant library of S. colletotrichum Analysis of the copy number of T-DNA insertion mutants (Southern hybridization) Mycelia of Colletotrichum mutants were inoculated to PDA/CM fluid culture medium containing 200 µg/mL cephalosporin, 50 µg/mL tetracycline, and 10 µg/mL chlorimuronethyl. The medium was shaken at 28°C and 150 rpm for 7 d. Filter paper was used for filtration, and mycelial pellets were collected. Total DNA of S. colletotrichum gloeosporiodes transformants was extracted according to the conventional CTAB method. A detail protocol of Southern blot was performed following the specifications of DIG-High Prime DNA Labeling and Detection Starter Kit I. Analysis of the growth rate of mutants Purified untransformed CH008 was cultivated, and 30 transformants were randomly selected and cultivated on PDA plates for 5 d. The concentration of spore fluid was adjusted to 10 4 spores/mL. A pipette was used to extract 2 µL of spore fluid, which was inoculated onto a PDA plate. The plate was incubated at 28°C. After 4 d of cultivation, the colony diameter was measured once a day. The difference in colony diameter of two adjacent days was determined to represent the colony growth rate, and each strain was analyzed three times. The growth of colonies was observed for 9 d, and photos of their morphology were taken. Sporulation quantity and spore morphology of mutants Thirty transformants and untransformed CH008 were cultivated according to the methods specified above. After 6 d, sterile water was used to dilute spore fluid. A blood counting chamber was used for counting. Spore morphology was observed, and differences in spore morphology were recorded. Conidial germination and appressorium formation of mutants Fresh conidia suspension liquid was obtained using a water washing method. Its concentration was adjusted to 1.0×10 5 and 1.0×10 6 conidia/mL. Suspension liquid was dropped to a clean glass slide, blotted with a piece of bibulous filter paper, and incubated at 28°C. A total of 100 conidia were statistically analyzed, and the appressorium formation rate of germinated spores was recorded. Analyses were performed in triplicate. Samples were observed at 4, 6, 8, 10, and 12 h. Analysis of the copy number of T-DNA insertion mutants (Southern hybridization) Mycelia of Colletotrichum mutants were inoculated to PDA/CM fluid culture medium containing 200 µg/mL cephalosporin, 50 µg/mL tetracycline, and 10 µg/mL chlorimuronethyl. The medium was shaken at 28°C and 150 rpm for 7 d. Filter paper was used for filtration, and mycelial pellets were collected. Total DNA of S. colletotrichum gloeosporiodes transformants was extracted according to the conventional CTAB method. A detail protocol of Southern blot was performed following the specifications of DIG-High Prime DNA Labeling and Detection Starter Kit I. Analysis of the growth rate of mutants Purified untransformed CH008 was cultivated, and 30 transformants were randomly selected and cultivated on PDA plates for 5 d. The concentration of spore fluid was adjusted to 10 4 spores/mL. A pipette was used to extract 2 µL of spore fluid, which was inoculated onto a PDA plate. The plate was incubated at 28°C. After 4 d of cultivation, the colony diameter was measured once a day. The difference in colony diameter of two adjacent days was determined to represent the colony growth rate, and each strain was analyzed three times. The growth of colonies was observed for 9 d, and photos of their morphology were taken. Sporulation quantity and spore morphology of mutants Thirty transformants and untransformed CH008 were cultivated according to the methods specified above. After 6 d, sterile water was used to dilute spore fluid. A blood counting chamber was used for counting. Spore morphology was observed, and differences in spore morphology were recorded. Conidial germination and appressorium formation of mutants Fresh conidia suspension liquid was obtained using a water washing method. Its concentration was adjusted to 1.0×10 5 and 1.0×10 6 conidia/mL. Suspension liquid was dropped to a clean glass slide, blotted with a piece of bibulous filter paper, and incubated at 28°C. A total of 100 conidia were statistically analyzed, and the appressorium formation rate of germinated spores was recorded. Analyses were performed in triplicate. Samples were observed at 4, 6, 8, 10, and 12 h. Selection of pathogenicity-defective transformants of S. colletotrichum mutant library and analysis of flanking sequences at T-DNA insertion sites Preliminary screening and re-screening of pathogenic defects of mutants For preliminary screening, two inoculation methods, namely, spore fluid and mycelium cake, were used. In the spore fluid method, sterile water was used to wash Colletotrichum conidium cultivated on PDA for 5 d to prepare 5×10 5 CFU/mL suspension liquid. Conidium liquid (1 µL) was dropped on S. guianensis leaves with punctured parts and normal ones. Mycelium cakes at the edge of Colletotrichum bacterial colony were cultivated on PDA for 5 d (diameter, 5 mm). The surface of a mycelium cake, which contained hyphae, was attached to S. guianensis leaves with punctured parts and normal ones to observe morbidity. This experiment was performed in triplicate. The re-screening method was similar to preliminary screening, except that it was performed using potted and complete plants. According to the comparison of different inoculation methods, the most appropriate one was selected and applied for preliminary screening and re-screening of the pathogenic defects of each mutant. Cloning of flanking sequences at T-DNA insertion sites of mutant strains with pathogenic defects TAIL-PCR was used for amplification of the flanking sequences at T-DNA insertion sites. Random primers, composition of nested primers at left and right boundaries, and PCR procedures were based on the methods of Mullins [36] . After the PCR products accepted a 1.0% AGE test, cloning and sequencing were conducted. BLAST comparison was implemented for the obtained sequence. Preliminary screening and re-screening of pathogenic defects of mutants For preliminary screening, two inoculation methods, namely, spore fluid and mycelium cake, were used. In the spore fluid method, sterile water was used to wash Colletotrichum conidium cultivated on PDA for 5 d to prepare 5×10 5 CFU/mL suspension liquid. Conidium liquid (1 µL) was dropped on S. guianensis leaves with punctured parts and normal ones. Mycelium cakes at the edge of Colletotrichum bacterial colony were cultivated on PDA for 5 d (diameter, 5 mm). The surface of a mycelium cake, which contained hyphae, was attached to S. guianensis leaves with punctured parts and normal ones to observe morbidity. This experiment was performed in triplicate. The re-screening method was similar to preliminary screening, except that it was performed using potted and complete plants. According to the comparison of different inoculation methods, the most appropriate one was selected and applied for preliminary screening and re-screening of the pathogenic defects of each mutant. Cloning of flanking sequences at T-DNA insertion sites of mutant strains with pathogenic defects TAIL-PCR was used for amplification of the flanking sequences at T-DNA insertion sites. Random primers, composition of nested primers at left and right boundaries, and PCR procedures were based on the methods of Mullins [36] . After the PCR products accepted a 1.0% AGE test, cloning and sequencing were conducted. BLAST comparison was implemented for the obtained sequence. Results Construction of T-DNA insertion mutant library of S. colletotrichum Optimum concentration of chlorimuronethyl In accordance with Figure 1 , S. colletotrichum CH008 was inhibited at 5 µg/mL chlorimuronethyl and this strain nearly showed no growth at 10 µg/mL chlorimuronethyl. Thus, the final concentration selected by the culture medium was 10 (preliminary screening) and 20 µg/mL (re-screening). 10.1371/journal.pone.0111172.g001 Figure 1 Tolerance test of a wild-type strain on medium containing concentration gradients of chlorimuronethyl. Note (from the top, clockwise): DCM plates (concentrations of 0, 5, 10, 20, and 30 µg/mL); cultivation was carried out at 28°C for 4 d. Effects of acceptor materials on transformation Results of the differences in concentrations of A. tumefaciens and Colletotrichum spore liquid were analyzed by ANOVA (similar to Duncan's new multiple range method below). According to the related results, the difference in experimental results was significant at p0.01), whereas different capital letters indicate significant differences (p0.01), whereas different capital letters indicate significant differences (p0.01), whereas different capital letters indicate significant differences (p<0.01). Effects of induction time on transformation By inducing the activation and expression of genes at the Vir region of Agrobacterium , AS promoted T-DNA processing and transfer so that Agrobacterium T-DNA could enter the target genome and integrate with it more easily. An appropriate induction time was directly related to the ability of AS to activate the Vir region sufficiently, and affected the efficiency of final recombination. The effect was most significant in this experiment, and the transformation efficiency peaked at 6 h of induction and transformation ( Figure 3 ). However, extending the induction time did not improve the transformation efficiency. This finding may be due to the fact that the culture time of Agrobacterium was too long, so Agrobacterium gradually entered a decline phase and affected the transformation efficiency. 10.1371/journal.pone.0111172.g003 Figure 3 Effect of induction time on transformation efficiency. Note: Different capital letters indicate significant differences (p<0.01). Effects of AS on transformation When the AS concentration was 100 µmol/L ( Figure 4 ), the results were significantly higher than those at other AS levels. However, the transformation efficiency did not improve as the concentration increased. 10.1371/journal.pone.0111172.g004 Figure 4 Effect of AS concentrations on transformation efficiency. Note: Different capital letters indicate significant differences (p<0.01). Effects of co-culture time on transformation According to Figure 5 , the transformation efficiency peaked when S. colletotrichum and Agrobacterium were co-cultured for 3–4 d, and the results were highly significant. 10.1371/journal.pone.0111172.g005 Figure 5 Effect of co-culture time on transformation efficiency. Note: Different capital letters indicate significant differences (p<0.01). Effects of temperature on transformation during co-culture In this experiment ( Figure 6 ), the optimum transformation efficiency was obtained at a co-culture temperature of 25°C. Its effect was most significant at this temperature. 10.1371/journal.pone.0111172.g006 Figure 6 Effect of co-culture temperature on transformation efficiency. Note: Different capital letters indicate significant differences (p<0.01). PCR Among 20 selected converter strains, the positive control and all transformants demonstrated bright and clear strips at 750 bp. Thus, the T-DNA insertion rate was 100% ( Figure 7 ). 10.1371/journal.pone.0111172.g007 Figure 7 Verification of PCR results. Note: Marker v is on the left, followed by the positive control of plasmid and transformants, and untransformed CH 008 is on the right. Verification of fluorescence of transformant's conidia Thirty transformants were randomly selected. After sporulation, conidia were examined under confocal fluorescence microscopy at an excitation wavelength of 400–500 nm ( Figure 8 ). Spores of 13 transformants were bright under fluorescence, which indicates that GFP genes were carried to and integrated with S. colletotrichum gloeosporiodes Penz genome, and exhibited good expression. 10.1371/journal.pone.0111172.g008 Figure 8 GFP fluorescence of S. colletotrichum T-DNA transformants. This study optimized the genetic transformation system conditions of S. colletotrichum . The optimized conditions were as follows: AGL-1 concentration of A. tumefaciens (OD 600 ), 0.8; concentration of Colletotrichum conidium, 1×10 6 conidia/mL; AS concentration, 100 mmol/L; induction time, 6 h; co-culture temperature, 25°C; and co-culture time, 3 d. The T-DNA insertion mutant library of S. colletotrichum was constructed successfully using the genetic transformation system. The transformation efficiency was determined to be 300–400 transformants/10 6 Colletotrichum spores. Analysis of S. colletotrichum T-DNA insertion mutant libraries The genetic transformation system was used to obtain 4616 S. colletotrichum genetic transformants. Some of these transformants were then selected for analysis. Analysis of the copy number of T-DNA insertion mutants The results of Southern blot are shown in Figure 9 . Among eight randomly selected transformants, six demonstrated a single strip and the remaining two had two strips. Untransformed CH 008 and one transformant had no strip. These results indicate that most T-DNA insertions had a single locus, which could aid in amplification for sequences of insertion sites via Tail-PCR. 10.1371/journal.pone.0111172.g009 Figure 9 Southern blot analysis of genomic DNA from untransformed CH008 (CK) and transformed isolates. Genetic stability After cultivating untransformed and mutant strains in PDA plates without chlorimuronethyl for 10 generations, they were inoculated to screening plates with chlorimuronethyl. The transformants grew normally ( Figure 9 ), whereas the untransformed strains did not grow (indicated by the arrow in Figure 10 ). These findings show that resistance could be stably inherited in transformants. 10.1371/journal.pone.0111172.g010 Figure 10 Determination of transformants' genetic stability (arrows indicate wild strains). Growth rate and colony morphology of transformants Untransformed CH008 was cultivated, and 30 transformants were randomly selected. Thus, we observed an obvious difference in the growth rate of transformants. Specifically, mutant 1561 had the lowest growth rate, and its growth rate was much lower than that of untransformed CH008. Transformant 328 exhibited the highest growth rate ( Table 1 ). Some strains demonstrated morphological variation, such as white hyphae that did not generate spores, greyish green hyphae, and yellow hyphae. However, the proportion of such strains was small. 10.1371/journal.pone.0111172.t001 Table 1 Comparison of the growth rate of untransformed strains and transformants. Approaching growth rate of untransformed strains High growth rate of untransformed strains Low growth rate of untransformed strains Ch008 0.80 cm/d 888 1.06 cm/d 1561 0.37 cm/d 1130 0.76 cm/d 328 1.40 cm/d 3532 0.53 cm/d 993 0.83 cm/d 2715 1.13 cm/d 678 0.40cm/d 1477 0.80 cm/d 3590 1.20cm/d 3425 0.60 cm/d 3200 0.87 cm/d 3416 1.23 cm/d 1801 0.56 cm/d 2181 0.73 cm/d 1761 1.16 cm/d 3393 0.46 cm/d 3616 0.86 cm/d 3605 1.00 cm/d 604 0.43 cm/d Note: Data in the table are average values of three measurements. Sporulation quantity of mutants The sporulation quantities of wild strains and mutants were determined, and the results are shown in Table 2 . The sporulation quantity of transformants 2181, 2881, 2561, 844, and 888 was much higher than that of untransformed CH008, whereas the sporulation quantity of transformants 1477, 1561, and 2097 was lower than that of untransformed CH008. Transformants 3443 and 993 showed nearly no sporulation quantity. 10.1371/journal.pone.0111172.t002 Table 2 Comparison of transformants' ability to produce spores. Strain number Ch008 2097 1447 1477 2181 961 2881 888 993 Spore concentration 1.05×10 6 4.00×10 4 2.15×10 5 1.00×10 6 2.15×10 7 6.03×10 4 1.67×10 7 1.50×10 5 0 Strain number 3590 3661 3605 2561 3443 844 1561 1761 604 Spore concentration 2.33×10 6 2.34×10 5 2.50×10 5 3.45×10 7 0 1.75×10 7 2.35×10 5 5.40×10 7 1.70×10 5 Conidial germination and appressorium formation in mutants The results show that wild-type conidia and 10 randomly selected transformant spores could germinate after some time. The time that most of the transformants took to germinate was shorter than that of the wild-type strains, and the average length of germ tube growth in some transformants was longer than that of the wild-type strains. Three mutant strains demonstrated both low germination and appressorium formation rates, namely, t-960, t-604, and t-2327. By contrast, seven mutant spores (t-2393, t-2515, t-906, t-888, t-1130, t-2416, and t-3616) showed no significant difference from wild-type strains. In addition, wild-type strains and most transformants grew one to two germinal tubes from both ends of spores, and 80% of them demonstrated appressorium formation. Spore germination of mutant t-906 was abnormal; this mutant had three to four germinal tubes and its germination rate was high, but it germinated few appressoria ( Table 3 and Figure 11 ). 10.1371/journal.pone.0111172.g011 Figure 11 Conidial germination and appressorium formation after 8 h. 10.1371/journal.pone.0111172.t003 Table 3 Comparison of conidial germination and germination rate. Strain name Initial germination time Germination rate Length and number of germinal tubes Ch008 8 h 90–100 55–110 µm, 1 t-2515 6 h 80–90 110–300 µm, 2 t-2393 8 h 70–80 200–300 µm, 2 t-2327 6 h 40–50 300–400 µm, 1 t-888 6 h 70–80 200–250 µm, 2 t-906 4 h 95–100 300–400 µm, 3–4 t-960 8 h 60–70 30–45 µm, 1 t-1130 8 h 70–80 100–255 µm, 2 t-2416 6 h 90–100 110–350 µm, 2 t-3616 6 h 70–80 20–100 µm, 1 t-604 4 h 40–60 110–300 µm, 2 Analysis of the copy number of T-DNA insertion mutants The results of Southern blot are shown in Figure 9 . Among eight randomly selected transformants, six demonstrated a single strip and the remaining two had two strips. Untransformed CH 008 and one transformant had no strip. These results indicate that most T-DNA insertions had a single locus, which could aid in amplification for sequences of insertion sites via Tail-PCR. 10.1371/journal.pone.0111172.g009 Figure 9 Southern blot analysis of genomic DNA from untransformed CH008 (CK) and transformed isolates. Genetic stability After cultivating untransformed and mutant strains in PDA plates without chlorimuronethyl for 10 generations, they were inoculated to screening plates with chlorimuronethyl. The transformants grew normally ( Figure 9 ), whereas the untransformed strains did not grow (indicated by the arrow in Figure 10 ). These findings show that resistance could be stably inherited in transformants. 10.1371/journal.pone.0111172.g010 Figure 10 Determination of transformants' genetic stability (arrows indicate wild strains). Growth rate and colony morphology of transformants Untransformed CH008 was cultivated, and 30 transformants were randomly selected. Thus, we observed an obvious difference in the growth rate of transformants. Specifically, mutant 1561 had the lowest growth rate, and its growth rate was much lower than that of untransformed CH008. Transformant 328 exhibited the highest growth rate ( Table 1 ). Some strains demonstrated morphological variation, such as white hyphae that did not generate spores, greyish green hyphae, and yellow hyphae. However, the proportion of such strains was small. 10.1371/journal.pone.0111172.t001 Table 1 Comparison of the growth rate of untransformed strains and transformants. Approaching growth rate of untransformed strains High growth rate of untransformed strains Low growth rate of untransformed strains Ch008 0.80 cm/d 888 1.06 cm/d 1561 0.37 cm/d 1130 0.76 cm/d 328 1.40 cm/d 3532 0.53 cm/d 993 0.83 cm/d 2715 1.13 cm/d 678 0.40cm/d 1477 0.80 cm/d 3590 1.20cm/d 3425 0.60 cm/d 3200 0.87 cm/d 3416 1.23 cm/d 1801 0.56 cm/d 2181 0.73 cm/d 1761 1.16 cm/d 3393 0.46 cm/d 3616 0.86 cm/d 3605 1.00 cm/d 604 0.43 cm/d Note: Data in the table are average values of three measurements. Sporulation quantity of mutants The sporulation quantities of wild strains and mutants were determined, and the results are shown in Table 2 . The sporulation quantity of transformants 2181, 2881, 2561, 844, and 888 was much higher than that of untransformed CH008, whereas the sporulation quantity of transformants 1477, 1561, and 2097 was lower than that of untransformed CH008. Transformants 3443 and 993 showed nearly no sporulation quantity. 10.1371/journal.pone.0111172.t002 Table 2 Comparison of transformants' ability to produce spores. Strain number Ch008 2097 1447 1477 2181 961 2881 888 993 Spore concentration 1.05×10 6 4.00×10 4 2.15×10 5 1.00×10 6 2.15×10 7 6.03×10 4 1.67×10 7 1.50×10 5 0 Strain number 3590 3661 3605 2561 3443 844 1561 1761 604 Spore concentration 2.33×10 6 2.34×10 5 2.50×10 5 3.45×10 7 0 1.75×10 7 2.35×10 5 5.40×10 7 1.70×10 5 Conidial germination and appressorium formation in mutants The results show that wild-type conidia and 10 randomly selected transformant spores could germinate after some time. The time that most of the transformants took to germinate was shorter than that of the wild-type strains, and the average length of germ tube growth in some transformants was longer than that of the wild-type strains. Three mutant strains demonstrated both low germination and appressorium formation rates, namely, t-960, t-604, and t-2327. By contrast, seven mutant spores (t-2393, t-2515, t-906, t-888, t-1130, t-2416, and t-3616) showed no significant difference from wild-type strains. In addition, wild-type strains and most transformants grew one to two germinal tubes from both ends of spores, and 80% of them demonstrated appressorium formation. Spore germination of mutant t-906 was abnormal; this mutant had three to four germinal tubes and its germination rate was high, but it germinated few appressoria ( Table 3 and Figure 11 ). 10.1371/journal.pone.0111172.g011 Figure 11 Conidial germination and appressorium formation after 8 h. 10.1371/journal.pone.0111172.t003 Table 3 Comparison of conidial germination and germination rate. Strain name Initial germination time Germination rate Length and number of germinal tubes Ch008 8 h 90–100 55–110 µm, 1 t-2515 6 h 80–90 110–300 µm, 2 t-2393 8 h 70–80 200–300 µm, 2 t-2327 6 h 40–50 300–400 µm, 1 t-888 6 h 70–80 200–250 µm, 2 t-906 4 h 95–100 300–400 µm, 3–4 t-960 8 h 60–70 30–45 µm, 1 t-1130 8 h 70–80 100–255 µm, 2 t-2416 6 h 90–100 110–350 µm, 2 t-3616 6 h 70–80 20–100 µm, 1 t-604 4 h 40–60 110–300 µm, 2 Selection of transformants of the S. colletotrichum mutant library with pathogenic defects and analysis of flanking sequences at T-DNA insertion sites Preliminary screening of mutants with pathogenic defects Comparison of inoculation methods showed that the most appropriate preliminary screening method was the use of the wild-type strain CH008 and transformant spore liquid to infect punctured parts of leaves of S. guianensis , and select mutants with lost pathogenicity. Using this method, we selected 1230 transformants and obtained 23 strains with pathogenic defects (18 strains with reduced pathogenicity and five strains with lost pathogenicity). Figure 12 shows that brown scabs formed at the infected part of wild strain CH008, whereas transformant 2430 lost its infection ability. 10.1371/journal.pone.0111172.g012 Figure 12 Determination of virulence during preliminary screening. Re-screening of pathogenicity determination For re-screening, we selected defective mutant strains whose pathogenicity was lost when preliminary screening was performed. The potted S. guianensis was inoculated again. Comparison of different re-screening inoculation methods showed that the most appropriate re-screening method was spraying the wild strain CH008 and transformant spore liquid to punctured parts of leaves of S. guianensis for verification and selection of mutants with lost pathogenicity. This method was used to re-screen five mutants with completely lost pathogenicity. The results were similar to those of preliminary screening. For example, mutant 2430 was used to infect S. guianensis plants. Two weeks later, pathogenic symptoms appeared. Scabs were observed on some leaves and stems of plants, and slowly expanded to most of the stem leaves ( Figure 13 ). 10.1371/journal.pone.0111172.g013 Figure 13 Incidence of different parts of plants after re-screening and inoculation. Cloning of T-DNA insertion flanking sequences of mutant strains with pathogenic defects TAIL-PCR amplification was used in 15 mutants with reduced pathogenicity and defects, and three transformants with peculiar strips were obtained, i.e., Anthrax transformants 2430, 913, and 3521. Only transformant 2430 was successful in transformation and sequencing; and the flanking sequences of 476 bp were obtained ( Figure 14 ). 10.1371/journal.pone.0111172.g014 Figure 14 Flanking sequences of T-DNA at the insertion sites of mutant 2430. BLAST was used to compare the sequence with the sequenced Colletotrichum whole genome database (unpublished), and the website 'The FGENESH Program' (Softberry Inc., Mount Kisco, NY, USA; http://linux1.Softberry.com/berry.phtml ) was used to predict its functions. Therefore, a hypothetical gene in the regional code of the nucleotide sequence was noted. BLAST was subsequently adopted to compare the sequences in NCBI; 401 nt was completely consistent with the partial sequence of Contig464 of the database ( Figure 15 ). T-DNA was a promoter subregion of the predicted gene. The full length of the predicted gene was 1251 bp. The code of the predicted gene was 416 aa, and the amino acid homology between the predicted gene and Magnaporthe oryzae gene (XP_003719674.1) was 79%. This type of gene codes aspartate transaminase. This code may play an important role in the infection process of pathogeny. Insertion of exogenous sections destroyed the gene's functions, so the mutants exhibited lost pathogenicity. 10.1371/journal.pone.0111172.g015 Figure 15 BLAST results of T-DNA flanking sequences in NCBI. Preliminary screening of mutants with pathogenic defects Comparison of inoculation methods showed that the most appropriate preliminary screening method was the use of the wild-type strain CH008 and transformant spore liquid to infect punctured parts of leaves of S. guianensis , and select mutants with lost pathogenicity. Using this method, we selected 1230 transformants and obtained 23 strains with pathogenic defects (18 strains with reduced pathogenicity and five strains with lost pathogenicity). Figure 12 shows that brown scabs formed at the infected part of wild strain CH008, whereas transformant 2430 lost its infection ability. 10.1371/journal.pone.0111172.g012 Figure 12 Determination of virulence during preliminary screening. Re-screening of pathogenicity determination For re-screening, we selected defective mutant strains whose pathogenicity was lost when preliminary screening was performed. The potted S. guianensis was inoculated again. Comparison of different re-screening inoculation methods showed that the most appropriate re-screening method was spraying the wild strain CH008 and transformant spore liquid to punctured parts of leaves of S. guianensis for verification and selection of mutants with lost pathogenicity. This method was used to re-screen five mutants with completely lost pathogenicity. The results were similar to those of preliminary screening. For example, mutant 2430 was used to infect S. guianensis plants. Two weeks later, pathogenic symptoms appeared. Scabs were observed on some leaves and stems of plants, and slowly expanded to most of the stem leaves ( Figure 13 ). 10.1371/journal.pone.0111172.g013 Figure 13 Incidence of different parts of plants after re-screening and inoculation. Cloning of T-DNA insertion flanking sequences of mutant strains with pathogenic defects TAIL-PCR amplification was used in 15 mutants with reduced pathogenicity and defects, and three transformants with peculiar strips were obtained, i.e., Anthrax transformants 2430, 913, and 3521. Only transformant 2430 was successful in transformation and sequencing; and the flanking sequences of 476 bp were obtained ( Figure 14 ). 10.1371/journal.pone.0111172.g014 Figure 14 Flanking sequences of T-DNA at the insertion sites of mutant 2430. BLAST was used to compare the sequence with the sequenced Colletotrichum whole genome database (unpublished), and the website 'The FGENESH Program' (Softberry Inc., Mount Kisco, NY, USA; http://linux1.Softberry.com/berry.phtml ) was used to predict its functions. Therefore, a hypothetical gene in the regional code of the nucleotide sequence was noted. BLAST was subsequently adopted to compare the sequences in NCBI; 401 nt was completely consistent with the partial sequence of Contig464 of the database ( Figure 15 ). T-DNA was a promoter subregion of the predicted gene. The full length of the predicted gene was 1251 bp. The code of the predicted gene was 416 aa, and the amino acid homology between the predicted gene and Magnaporthe oryzae gene (XP_003719674.1) was 79%. This type of gene codes aspartate transaminase. This code may play an important role in the infection process of pathogeny. Insertion of exogenous sections destroyed the gene's functions, so the mutants exhibited lost pathogenicity. 10.1371/journal.pone.0111172.g015 Figure 15 BLAST results of T-DNA flanking sequences in NCBI. Discussion In 2011, genetic transformation of over 60 fungi was realized successfully by ATMT [38] . When transformation is carried out, each fungus has optimum transformation conditions. The optimum transformation and screening conditions should be explored to construct a high-quality genetic mutant library with a large quantity in a short time period. Moreover, when high-quality and efficient transformation and optimization systems are used to construct necessary mutant libraries, manpower and material resources are not only saved but the library-establishment cycle is also shortened. Such systems can also reduce difficulty in screening and cloning related pathogenic genes [39] . The cultivation, growth status, and purity of Agrobacterium have significant effects on transformation, and they are important for preparing Agrobacterium infection liquid with high purity, vigorous growth, and powerful infection capability [40] . Agrobacterium in the middle and late logarithmic phases is considered to be the optimum infectious bacteria. However, the time at which different species of Agrobacterium reach the logarithmic phase, as well as their concentrations, can differ. In the present study, after AGL-1 was cultured in MM for 36 h and cultivated in IM for 6 h, its OD 600 value ranged from 0.6 to 1.0. By selecting Agrobacterium within this range and adjusting to an appropriate concentration, the obtained transformation efficiency was high. AS is currently the most common Vir gene inducer. Numerous studies showed an optimum inductive effect at a concentration of 50–200 µmol/L. When the pH of the culture medium containing AS is 5.0–5.6, the induction of genes at the Vir region of A. tumefaciens peaks [41] . IM in the present study showed a final selected concentration of 100 µmol/L and pH of 5.2. This finding verifies the conclusion drawn by Holford et al., who reported optimum transformation effects at pH 5.2 [42] . When the AS concentration was 100 µmol/L, the results were significantly higher than those at other levels. However, the transformation efficiency did not improve with increasing AS concentration. Excess AS concentrations may have toxic effects on explants, and influence further improvement in the transformation efficiency [41] . The co-culture time is one of the most important factors affecting the success of the transformation of A. tumefaciens . The process of transformation mediated by A. tumefaciens takes some time, and transformants are unable to form at a very short co-culture time [35] , [43] – [45] . In the present study, the number of transformants increased with increasing co-culture time. The number of transformants peaked when S. colletotrichum and Agrobacterium were co-cultured for 3–4 d, and the results were highly significant. Continuous co-culture can result in generation of false-positive clones [46] , resulting in very large bacterial colonies that cannot be selected easily [47] . Campoy et al. revealed that the number of transformants peaks when the co-culture temperature is consistent with the optimum growth temperature of acceptors; reducing or increasing the co-culture temperature reduces the number of transformants [48] . The results of this study were consistent with those observed by Campoy et al. Furthermore, selection of transformation acceptors is critical. Many forms, such as protoplast, mycelium, and conidium, can be used as acceptor materials [26] . Acceptor strains can be transformed by exogenous genes only when they stay in the phase of cell division. High requirements must be met in the preparation of protoplast, and this process is quite complicated. Materials such as mycelium and conidium can also achieve ideal transformation efficiency. Very high cell concentration of acceptor fungi can lead to excess fungal growth, so transformants cannot be selected [24] ; very high Agrobacterium concentrations can result in serious Agrobacterium pollution [37] . The results of Southern blot show that the T-DNA insertion rate in transformants was 90%. The transformants, whose T-DNA insertion was at a single site, accounted for about 67%. The rate of a single copy was much higher than that of Colletotrichum graminicola (16%), slightly higher than that of other anthracnose species (65%) [49] , [50] , but lower than that obtained by Jia Peisong [51] and Wang Haiyan [52] (100%). Single-site insertion in genome is usually expected because the derived phenotype is related to changes in single sites in the genome [24] , which can help in finding marker genes from the fungal genome. Some studies implied that the insertion rate of T-DNA single copy numbers has an inverse relation with the co-culture time to some extent [53] . However, the co-culture time must not be shortened too much in single-copy insertion so that the transformation efficiency does not decrease. Moreover, Examination of a large number of mutants using PCR and Southern blot is time-consuming and laborious, and generally provides only a subjective measure of infectivity. Therefore it would be worthwhile to develop the high-throughput technologies for molecular verification of Stylosanthes anthracnose mutants, such as Luminex or transposon sequencing, Lin Tao et al. examined 434 signature-tagged mutagenesis mutants using Luminex-based multiplex PCR, which is an efficient and timesaving method [54] – [56] . This research showed that the growth rates and sporulation quantity of some transformants changed. T-DNA insertion causes inactivation for a certain gene of transformants, and this variation results in changes in other characters. However, most transformants did not demonstrate a significant difference from untransformed strains. Conidial generation is a premise for infection and pathopoiesia of many pathogenic bacteria; the quantity and germination rate of conidia, as well as appressorium formation, affect their pathogenicity to some extent [57] , [58] . In the future, relations among the pathogenicity of transformants, growth rate of mutants, sporulation quantity, conidial germination, appressorium formation, and pathopoiesia will be further examined. The flanking sequences of T-DNA insertion sites were cloned. TAIL-PCR is commonly used to amplify unknown flanking sequences of the known T-DNA sequence. This technique can design reliable random degenerate primers that are appropriate for the background of genomes that need to be detected [59] . The present study identified the optimum random primers (i.e., AD4 and AD8), and four right flanking sequences and one left flanking sequence were obtained. The reason for these results might be the high probability at which the left border of T-DNA is cut off, which was consistent with several fungal results that have been reported [36] , [49] , [60] . The left border of T-DNA may be unnecessary for T-DNA transfer, whereas the right border is essential for T-DNA transfer; thus, transfer starts at the right border and continues toward the left [61] . TAIL-PCR was used in the present study to amplify the flanking sequences of five mutant genomes, and five flanking sequences were obtained. For one sequence, BLAST comparison showed that its pathogenicity was lost and the flank section amplified by t-2430 was approximately 0.5 kb. By analyzing and comparing the obtained sequences, these genes and genes of aspartate transaminase coded by the pathogenic bacterium M. oryzae demonstrated high homology. This enzyme mainly exists in the mitochondria of cells, and exerts important catalytic actions during nitrogen metabolism. The insertion of exogenous sections may alter the functions of genes, affect the coding of transferase, disturb some important metabolisms of the mutant, and result in the loss of pathogenicity in pathogenic infection processes.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7187339/

Multiphoton intravital microscopy in small animals: motion artefact challenges and technical solutions

Summary Since its invention 29 years ago, two‐photon laser‐scanning microscopy has evolved from a promising imaging technique, to an established widely available imaging modality used throughout the biomedical research community. The establishment of two‐photon microscopy as the preferred method for imaging fluorescently labelled cells and structures in living animals can be attributed to the biophysical mechanism by which the generation of fluorescence is accomplished. The use of powerful lasers capable of delivering infrared light pulses within femtosecond intervals, facilitates the nonlinear excitation of fluorescent molecules only at the focal plane and determines by objective lens position. This offers numerous benefits for studies of biological samples at high spatial and temporal resolutions with limited photo‐damage and superior tissue penetration. Indeed, these attributes have established two‐photon microscopy as the ideal method for live‐animal imaging in several areas of biology and have led to a whole new field of study dedicated to imaging biological phenomena in intact tissues and living organisms. However, despite its appealing features, two‐photon intravital microscopy is inherently limited by tissue motion from heartbeat, respiratory cycles, peristalsis, muscle/vascular tone and physiological functions that change tissue geometry. Because these movements impede temporal and spatial resolution, they must be properly addressed to harness the full potential of two‐photon intravital microscopy and enable accurate data analysis and interpretation. In addition, the sources and features of these motion artefacts are varied, sometimes unpredictable and unique to specific organs and multiple complex strategies have previously been devised to address them. This review will discuss these motion artefacts requirement and technical solutions for their correction and after intravital two‐photon microscopy. Introduction Two‐photon laser‐scanning microscopy was developed in 1990 and has become the method of choice to investigate biological processes in live animals, notably because its nonlinear nature confers it several appealing features over one‐photon confocal microscopy (Denk et al ., 1990 ). Indeed, for the excitation process to occur at the sample two photons must be coincident both in space and time on the fluorescent molecule on an extremely short time scale. Given the low probability of such an event occurring, the photon density arriving at the sample must be enormous. This is possible with mode‐locked lasers providing a train of pulses of short duration (∼100fs) with high peak power and repetition rate (∼80 MHz) of near‐infrared light. Notably, the latter provides better tissue penetration than its visible counterpart and makes it especially suitable for thick tissue imaging. Moreover, since two photons are required to produce signal the latter scales with the square of laser power, hence no fluorescence is generated outside the focal volume and unwanted background is minimized. In addition, the detection layout in two‐photon microscopy does not require a physical pinhole since no fluorescence is generated outside the focus in the first place (Sanderson et al ., 2014 ). This enables scattered photons to contribute to the useful signal since they originate from the focus by virtue of the nonlinear process described above. Therefore, two‐photon microscopy is highly appealing for intravital studies and possesses a remarkable ability to access moderately unperturbed environments with excellent temporal resolution facilitating the observation of dynamic biological events over long periods of time (Bullen, 2008 ; Niesner & Hauser, 2011 ; Pittet & Weissleder, 2011 ; Ellenbroek & van Rheenen, 2014 ). However, the extent to which the environment is perturbed is highly dependent on the location of the organ to be imaged. Indeed, gaining optical access deep into tissues that are themselves isolated creates substantial challenges in acquiring high‐quality images in a minimally invasive manner (Helmchen & Denk, 2005 ; Grewe & Helmchen, 2009 ; Megens et al ., 2011 ). For example, the liver, kidney, gut and spleen are relatively simple to image because they can be readily exteriorized and placed under an objective lens. By contrast, the brain, spinal cord, heart and lungs are much more difficult to image, because they require elaborate surgical procedures that likely impact their functions. These challenges can create a thin line between successfully imaging a physiological process in a remote organ and observing unwanted direct or indirect effects of an intrusive procedure on an organ. Therefore, to conduct intravital imaging experiments, we must find a balance between minimizing invasiveness during sample preparation and preserving our ability to reliably record physiological phenomena at high resolution. The latter is especially challenging in living animals, because – regardless of their location in the body – all organs are inherently subjected to disturbances caused by various physiological processes, such as the heartbeat, respiratory cycles, vascular tone or peristalsis. These disturbances inevitably cause motion artefacts that corrupt frames and must be eliminated or compensated to properly and accurately analyze data (Lucotte & Balaban, 2014 ; Vinegoni et al ., 2014 ). Researchers have recently begun devoting considerable effort into limiting the burden of motion artefacts. In this review, we will summarize technical and software solutions, particularly those that apply to studies in neuroscience, immunology and cell biology, and provide an overview of the organ‐specific tools available for stabilizing intravital optical imaging in small animals (mostly in mice, but also in rats and rabbits). The burden of motion artefacts For intravital imaging at the cellular level, we must take into account certain challenges, such as physical access to the organ, tissue penetration and motion artefacts not typically encountered in vitro . Motion artefacts are arguably the most cumbersome to tackle, mostly because each organ has a unique pattern, frequency or amplitude of movement and an unpredictable nature. In addition, because tissue movement is not bound to rigid, planar translations, the experimenter may not notice physical deformations, such as rotation and scale changes, until the end of the imaging session, which can further complicate their analysis. Moreover, motion artefacts are prone to providing erroneous impressions of movement within the organ that are not necessarily caused by a biological process, but, rather, are false positives that must be considered during data analysis and interpretation. This phenomenon is amplified when small motile structures, such as microglial processes, growing axons or dendritic spines, are observed at high magnification, when minor disturbances in the imaging (XY) or axial (Z) planes can mislead the representation of their genuine displacements. This effect is crucial for morphological, speed or any dynamic analysis of phenomena in moving environments. Several strategies address these issues and will be discussed in the following sections, including their potential use, combination and applicability to different organs and types of artefacts. We will also discuss compensation approaches, such as organ immobilization, active compensation, gated/triggered acquisition and image processing. The burden of motion artefacts For intravital imaging at the cellular level, we must take into account certain challenges, such as physical access to the organ, tissue penetration and motion artefacts not typically encountered in vitro . Motion artefacts are arguably the most cumbersome to tackle, mostly because each organ has a unique pattern, frequency or amplitude of movement and an unpredictable nature. In addition, because tissue movement is not bound to rigid, planar translations, the experimenter may not notice physical deformations, such as rotation and scale changes, until the end of the imaging session, which can further complicate their analysis. Moreover, motion artefacts are prone to providing erroneous impressions of movement within the organ that are not necessarily caused by a biological process, but, rather, are false positives that must be considered during data analysis and interpretation. This phenomenon is amplified when small motile structures, such as microglial processes, growing axons or dendritic spines, are observed at high magnification, when minor disturbances in the imaging (XY) or axial (Z) planes can mislead the representation of their genuine displacements. This effect is crucial for morphological, speed or any dynamic analysis of phenomena in moving environments. Several strategies address these issues and will be discussed in the following sections, including their potential use, combination and applicability to different organs and types of artefacts. We will also discuss compensation approaches, such as organ immobilization, active compensation, gated/triggered acquisition and image processing. Animal restraint and tissue immobilization In general, the extent of motion artefacts is heavily influenced by the immobilization strategy used to restrict displacement of the animal and organ of interest while avoiding perturbations to its normal functions. The ideal scenario involves minimally invasive surgery to access the sample and preserve its physiological environment. Because the location of an organ dictates the best approach to minimize shifting, the following three sections discuss many scenarios for physical restraint of abdominal organs (kidney, spleen, liver and intestine), organs located inside/close to the chest cavity (heart, lung and spinal cord) and the brain, which is uniquely located within the skull, away from extensive movements caused by breathing and heartbeat. Typical examples of animal restraint strategies are described in Table 1 . Table 1 Available restraining approaches for TP‐IVM Organ Biological events observed Restraining method Refs Gut ‐ Immune cell trafficking ‐ Myenteric neurogenesis ‐ Mucosal homeostasis and cellular interactions ‐ Externalized gut, pinned, sutured or sandwiched ‐ Gut stabilized with animal's weight ‐ Topical application of papaverine (Klinger et al ., 2012 ; Xu et al ., 2012 ; Goto et al ., 2013 ; Mizuno et al ., 2014 ; Motegi et al ., 2020 ) Abdominal organs ‐ Gene expression ‐ Cellular trafficking ‐ Exteriorized organ ‐ Custom microstage device ‐ Abdominal imaging window (Denk et al ., 1990 , Ritsma et al ., 2014 ) Spinal cord ‐ Resident and peripheral immune cell behaviour trafficking ‐ Extravasation and interactions with tissue ‐ Vascular permeability ‐ Axonal growth ‐ Spinal cord clamps ‐ Implanted custom Chamber ‐ Embedding with agar (Kerschensteiner et al ., 2005 ; Odoardi et al ., 2007 ; Johannssen & Helmchen, 2010 ; Kim et al ., 2010 ; Nikić et al ., 2011 ; Davalos & Akassoglou, 2012 ; Farrar et al ., 2012 ; Fenrich et al ., 2013 ; Coisne et al ., 2013 ; Evans et al ., 2014 ; Schaffran et al ., 2019 ), (Figley et al ., 2013 ) Lung ‐ Immune cell interactions ‐ Leukocyte trafficking and extravasation ‐ Cover slip over exposed lung ‐ Thoracic suction window ‐ Endotracheal tube (Kreisel et al ., 2010 ; Fiole et al ., 2014 ; Veres et al ., 2017 ), (Looney et al ., 2011 , Ueki et al ., 2018 ) Heart ‐ Leukocyte recruitment and trafficking dynamics ‐ Cell‐endothelium interactions ‐ Collagen‐rich structures ‐ Ring‐shaped stabilizer ‐ Compressive cover slip ‐ Gluing device (Li et al ., 2012 ; Lee et al ., 2012 ; Diaz et al ., 2014 ; Matsuura et al ., 2018 ) Brain ‐ Neuronal plasticity ‐ Ca 2+ imaging ‐ Neuronal activity ‐ Vascular structure and permeability ‐ Electrophysiological recordings ‐ Axonal growth ‐ Glial cell behaviour ‐ Stereotactic frame ‐ Head plate ‐ Heat post clamps ‐ Ear‐fixed bars ‐ Skull glued to support (Zhang et al ., 2012 ; Manglani & McGavern, 2018 ), (Dorand et al ., 2014 , Alieva et al ., 2014 ), (Drew et al ., 2010 ; Shih et al ., 2012 ), (Mizrahi et al ., 2004 , Gu et al ., 2014 ) John Wiley & Sons, Ltd. Gut and other abdominal organs Physical stabilization of abdominal organs is much easier when they are exteriorized (Marques et al ., 2015 ), even though this approach is considerably invasive. These organs can be positioned in almost any configuration, which enables whole‐organ imaging while avoiding the need to observe limited sections. However, this approach comes at the expense of larger degrees of freedom for movement and, thus, requires appropriate stabilization devices. Conventionally, organs are placed in a custom holder and pressure is applied with a cover slip or they can be sandwiched between two cover slips so that imaging is performed onto a planar surface (Klinger et al ., 2012 ; Xu et al ., 2012 ; Goto et al ., 2013 ; Mizuno et al ., 2014 ; Motegi et al ., 2020 ). Although this approach limits large‐scale XY or Z drift, it is unlikely to be suitable for smaller shifts. Recently, researchers combined this setup with single‐photon confocal microscopy for intravital imaging of the mouse liver, a relatively low‐cost alternative to two‐photon microscope systems (Soulet et al ., 2013 ). This approach can be used for intravital microscopy depending on the tissue. However, in the mouse liver, depth imaging is limited to approximately 100 μm from the organ surface (Marques et al ., 2015 ). Deeper imaging into living organs, requires limiting the extensive light scattering and absorption that are inherent with single photon excitation, but limited when two‐photon excitation is used (Helmchen & Denk, 2005 ; Cao et al ., 2012 ). Cao and colleagues developed a microstage device that uses ball joints to freely translate and rotate partially externalized organs in three dimensions (3D) (Megens et al ., 2010 ). When exteriorizing abdominal organs, one should consider several practical issues, including carefully regulating the chamber's temperature and providing physiological fluids that prevent organs from drying out. The use of an inverted microscope can provide additional benefits, as the animal's weight can be used to apply an even force on the organ to help dampen motion (Xu et al ., 2012 ; Mizuno et al ., 2014 ). Despite reducing overall movements, physical immobilization of exteriorized organs will likely not compensate for stable high‐resolution imaging (Soulet et al ., 2013 ). To increase stability, glue or sutures are often used, although care must be taken to prevent unwanted immune activation and perturbations to the environment (Goto et al ., 2013 ; Mizuno et al ., 2014 ). Even then, triggered acquisition or image processing might be required. For example, in the gut, peristalsis causes unpredictable, high‐amplitude movements that are not fully compensated by physical restraint alone (Chèvre et al ., 2014 ; Ritsma et al ., 2014 ). Goto et al . used papaverine, an opium alkaloid that treats spasms of the gastrointestinal tract, to suppress movement of the ileum (Goto et al ., 2013 ), which might affect intestinal functions in unpredictable ways. Alternatively, implanted chambers can adequately stabilize organs and allow imaging of the same region for extended periods of time. For example, Ritsma and colleagues described an implant that provides optical access to the kidney, liver, intestine, pancreas or spleen for up to 5 weeks (Davalos et al ., 2008 ). By securing organs to the inner wall of the chamber with glue and using a custom‐made imaging box, they reduced movements during data acquisition in a modestly invasive fashion. Moreover, placement of sterile gauze between organs and rigid structures, such as the diaphragm or rib cage, can help decrease breathing artefacts. However, since inflammatory reactions are difficult to avoid outside the chamber, appropriate controls are required to ensure that the implant does not change the organ's functions. Organs located inside/close to the chest cavity Spinal cord Organs located inside (heart, lungs) or close to (spinal cord) the chest cavity are the most challenging to tackle in terms of movement compensation. However, the spinal cord is notably unique because it is physically confined in the vertebral column and thus benefits from rigid support that helps reduce macroscopic movements. Because breathing generates most movements in the spinal cord, every effort should be made to promote smooth respiration, including careful choice of anesthetics and animal positioning (Kerschensteiner et al ., 2005 ; Odoardi et al ., 2007 ; Johannssen & Helmchen, 2010 ; Davalos & Akassoglou, 2012 ). As physical restraint of the spinal cord itself is not possible, vertebrae can be targeted to dampen its movement. For example, some studies described a stabilization method for short‐term imaging in which mice are suspended with spinal clamps to elevate the abdomen and allow them to breath freely (Johannssen & Helmchen, 2010 ; Nikić et al ., 2011 ; Coisne et al ., 2013 ; Schaffran et al ., 2019 ). Since residual motion artefacts caused by heartbeat and breathing are generally not eliminated altogether, they can be further reduced by embedding the spinal cord in agar (Kerschensteiner et al ., 2005 ). This strategy is widely used to image large regions of interest for short periods of time and can be applied to many segment of the spinal cord (Davalos et al ., 2012 ; Aubé et al ., 2014 ; Evans et al ., 2014 ; Haghayegh Jahromi et al ., 2017 ). Selecting different regions of the spinal cord for imaging also poses different challenges and advantages. For example, the thoracic spinal cord naturally curves outwards from the animal's body and thus requires less muscle displacement for exposure; but it is also closer to the heart and lungs, making it more sensitive to movement artefact. In the lumbar spinal region, the larger separation between the cord and surrounding vertebrae reduces the risk of puncturing or damaging the cord during laminectomy. However, it is covered by a thicker muscular layer due to its inward curvature, thus requiring a more invasive surgical procedure. Indeed, to obtain optical access to the specimen for imaging, one must expose the area to be imaged by removing or preferably displacing tissue and/or bone covering the spinal cord. This needs to be repeated for longitudinal imaging experiments of the same cells or spinal cord areas, for every imaging session (Johannssen & Helmchen, 2010 ). Several approaches have been devised and used to carry out repetitive imaging of the mouse spinal cord. Experiments that require only one or two imaging sessions can be performed by imaging through the interlaminar space, particularly in the thoracic spinal cord, provided of course that the limited field of view is compatible with the experimental design (Kim et al ., 2010 ; Farrar et al ., 2012 ). For longer‐term imaging requiring several reimaging sessions, the best strategy is to use a chronic implant to reduce the risks associated with repetitive surgical interventions. For example, Farrar et al . robustly stabilized animals with V‐shaped stainless‐steel clamps that sandwich vertebrae (Nikić et al ., 2011 ; Fenrich et al ., 2013 ). This design is limited by the relative complexity of construction, modest success rate and the need to administer antiinflammatory or immunosuppressant drugs to animals to prevent fibrosis over the laminectomy site. Alternatively, Fenrich et al . developed a window‐chamber design with staples and paperclips as supporting elements. This approach yields moderately higher success rates, longer imaging periods and is easier to use than the design by Farrar et al . (Figley et al ., 2013 ). Finally, Figley et al . designed a transparent 3D‐printed window built from nonmetallic materials for compatibility with nonfluorescent or laser‐scanning imaging modalities such as optical coherence tomography and photoacoustic imaging (Veres et al ., 2017a ). These window designs support stable intravital imaging of the mouse spinal cord; however, they can hardly eliminate minute drifts of the spinal cord and microscopic displacements caused by vascular tone, which can be corrected with image processing. Heart and lungs The heart and lungs present unique challenges for physical restraint. Because their complete immobilization is not possible, restraining methods aim to reduce movements and thereby facilitate the processing of image sequences offline. In addition, with gated acquisition, cardiac rhythm and breathing activity can be monitored to help synchronize image collection with periods of minimal organ movement. This approach will be discussed further in later sections. To access the heart and lungs, the chest cavity must be opened, which introduces challenges beyond surgical intervention. Inside the thorax, there is a vacuum created by the moving diaphragm to force air flow in and out of the lungs. Disrupting this vacuum to access the heart or lungs without respiratory support will eventually cause the lungs to collapse. Hence, to access those organs, one must consider the unique environment inside the thorax and maintain the vacuum with appropriate sealing and mechanical ventilation. A first approach was to use an endotracheal tube to stabilize the trachea during imaging (Fiole et al ., 2014 ). However, this method is not ideal to maintain the physiological integrity of the trachea wall and mucosa. New methods have therefore been developed to optimize the imaging window directly in the opening of the thoracotomy with a trachea holding device and a glued glass cover slip (Kreisel et al ., 2010 ; Looney et al ., 2011 ; Veres et al ., 2017b ). To collect image frames with minimal artefacts, this approach must be combined with data acquisition synchronized with respiration. As a more invasive, yet robust, alternative, a thoracic window can minimize movements and stabilize imaging opportunities. This type of window was introduced in 1939 for imaging of the cat lung and was revisited for the mouse in 2011 by Looney and colleagues based on Wagner's design for the dog (Ueki et al ., 2018 ). The suction‐based device applies gentle pressure (20–30 mm Hg) onto the lung so that it stays loosely adhered to the organ and allows it to expand and retract freely (Lee et al ., 2012 ). Residual movements in 3D are substantially reduced to a magnitude of 5–10 μm, which enables high‐resolution imaging of immune trafficking and dynamics. This thoracic window permits imaging during the entire respiratory cycle without interrupting the ventilation supplied to the animal, which is highly desirable, especially during prolonged imaging sessions. The caveats related to this approach include invasiveness, technical difficulty and the fact that mice are not spontaneously breathing. Similarly, the beating heart has high‐magnitude rhythmic movements that create considerable challenges for stable high‐resolution imaging. Most methods that stabilize the heart share the same principle: lightly gluing the organ to a rigid support and monitoring the animal's electrocardiogram and/or respiratory function and trigger data acquisition during periods with less movement. This period generally occurs during diastole, when the heart temporarily relaxes before the next systole. For example, Lee and colleagues developed a ring‐shaped stabilizer whose base is fixed to the myocardium, which significantly dampens large‐amplitude movements; however, cardiac contractions pressing against the device create residual motion artefacts (Aguirre et al ., 2014 ; Matsuura et al ., 2018 ). Jones et al . imaged single cardiomyocytes with a similar setup to that of Aguirre et al . These authors also administered the neuromuscular blocker pancuronium to mice and rabbits to reduce breathing‐induced movements (Li et al ., 2012 ; Diaz et al ., 2014 ; Jones et al ., 2018 ). However, care must be taken when using muscle‐paralyzing agents because they can induce considerable changes in heart rate or blood pressure (Li et al ., 2012 ). Li et al . successfully imaged neutrophil trafficking in beating transplanted hearts by securing the mouse between two metal plates and pressing on the organ with a cover slip (Drechsler et al ., 2010 ). The major arteries surrounding the heart are also of interest in many pathological conditions, especially atherosclerosis (Chèvre et al ., 2014 ; Ritsma et al ., 2014 ; Manglani & McGavern, 2018 ). Chèvre et al . performed multichannel, high‐speed imaging of the isolated carotid artery, using a bevel support and by pressing against the vessel with a cover slip, to collect high‐resolution images of atherosclerotic lesions in mice (Ritsma et al ., 2014 ). Exposed carotid arteries were also observed without a specific restraining apparatus; however, to obtain stable images, fast, gated acquisition had to be performed, in conjunction with extensive processing and at the expense of spatial resolution (Chèvre et al ., 2014 ; Manglani & McGavern, 2018 ). These techniques can be combined with heart rate‐ and breathing‐triggered acquisition, which will be discussed in more detail in a later section. Brain Similar to the spinal cord, the brain is tightly contained by a bone structure that naturally provides physical restraint. Its remoteness from the heart and chest cavity also helps create an environment moderately affected by motion artefacts. Physical restraint of the brain itself is futile, because it must be allowed to move freely in the subarachnoid space; therefore, techniques that stabilize the skull have the sole purpose of stably supporting the organ for imaging. Two approaches have been developed towards this end: open‐skull and thinned‐skull windows. As their names imply, the first involves a complete craniotomy, whereas the second does not. Since the brain was the first organ imaged in vivo by two‐photon microscopy over 20 years ago, both methods have been used extensively by researchers in the neuroscience community and implementation has been extensively described (Xu et al ., 2007 ; Zhang et al ., 2012 ; Alieva et al ., 2014 ; Dorand et al ., 2014 ). Each method has its own merits and drawbacks, and selecting one over the other depends on experimental design and goals; for example, thinned‐skull windows allow imaging of the intact mouse cortex over several hours, but can cause pial injury if they are rethinned more than two to three times. Conversely, open‐skull windows allow for numerous reimaging sessions over days to weeks or months, but their original implantation causes substantial pial inflammation that requires at least 10 days to subside and allow for 'physiological' imaging conditions (Shih et al ., 2012 ). From an image quality perspective, the skull thinning procedure can cause irregularities that induce spherical aberrations and are thus more challenging to perform successfully than open‐skull windows. Recent attempts have been made to combine the best qualities of both methods by creating a thinned‐skull window that is polished and reinforced by a glass cover slip that is glued over the thinned area to prevent bone regrowth (Mizrahi et al ., 2004 ; Drew et al ., 2010 ; Dorand et al ., 2014 ). Of interest is that brain imaging methods have also been developed for in vivo study of deeper brain structures such as the hippocampus in mice, as this region is of extensive interest for both physiological and pathological brain functions (Gu et al ., 2014 ). This method is also quite challenging, since it requires surgical removal of the cortical tissue above these deeper brain regions, and even longer waiting periods than those for window implantation for inflammation resolution (Park et al ., 2019 ). All these methods stabilize the organ, using a common fixation approach through either a stereotactic frame or a head‐restraining apparatus. Moreover, translucent substances, such as silicone elastomer or low‐melting point agarose, can fill the space between the cortical surface and cover slip to help maintain their adhesion over extended periods of time (Holtmaat et al ., 2009 ; Yildirim et al ., 2019 ). These substances also minimize the refractive‐indices mismatch between the cover glass and brain tissue, thereby reducing optical aberrations. Generally, brain‐related motion artefacts do not create a major issue in anesthetized mice because of their low amplitude and frequency, ranging from about 2 to 5 cycles/s. However, insufficient anesthesia induces accelerated breathing and heartbeat that translate into visible drift and artefacts. We stress that the level of anesthesia must be carefully monitored during the entire experiment (i.e. surgery and imaging) to avoid pain and distress to animals. Intravital imaging of the brain becomes a bigger challenge in awake behaving animals, in which tissue displacements are more pronounced and difficult to tackle because their amplitude is likely larger than the structures under observation. However, observation of the brain of nonanesthetized animals with optical microscopy better associates behaviours or external stimuli to communication patterns among neuronal populations. For this purpose, calcium (Ca 2+ ) transients and action potentials are generally recorded with genetically encoded probes, adeno‐associated virus vectors or by bulk loading fluorescent indicators into the brain (Garaschuk et al ., 2006 ; Andermann et al ., 2011 ; St‐Pierre et al ., 2014 ). The volume spanned by neuronal networks in specific brain compartments is substantial. Thus, the ideal experimental paradigm enables fast, stable, sensitive and high‐resolution imaging of neuronal structures over large areas and depths. However, for practical and technical considerations these features are hard to achieve simultaneously in vivo , so precedence must be given to some over others. The ideal paradigm primarily discriminates the approaches developed to capture neuronal activity in awake animals, which can be loosely sorted into tabletop and portable systems. Tabletop systems usually involve head‐fixed mice under a conventional microscope and a laser source that allows two‐photon excitation and its associated benefits. In these mice, access to the brain and its restraint are similar to anesthetized mice; however, awake animals are exposed to different environments suitable for studying the motor, visual, auditory or barrel cortices, such as during locomotion or sensory stimulation (Sato et al ., 2007 ; Kerlin et al ., 2010 ; Scott et al ., 2014 ; Nadella et al ., 2016 ). Recent technical developments use new microscope technology to decrease motion artefact by increasing the speed of acquisition. For example, random‐access scanning microscopy enable high‐speed volumetric imaging of brain with reduces intraframe motion artefacts (Botcherby et al ., 2006 ). Another available technic is the use of a Bessel beam to reduce the acquisition time in the optical Z‐axis by increasing the length of the point spread function of the focused excitation beam to generate an extended depth of field image (Song et al ., 2017 ). Recently, Song et al . created a method to address loss of depth information limitation of Bessel beam by splitting the beam. Referred as volumetric Two‐photon Imaging of Neurons using Stereoscopy, the two excitation beams create a stereoscopic 'V'‐shaped point spread function configuration that preserves depth information. The combination of Bessel beam and volumetric Two‐photon Imaging of Neurons using Stereoscopy allows a high‐speed image acquisition and a lower exposure time without loss of 3D information (Birkner et al ., 2017 ). Another method using ultra‐short laser pulses can increase rapidly the fluorescent indicator energy with lower power laser. This method presents another solution to increase the acquisition speed, to decrease the phototoxicity and to increase the depth penetration (Sato et al ., 2017 ). Recently, researchers performed dual‐axis Ca2+ imaging in distinct areas of the visual cortex using gradient refractive index microendoscopes and optogenetic perturbation of targeted neuronal networks. This was performed alongside Ca 2+ imaging that probed the activity in each area during different activity states (Lecoq et al ., 2014 ; Packer et al ., 2015 ). Interestingly, along with image registration, Packer and colleagues used microscope parts that moved the optical path relative to the sample to reduce the impact of motion artefacts (Gulati et al ., 2017 ). This paradigm is also common in portable‐microscope systems that fix fibre‐based and integrated microscopes to the animal's head, thereby allowing for behavioural testing of the mice during imaging, since they are not head‐restrained and can move freely (Ghosh et al ., 2011 ; Helmchen et al ., 2013 ; Ziv et al ., 2013 ; Chen et al ., 2018 ). Moreover, integrated microscopes with light‐emitting diode (LED) illumination and a large depth of field, enable experimenters to perform video‐rate image acquisition of extensive fields of view with fewer effects from motion artefacts (Ziv et al ., 2013 ). When equipped with a microendoscope, miniaturized microscopes support relatively stable Ca 2+ imaging in deep brain areas, such as the hippocampus, while animals are awake and behaving (Kerr & Nimmerjahn, 2012 ; Helmchen et al ., 2013 ; Berdyyeva et al ., 2014 ). Without restraining devices or anesthetic, miniature microscope avoids movement of the head and cause less intraframe motion artefacts (Sawinski et al ., 2009 ; Zong et al ., 2017 ). The main drawbacks associated with these systems are their limited spatial resolution and depth of field, as well as a limited single‐photon/colour excitation and the impossibility of moving the lens during imaging. These drawbacks will undoubtedly be addressed in the near future and should render integrated microscopes as widespread, complementary tools to two‐photon excitation systems mainly for the study of the brain and the spinal cord (Paukert & Bergles, 2012 ; Sekiguchi et al ., 2016 ; Nelson et al ., 2019 ). Gut and other abdominal organs Physical stabilization of abdominal organs is much easier when they are exteriorized (Marques et al ., 2015 ), even though this approach is considerably invasive. These organs can be positioned in almost any configuration, which enables whole‐organ imaging while avoiding the need to observe limited sections. However, this approach comes at the expense of larger degrees of freedom for movement and, thus, requires appropriate stabilization devices. Conventionally, organs are placed in a custom holder and pressure is applied with a cover slip or they can be sandwiched between two cover slips so that imaging is performed onto a planar surface (Klinger et al ., 2012 ; Xu et al ., 2012 ; Goto et al ., 2013 ; Mizuno et al ., 2014 ; Motegi et al ., 2020 ). Although this approach limits large‐scale XY or Z drift, it is unlikely to be suitable for smaller shifts. Recently, researchers combined this setup with single‐photon confocal microscopy for intravital imaging of the mouse liver, a relatively low‐cost alternative to two‐photon microscope systems (Soulet et al ., 2013 ). This approach can be used for intravital microscopy depending on the tissue. However, in the mouse liver, depth imaging is limited to approximately 100 μm from the organ surface (Marques et al ., 2015 ). Deeper imaging into living organs, requires limiting the extensive light scattering and absorption that are inherent with single photon excitation, but limited when two‐photon excitation is used (Helmchen & Denk, 2005 ; Cao et al ., 2012 ). Cao and colleagues developed a microstage device that uses ball joints to freely translate and rotate partially externalized organs in three dimensions (3D) (Megens et al ., 2010 ). When exteriorizing abdominal organs, one should consider several practical issues, including carefully regulating the chamber's temperature and providing physiological fluids that prevent organs from drying out. The use of an inverted microscope can provide additional benefits, as the animal's weight can be used to apply an even force on the organ to help dampen motion (Xu et al ., 2012 ; Mizuno et al ., 2014 ). Despite reducing overall movements, physical immobilization of exteriorized organs will likely not compensate for stable high‐resolution imaging (Soulet et al ., 2013 ). To increase stability, glue or sutures are often used, although care must be taken to prevent unwanted immune activation and perturbations to the environment (Goto et al ., 2013 ; Mizuno et al ., 2014 ). Even then, triggered acquisition or image processing might be required. For example, in the gut, peristalsis causes unpredictable, high‐amplitude movements that are not fully compensated by physical restraint alone (Chèvre et al ., 2014 ; Ritsma et al ., 2014 ). Goto et al . used papaverine, an opium alkaloid that treats spasms of the gastrointestinal tract, to suppress movement of the ileum (Goto et al ., 2013 ), which might affect intestinal functions in unpredictable ways. Alternatively, implanted chambers can adequately stabilize organs and allow imaging of the same region for extended periods of time. For example, Ritsma and colleagues described an implant that provides optical access to the kidney, liver, intestine, pancreas or spleen for up to 5 weeks (Davalos et al ., 2008 ). By securing organs to the inner wall of the chamber with glue and using a custom‐made imaging box, they reduced movements during data acquisition in a modestly invasive fashion. Moreover, placement of sterile gauze between organs and rigid structures, such as the diaphragm or rib cage, can help decrease breathing artefacts. However, since inflammatory reactions are difficult to avoid outside the chamber, appropriate controls are required to ensure that the implant does not change the organ's functions. Organs located inside/close to the chest cavity Spinal cord Organs located inside (heart, lungs) or close to (spinal cord) the chest cavity are the most challenging to tackle in terms of movement compensation. However, the spinal cord is notably unique because it is physically confined in the vertebral column and thus benefits from rigid support that helps reduce macroscopic movements. Because breathing generates most movements in the spinal cord, every effort should be made to promote smooth respiration, including careful choice of anesthetics and animal positioning (Kerschensteiner et al ., 2005 ; Odoardi et al ., 2007 ; Johannssen & Helmchen, 2010 ; Davalos & Akassoglou, 2012 ). As physical restraint of the spinal cord itself is not possible, vertebrae can be targeted to dampen its movement. For example, some studies described a stabilization method for short‐term imaging in which mice are suspended with spinal clamps to elevate the abdomen and allow them to breath freely (Johannssen & Helmchen, 2010 ; Nikić et al ., 2011 ; Coisne et al ., 2013 ; Schaffran et al ., 2019 ). Since residual motion artefacts caused by heartbeat and breathing are generally not eliminated altogether, they can be further reduced by embedding the spinal cord in agar (Kerschensteiner et al ., 2005 ). This strategy is widely used to image large regions of interest for short periods of time and can be applied to many segment of the spinal cord (Davalos et al ., 2012 ; Aubé et al ., 2014 ; Evans et al ., 2014 ; Haghayegh Jahromi et al ., 2017 ). Selecting different regions of the spinal cord for imaging also poses different challenges and advantages. For example, the thoracic spinal cord naturally curves outwards from the animal's body and thus requires less muscle displacement for exposure; but it is also closer to the heart and lungs, making it more sensitive to movement artefact. In the lumbar spinal region, the larger separation between the cord and surrounding vertebrae reduces the risk of puncturing or damaging the cord during laminectomy. However, it is covered by a thicker muscular layer due to its inward curvature, thus requiring a more invasive surgical procedure. Indeed, to obtain optical access to the specimen for imaging, one must expose the area to be imaged by removing or preferably displacing tissue and/or bone covering the spinal cord. This needs to be repeated for longitudinal imaging experiments of the same cells or spinal cord areas, for every imaging session (Johannssen & Helmchen, 2010 ). Several approaches have been devised and used to carry out repetitive imaging of the mouse spinal cord. Experiments that require only one or two imaging sessions can be performed by imaging through the interlaminar space, particularly in the thoracic spinal cord, provided of course that the limited field of view is compatible with the experimental design (Kim et al ., 2010 ; Farrar et al ., 2012 ). For longer‐term imaging requiring several reimaging sessions, the best strategy is to use a chronic implant to reduce the risks associated with repetitive surgical interventions. For example, Farrar et al . robustly stabilized animals with V‐shaped stainless‐steel clamps that sandwich vertebrae (Nikić et al ., 2011 ; Fenrich et al ., 2013 ). This design is limited by the relative complexity of construction, modest success rate and the need to administer antiinflammatory or immunosuppressant drugs to animals to prevent fibrosis over the laminectomy site. Alternatively, Fenrich et al . developed a window‐chamber design with staples and paperclips as supporting elements. This approach yields moderately higher success rates, longer imaging periods and is easier to use than the design by Farrar et al . (Figley et al ., 2013 ). Finally, Figley et al . designed a transparent 3D‐printed window built from nonmetallic materials for compatibility with nonfluorescent or laser‐scanning imaging modalities such as optical coherence tomography and photoacoustic imaging (Veres et al ., 2017a ). These window designs support stable intravital imaging of the mouse spinal cord; however, they can hardly eliminate minute drifts of the spinal cord and microscopic displacements caused by vascular tone, which can be corrected with image processing. Heart and lungs The heart and lungs present unique challenges for physical restraint. Because their complete immobilization is not possible, restraining methods aim to reduce movements and thereby facilitate the processing of image sequences offline. In addition, with gated acquisition, cardiac rhythm and breathing activity can be monitored to help synchronize image collection with periods of minimal organ movement. This approach will be discussed further in later sections. To access the heart and lungs, the chest cavity must be opened, which introduces challenges beyond surgical intervention. Inside the thorax, there is a vacuum created by the moving diaphragm to force air flow in and out of the lungs. Disrupting this vacuum to access the heart or lungs without respiratory support will eventually cause the lungs to collapse. Hence, to access those organs, one must consider the unique environment inside the thorax and maintain the vacuum with appropriate sealing and mechanical ventilation. A first approach was to use an endotracheal tube to stabilize the trachea during imaging (Fiole et al ., 2014 ). However, this method is not ideal to maintain the physiological integrity of the trachea wall and mucosa. New methods have therefore been developed to optimize the imaging window directly in the opening of the thoracotomy with a trachea holding device and a glued glass cover slip (Kreisel et al ., 2010 ; Looney et al ., 2011 ; Veres et al ., 2017b ). To collect image frames with minimal artefacts, this approach must be combined with data acquisition synchronized with respiration. As a more invasive, yet robust, alternative, a thoracic window can minimize movements and stabilize imaging opportunities. This type of window was introduced in 1939 for imaging of the cat lung and was revisited for the mouse in 2011 by Looney and colleagues based on Wagner's design for the dog (Ueki et al ., 2018 ). The suction‐based device applies gentle pressure (20–30 mm Hg) onto the lung so that it stays loosely adhered to the organ and allows it to expand and retract freely (Lee et al ., 2012 ). Residual movements in 3D are substantially reduced to a magnitude of 5–10 μm, which enables high‐resolution imaging of immune trafficking and dynamics. This thoracic window permits imaging during the entire respiratory cycle without interrupting the ventilation supplied to the animal, which is highly desirable, especially during prolonged imaging sessions. The caveats related to this approach include invasiveness, technical difficulty and the fact that mice are not spontaneously breathing. Similarly, the beating heart has high‐magnitude rhythmic movements that create considerable challenges for stable high‐resolution imaging. Most methods that stabilize the heart share the same principle: lightly gluing the organ to a rigid support and monitoring the animal's electrocardiogram and/or respiratory function and trigger data acquisition during periods with less movement. This period generally occurs during diastole, when the heart temporarily relaxes before the next systole. For example, Lee and colleagues developed a ring‐shaped stabilizer whose base is fixed to the myocardium, which significantly dampens large‐amplitude movements; however, cardiac contractions pressing against the device create residual motion artefacts (Aguirre et al ., 2014 ; Matsuura et al ., 2018 ). Jones et al . imaged single cardiomyocytes with a similar setup to that of Aguirre et al . These authors also administered the neuromuscular blocker pancuronium to mice and rabbits to reduce breathing‐induced movements (Li et al ., 2012 ; Diaz et al ., 2014 ; Jones et al ., 2018 ). However, care must be taken when using muscle‐paralyzing agents because they can induce considerable changes in heart rate or blood pressure (Li et al ., 2012 ). Li et al . successfully imaged neutrophil trafficking in beating transplanted hearts by securing the mouse between two metal plates and pressing on the organ with a cover slip (Drechsler et al ., 2010 ). The major arteries surrounding the heart are also of interest in many pathological conditions, especially atherosclerosis (Chèvre et al ., 2014 ; Ritsma et al ., 2014 ; Manglani & McGavern, 2018 ). Chèvre et al . performed multichannel, high‐speed imaging of the isolated carotid artery, using a bevel support and by pressing against the vessel with a cover slip, to collect high‐resolution images of atherosclerotic lesions in mice (Ritsma et al ., 2014 ). Exposed carotid arteries were also observed without a specific restraining apparatus; however, to obtain stable images, fast, gated acquisition had to be performed, in conjunction with extensive processing and at the expense of spatial resolution (Chèvre et al ., 2014 ; Manglani & McGavern, 2018 ). These techniques can be combined with heart rate‐ and breathing‐triggered acquisition, which will be discussed in more detail in a later section. Brain Similar to the spinal cord, the brain is tightly contained by a bone structure that naturally provides physical restraint. Its remoteness from the heart and chest cavity also helps create an environment moderately affected by motion artefacts. Physical restraint of the brain itself is futile, because it must be allowed to move freely in the subarachnoid space; therefore, techniques that stabilize the skull have the sole purpose of stably supporting the organ for imaging. Two approaches have been developed towards this end: open‐skull and thinned‐skull windows. As their names imply, the first involves a complete craniotomy, whereas the second does not. Since the brain was the first organ imaged in vivo by two‐photon microscopy over 20 years ago, both methods have been used extensively by researchers in the neuroscience community and implementation has been extensively described (Xu et al ., 2007 ; Zhang et al ., 2012 ; Alieva et al ., 2014 ; Dorand et al ., 2014 ). Each method has its own merits and drawbacks, and selecting one over the other depends on experimental design and goals; for example, thinned‐skull windows allow imaging of the intact mouse cortex over several hours, but can cause pial injury if they are rethinned more than two to three times. Conversely, open‐skull windows allow for numerous reimaging sessions over days to weeks or months, but their original implantation causes substantial pial inflammation that requires at least 10 days to subside and allow for 'physiological' imaging conditions (Shih et al ., 2012 ). From an image quality perspective, the skull thinning procedure can cause irregularities that induce spherical aberrations and are thus more challenging to perform successfully than open‐skull windows. Recent attempts have been made to combine the best qualities of both methods by creating a thinned‐skull window that is polished and reinforced by a glass cover slip that is glued over the thinned area to prevent bone regrowth (Mizrahi et al ., 2004 ; Drew et al ., 2010 ; Dorand et al ., 2014 ). Of interest is that brain imaging methods have also been developed for in vivo study of deeper brain structures such as the hippocampus in mice, as this region is of extensive interest for both physiological and pathological brain functions (Gu et al ., 2014 ). This method is also quite challenging, since it requires surgical removal of the cortical tissue above these deeper brain regions, and even longer waiting periods than those for window implantation for inflammation resolution (Park et al ., 2019 ). All these methods stabilize the organ, using a common fixation approach through either a stereotactic frame or a head‐restraining apparatus. Moreover, translucent substances, such as silicone elastomer or low‐melting point agarose, can fill the space between the cortical surface and cover slip to help maintain their adhesion over extended periods of time (Holtmaat et al ., 2009 ; Yildirim et al ., 2019 ). These substances also minimize the refractive‐indices mismatch between the cover glass and brain tissue, thereby reducing optical aberrations. Generally, brain‐related motion artefacts do not create a major issue in anesthetized mice because of their low amplitude and frequency, ranging from about 2 to 5 cycles/s. However, insufficient anesthesia induces accelerated breathing and heartbeat that translate into visible drift and artefacts. We stress that the level of anesthesia must be carefully monitored during the entire experiment (i.e. surgery and imaging) to avoid pain and distress to animals. Intravital imaging of the brain becomes a bigger challenge in awake behaving animals, in which tissue displacements are more pronounced and difficult to tackle because their amplitude is likely larger than the structures under observation. However, observation of the brain of nonanesthetized animals with optical microscopy better associates behaviours or external stimuli to communication patterns among neuronal populations. For this purpose, calcium (Ca 2+ ) transients and action potentials are generally recorded with genetically encoded probes, adeno‐associated virus vectors or by bulk loading fluorescent indicators into the brain (Garaschuk et al ., 2006 ; Andermann et al ., 2011 ; St‐Pierre et al ., 2014 ). The volume spanned by neuronal networks in specific brain compartments is substantial. Thus, the ideal experimental paradigm enables fast, stable, sensitive and high‐resolution imaging of neuronal structures over large areas and depths. However, for practical and technical considerations these features are hard to achieve simultaneously in vivo , so precedence must be given to some over others. The ideal paradigm primarily discriminates the approaches developed to capture neuronal activity in awake animals, which can be loosely sorted into tabletop and portable systems. Tabletop systems usually involve head‐fixed mice under a conventional microscope and a laser source that allows two‐photon excitation and its associated benefits. In these mice, access to the brain and its restraint are similar to anesthetized mice; however, awake animals are exposed to different environments suitable for studying the motor, visual, auditory or barrel cortices, such as during locomotion or sensory stimulation (Sato et al ., 2007 ; Kerlin et al ., 2010 ; Scott et al ., 2014 ; Nadella et al ., 2016 ). Recent technical developments use new microscope technology to decrease motion artefact by increasing the speed of acquisition. For example, random‐access scanning microscopy enable high‐speed volumetric imaging of brain with reduces intraframe motion artefacts (Botcherby et al ., 2006 ). Another available technic is the use of a Bessel beam to reduce the acquisition time in the optical Z‐axis by increasing the length of the point spread function of the focused excitation beam to generate an extended depth of field image (Song et al ., 2017 ). Recently, Song et al . created a method to address loss of depth information limitation of Bessel beam by splitting the beam. Referred as volumetric Two‐photon Imaging of Neurons using Stereoscopy, the two excitation beams create a stereoscopic 'V'‐shaped point spread function configuration that preserves depth information. The combination of Bessel beam and volumetric Two‐photon Imaging of Neurons using Stereoscopy allows a high‐speed image acquisition and a lower exposure time without loss of 3D information (Birkner et al ., 2017 ). Another method using ultra‐short laser pulses can increase rapidly the fluorescent indicator energy with lower power laser. This method presents another solution to increase the acquisition speed, to decrease the phototoxicity and to increase the depth penetration (Sato et al ., 2017 ). Recently, researchers performed dual‐axis Ca2+ imaging in distinct areas of the visual cortex using gradient refractive index microendoscopes and optogenetic perturbation of targeted neuronal networks. This was performed alongside Ca 2+ imaging that probed the activity in each area during different activity states (Lecoq et al ., 2014 ; Packer et al ., 2015 ). Interestingly, along with image registration, Packer and colleagues used microscope parts that moved the optical path relative to the sample to reduce the impact of motion artefacts (Gulati et al ., 2017 ). This paradigm is also common in portable‐microscope systems that fix fibre‐based and integrated microscopes to the animal's head, thereby allowing for behavioural testing of the mice during imaging, since they are not head‐restrained and can move freely (Ghosh et al ., 2011 ; Helmchen et al ., 2013 ; Ziv et al ., 2013 ; Chen et al ., 2018 ). Moreover, integrated microscopes with light‐emitting diode (LED) illumination and a large depth of field, enable experimenters to perform video‐rate image acquisition of extensive fields of view with fewer effects from motion artefacts (Ziv et al ., 2013 ). When equipped with a microendoscope, miniaturized microscopes support relatively stable Ca 2+ imaging in deep brain areas, such as the hippocampus, while animals are awake and behaving (Kerr & Nimmerjahn, 2012 ; Helmchen et al ., 2013 ; Berdyyeva et al ., 2014 ). Without restraining devices or anesthetic, miniature microscope avoids movement of the head and cause less intraframe motion artefacts (Sawinski et al ., 2009 ; Zong et al ., 2017 ). The main drawbacks associated with these systems are their limited spatial resolution and depth of field, as well as a limited single‐photon/colour excitation and the impossibility of moving the lens during imaging. These drawbacks will undoubtedly be addressed in the near future and should render integrated microscopes as widespread, complementary tools to two‐photon excitation systems mainly for the study of the brain and the spinal cord (Paukert & Bergles, 2012 ; Sekiguchi et al ., 2016 ; Nelson et al ., 2019 ). Spinal cord Organs located inside (heart, lungs) or close to (spinal cord) the chest cavity are the most challenging to tackle in terms of movement compensation. However, the spinal cord is notably unique because it is physically confined in the vertebral column and thus benefits from rigid support that helps reduce macroscopic movements. Because breathing generates most movements in the spinal cord, every effort should be made to promote smooth respiration, including careful choice of anesthetics and animal positioning (Kerschensteiner et al ., 2005 ; Odoardi et al ., 2007 ; Johannssen & Helmchen, 2010 ; Davalos & Akassoglou, 2012 ). As physical restraint of the spinal cord itself is not possible, vertebrae can be targeted to dampen its movement. For example, some studies described a stabilization method for short‐term imaging in which mice are suspended with spinal clamps to elevate the abdomen and allow them to breath freely (Johannssen & Helmchen, 2010 ; Nikić et al ., 2011 ; Coisne et al ., 2013 ; Schaffran et al ., 2019 ). Since residual motion artefacts caused by heartbeat and breathing are generally not eliminated altogether, they can be further reduced by embedding the spinal cord in agar (Kerschensteiner et al ., 2005 ). This strategy is widely used to image large regions of interest for short periods of time and can be applied to many segment of the spinal cord (Davalos et al ., 2012 ; Aubé et al ., 2014 ; Evans et al ., 2014 ; Haghayegh Jahromi et al ., 2017 ). Selecting different regions of the spinal cord for imaging also poses different challenges and advantages. For example, the thoracic spinal cord naturally curves outwards from the animal's body and thus requires less muscle displacement for exposure; but it is also closer to the heart and lungs, making it more sensitive to movement artefact. In the lumbar spinal region, the larger separation between the cord and surrounding vertebrae reduces the risk of puncturing or damaging the cord during laminectomy. However, it is covered by a thicker muscular layer due to its inward curvature, thus requiring a more invasive surgical procedure. Indeed, to obtain optical access to the specimen for imaging, one must expose the area to be imaged by removing or preferably displacing tissue and/or bone covering the spinal cord. This needs to be repeated for longitudinal imaging experiments of the same cells or spinal cord areas, for every imaging session (Johannssen & Helmchen, 2010 ). Several approaches have been devised and used to carry out repetitive imaging of the mouse spinal cord. Experiments that require only one or two imaging sessions can be performed by imaging through the interlaminar space, particularly in the thoracic spinal cord, provided of course that the limited field of view is compatible with the experimental design (Kim et al ., 2010 ; Farrar et al ., 2012 ). For longer‐term imaging requiring several reimaging sessions, the best strategy is to use a chronic implant to reduce the risks associated with repetitive surgical interventions. For example, Farrar et al . robustly stabilized animals with V‐shaped stainless‐steel clamps that sandwich vertebrae (Nikić et al ., 2011 ; Fenrich et al ., 2013 ). This design is limited by the relative complexity of construction, modest success rate and the need to administer antiinflammatory or immunosuppressant drugs to animals to prevent fibrosis over the laminectomy site. Alternatively, Fenrich et al . developed a window‐chamber design with staples and paperclips as supporting elements. This approach yields moderately higher success rates, longer imaging periods and is easier to use than the design by Farrar et al . (Figley et al ., 2013 ). Finally, Figley et al . designed a transparent 3D‐printed window built from nonmetallic materials for compatibility with nonfluorescent or laser‐scanning imaging modalities such as optical coherence tomography and photoacoustic imaging (Veres et al ., 2017a ). These window designs support stable intravital imaging of the mouse spinal cord; however, they can hardly eliminate minute drifts of the spinal cord and microscopic displacements caused by vascular tone, which can be corrected with image processing. Heart and lungs The heart and lungs present unique challenges for physical restraint. Because their complete immobilization is not possible, restraining methods aim to reduce movements and thereby facilitate the processing of image sequences offline. In addition, with gated acquisition, cardiac rhythm and breathing activity can be monitored to help synchronize image collection with periods of minimal organ movement. This approach will be discussed further in later sections. To access the heart and lungs, the chest cavity must be opened, which introduces challenges beyond surgical intervention. Inside the thorax, there is a vacuum created by the moving diaphragm to force air flow in and out of the lungs. Disrupting this vacuum to access the heart or lungs without respiratory support will eventually cause the lungs to collapse. Hence, to access those organs, one must consider the unique environment inside the thorax and maintain the vacuum with appropriate sealing and mechanical ventilation. A first approach was to use an endotracheal tube to stabilize the trachea during imaging (Fiole et al ., 2014 ). However, this method is not ideal to maintain the physiological integrity of the trachea wall and mucosa. New methods have therefore been developed to optimize the imaging window directly in the opening of the thoracotomy with a trachea holding device and a glued glass cover slip (Kreisel et al ., 2010 ; Looney et al ., 2011 ; Veres et al ., 2017b ). To collect image frames with minimal artefacts, this approach must be combined with data acquisition synchronized with respiration. As a more invasive, yet robust, alternative, a thoracic window can minimize movements and stabilize imaging opportunities. This type of window was introduced in 1939 for imaging of the cat lung and was revisited for the mouse in 2011 by Looney and colleagues based on Wagner's design for the dog (Ueki et al ., 2018 ). The suction‐based device applies gentle pressure (20–30 mm Hg) onto the lung so that it stays loosely adhered to the organ and allows it to expand and retract freely (Lee et al ., 2012 ). Residual movements in 3D are substantially reduced to a magnitude of 5–10 μm, which enables high‐resolution imaging of immune trafficking and dynamics. This thoracic window permits imaging during the entire respiratory cycle without interrupting the ventilation supplied to the animal, which is highly desirable, especially during prolonged imaging sessions. The caveats related to this approach include invasiveness, technical difficulty and the fact that mice are not spontaneously breathing. Similarly, the beating heart has high‐magnitude rhythmic movements that create considerable challenges for stable high‐resolution imaging. Most methods that stabilize the heart share the same principle: lightly gluing the organ to a rigid support and monitoring the animal's electrocardiogram and/or respiratory function and trigger data acquisition during periods with less movement. This period generally occurs during diastole, when the heart temporarily relaxes before the next systole. For example, Lee and colleagues developed a ring‐shaped stabilizer whose base is fixed to the myocardium, which significantly dampens large‐amplitude movements; however, cardiac contractions pressing against the device create residual motion artefacts (Aguirre et al ., 2014 ; Matsuura et al ., 2018 ). Jones et al . imaged single cardiomyocytes with a similar setup to that of Aguirre et al . These authors also administered the neuromuscular blocker pancuronium to mice and rabbits to reduce breathing‐induced movements (Li et al ., 2012 ; Diaz et al ., 2014 ; Jones et al ., 2018 ). However, care must be taken when using muscle‐paralyzing agents because they can induce considerable changes in heart rate or blood pressure (Li et al ., 2012 ). Li et al . successfully imaged neutrophil trafficking in beating transplanted hearts by securing the mouse between two metal plates and pressing on the organ with a cover slip (Drechsler et al ., 2010 ). The major arteries surrounding the heart are also of interest in many pathological conditions, especially atherosclerosis (Chèvre et al ., 2014 ; Ritsma et al ., 2014 ; Manglani & McGavern, 2018 ). Chèvre et al . performed multichannel, high‐speed imaging of the isolated carotid artery, using a bevel support and by pressing against the vessel with a cover slip, to collect high‐resolution images of atherosclerotic lesions in mice (Ritsma et al ., 2014 ). Exposed carotid arteries were also observed without a specific restraining apparatus; however, to obtain stable images, fast, gated acquisition had to be performed, in conjunction with extensive processing and at the expense of spatial resolution (Chèvre et al ., 2014 ; Manglani & McGavern, 2018 ). These techniques can be combined with heart rate‐ and breathing‐triggered acquisition, which will be discussed in more detail in a later section. Brain Similar to the spinal cord, the brain is tightly contained by a bone structure that naturally provides physical restraint. Its remoteness from the heart and chest cavity also helps create an environment moderately affected by motion artefacts. Physical restraint of the brain itself is futile, because it must be allowed to move freely in the subarachnoid space; therefore, techniques that stabilize the skull have the sole purpose of stably supporting the organ for imaging. Two approaches have been developed towards this end: open‐skull and thinned‐skull windows. As their names imply, the first involves a complete craniotomy, whereas the second does not. Since the brain was the first organ imaged in vivo by two‐photon microscopy over 20 years ago, both methods have been used extensively by researchers in the neuroscience community and implementation has been extensively described (Xu et al ., 2007 ; Zhang et al ., 2012 ; Alieva et al ., 2014 ; Dorand et al ., 2014 ). Each method has its own merits and drawbacks, and selecting one over the other depends on experimental design and goals; for example, thinned‐skull windows allow imaging of the intact mouse cortex over several hours, but can cause pial injury if they are rethinned more than two to three times. Conversely, open‐skull windows allow for numerous reimaging sessions over days to weeks or months, but their original implantation causes substantial pial inflammation that requires at least 10 days to subside and allow for 'physiological' imaging conditions (Shih et al ., 2012 ). From an image quality perspective, the skull thinning procedure can cause irregularities that induce spherical aberrations and are thus more challenging to perform successfully than open‐skull windows. Recent attempts have been made to combine the best qualities of both methods by creating a thinned‐skull window that is polished and reinforced by a glass cover slip that is glued over the thinned area to prevent bone regrowth (Mizrahi et al ., 2004 ; Drew et al ., 2010 ; Dorand et al ., 2014 ). Of interest is that brain imaging methods have also been developed for in vivo study of deeper brain structures such as the hippocampus in mice, as this region is of extensive interest for both physiological and pathological brain functions (Gu et al ., 2014 ). This method is also quite challenging, since it requires surgical removal of the cortical tissue above these deeper brain regions, and even longer waiting periods than those for window implantation for inflammation resolution (Park et al ., 2019 ). All these methods stabilize the organ, using a common fixation approach through either a stereotactic frame or a head‐restraining apparatus. Moreover, translucent substances, such as silicone elastomer or low‐melting point agarose, can fill the space between the cortical surface and cover slip to help maintain their adhesion over extended periods of time (Holtmaat et al ., 2009 ; Yildirim et al ., 2019 ). These substances also minimize the refractive‐indices mismatch between the cover glass and brain tissue, thereby reducing optical aberrations. Generally, brain‐related motion artefacts do not create a major issue in anesthetized mice because of their low amplitude and frequency, ranging from about 2 to 5 cycles/s. However, insufficient anesthesia induces accelerated breathing and heartbeat that translate into visible drift and artefacts. We stress that the level of anesthesia must be carefully monitored during the entire experiment (i.e. surgery and imaging) to avoid pain and distress to animals. Intravital imaging of the brain becomes a bigger challenge in awake behaving animals, in which tissue displacements are more pronounced and difficult to tackle because their amplitude is likely larger than the structures under observation. However, observation of the brain of nonanesthetized animals with optical microscopy better associates behaviours or external stimuli to communication patterns among neuronal populations. For this purpose, calcium (Ca 2+ ) transients and action potentials are generally recorded with genetically encoded probes, adeno‐associated virus vectors or by bulk loading fluorescent indicators into the brain (Garaschuk et al ., 2006 ; Andermann et al ., 2011 ; St‐Pierre et al ., 2014 ). The volume spanned by neuronal networks in specific brain compartments is substantial. Thus, the ideal experimental paradigm enables fast, stable, sensitive and high‐resolution imaging of neuronal structures over large areas and depths. However, for practical and technical considerations these features are hard to achieve simultaneously in vivo , so precedence must be given to some over others. The ideal paradigm primarily discriminates the approaches developed to capture neuronal activity in awake animals, which can be loosely sorted into tabletop and portable systems. Tabletop systems usually involve head‐fixed mice under a conventional microscope and a laser source that allows two‐photon excitation and its associated benefits. In these mice, access to the brain and its restraint are similar to anesthetized mice; however, awake animals are exposed to different environments suitable for studying the motor, visual, auditory or barrel cortices, such as during locomotion or sensory stimulation (Sato et al ., 2007 ; Kerlin et al ., 2010 ; Scott et al ., 2014 ; Nadella et al ., 2016 ). Recent technical developments use new microscope technology to decrease motion artefact by increasing the speed of acquisition. For example, random‐access scanning microscopy enable high‐speed volumetric imaging of brain with reduces intraframe motion artefacts (Botcherby et al ., 2006 ). Another available technic is the use of a Bessel beam to reduce the acquisition time in the optical Z‐axis by increasing the length of the point spread function of the focused excitation beam to generate an extended depth of field image (Song et al ., 2017 ). Recently, Song et al . created a method to address loss of depth information limitation of Bessel beam by splitting the beam. Referred as volumetric Two‐photon Imaging of Neurons using Stereoscopy, the two excitation beams create a stereoscopic 'V'‐shaped point spread function configuration that preserves depth information. The combination of Bessel beam and volumetric Two‐photon Imaging of Neurons using Stereoscopy allows a high‐speed image acquisition and a lower exposure time without loss of 3D information (Birkner et al ., 2017 ). Another method using ultra‐short laser pulses can increase rapidly the fluorescent indicator energy with lower power laser. This method presents another solution to increase the acquisition speed, to decrease the phototoxicity and to increase the depth penetration (Sato et al ., 2017 ). Recently, researchers performed dual‐axis Ca2+ imaging in distinct areas of the visual cortex using gradient refractive index microendoscopes and optogenetic perturbation of targeted neuronal networks. This was performed alongside Ca 2+ imaging that probed the activity in each area during different activity states (Lecoq et al ., 2014 ; Packer et al ., 2015 ). Interestingly, along with image registration, Packer and colleagues used microscope parts that moved the optical path relative to the sample to reduce the impact of motion artefacts (Gulati et al ., 2017 ). This paradigm is also common in portable‐microscope systems that fix fibre‐based and integrated microscopes to the animal's head, thereby allowing for behavioural testing of the mice during imaging, since they are not head‐restrained and can move freely (Ghosh et al ., 2011 ; Helmchen et al ., 2013 ; Ziv et al ., 2013 ; Chen et al ., 2018 ). Moreover, integrated microscopes with light‐emitting diode (LED) illumination and a large depth of field, enable experimenters to perform video‐rate image acquisition of extensive fields of view with fewer effects from motion artefacts (Ziv et al ., 2013 ). When equipped with a microendoscope, miniaturized microscopes support relatively stable Ca 2+ imaging in deep brain areas, such as the hippocampus, while animals are awake and behaving (Kerr & Nimmerjahn, 2012 ; Helmchen et al ., 2013 ; Berdyyeva et al ., 2014 ). Without restraining devices or anesthetic, miniature microscope avoids movement of the head and cause less intraframe motion artefacts (Sawinski et al ., 2009 ; Zong et al ., 2017 ). The main drawbacks associated with these systems are their limited spatial resolution and depth of field, as well as a limited single‐photon/colour excitation and the impossibility of moving the lens during imaging. These drawbacks will undoubtedly be addressed in the near future and should render integrated microscopes as widespread, complementary tools to two‐photon excitation systems mainly for the study of the brain and the spinal cord (Paukert & Bergles, 2012 ; Sekiguchi et al ., 2016 ; Nelson et al ., 2019 ). Gated acquisition and active compensation Despite physical restraint, motion artefacts will likely hinder high‐resolution imaging in organs prone to high‐amplitude displacements. Gated acquisition and active compensation can help address this and provide stable‐imaging conditions, albeit with reduced temporal or spatial resolution. Representative examples are provided in Table 2 . Table 2 Gated acquisition and active compensation systems John Wiley & Sons, Ltd. Gated acquisition Gated acquisition collects images at dedicated time points, when movement amplitude or frequency is minimal. This method is generally used to monitor the heart and lungs, because it can precisely monitor their activity and perform imaging during specific cycles of heartbeat, respiration or both. This idea supports prospective gating. Although gated acquisition helps acquire images without major artefacts, its temporal resolution is compromised and, thus, can overlook phenomena that occur outside of these 'stable' cycles. Nevertheless, this approach is relatively simple to implement and can address many biological questions. Interestingly, Paukert and Bergles showed that microdisplacements occurring in the brain were primarily caused by heartbeat. Based on this finding, they used electrocardiogram‐triggered interlaced scanning to observe dendritic spines (Jenkins et al ., 2013 ). An alternative method to prospective gating involves acquiring as many images as possible while monitoring the animal's electrocardiogram or breathing activity and then selecting frames that were recorded during the cycles of interest. The throughput is even higher when the microscope is equipped with a fast‐scanning system, such as a resonant scanner or spinning polygonal mirror. This approach, coined retrospective gating, revealed high‐resolution interactions between macrophages and dendritic cells in the mouse lung without stabilizing the organ. Of note is that only one image was selected every minutes within 5 s‐acquisition periods, which decreased temporal resolution, but created stable image sequences with good spatial resolution (Kreisel et al ., 2010 ). As a complement to gated acquisition, the heart can be paced with electrodes or infrared laser light to dictate heartbeat frequency as shown here in rabbits and mouse (Laffray et al ., 2011 ; Kelly et al ., 2018 ). However, these procedures require special equipment and change the normal function of the heart, which would not be suitable for many intravital imaging purposes. Active compensation Motion artefacts that stem from relatively high‐amplitude movements can be detected and corrected in real‐time with a number of hardware and software solutions. This concept supports active compensation, in which the microscope stage and/or objective lens move to accommodate sample displacements in 3D. Because microscope stages have high inertia, moving the objective lens is inherently faster and allows the sample to remain stationary, which can be critical when sudden movements occur. A contact‐free sensor of optical displacement has been described to actively synchronize the objective's axial position with the moving surface of the rat spinal cord, enabling stable Ca 2+ imaging in vivo (Lee et al ., 2008a ). With this approach, the circular shape of the reference laser beam becomes hemicircular as it moves away from focus, thus providing an optical means to monitor the position of the tissue with respect to that of the objective. In this case, solely adjusting the objective's axial position (perpendicular to the imaging plane) implies more steps to compensate for X‐Y displacements. One group used a piezoactuator‐driven pentagon with a high‐speed loop that provides visual feedback by tracking beads in the kidney and compensating for planar displacements (Lee et al ., 2008b ). This group also added a cantilever‐based sensor to their system to guide objective displacement in the axial direction (Schroeder et al ., 2010 ). Despite notably reduced amplitude of X‐Y movement, this approach requires contact with the sample, which lessens compensation for vertical motion and relies on charge‐coupled device camera detection. Therefore, this approach would require extensive hardware modifications to be applied to two‐photon excitation microscopy. Methods based on image analysis can also actively monitor changes in reference volumes and adjust the position of the stage and/or the objective lens in real time (Bakalar et al ., 2012 ; Lee et al ., 2014 ). Schroeder and colleagues' design is based on 2D cross‐correlation between two orthogonal imaging planes in a given volume and can correct for X‐Y displacements only with a speed limited to 2 µm/s ( Bakalar et al ., 2012 ). The authors improved the performance of their system with resonant scanners for video‐rate acquisition coupled to real‐time calculations of 3D cross‐correlation between volumes of interest over time (Lee et al ., 2014 ). They performed calculations on graphical processing units to estimate tissue motion from entire z‐stacks (i.e. in 3D), providing considerably more information for the motion compensation algorithm than 2D‐imaging planes. However, this gain in accuracy comes at the expense of increased computational time, implying that high‐resolution images might need to be scaled‐down to keep up with the large amount of data generated by the resonant scanner. This issue will undoubtedly be addressed in the near future. This innovative approach is also limited by the displacement range of the objective (limited to about 150 µm) and less‐efficient motion compensation for displacement rates greater than 200 µm/min, which might preclude its use in some organs, such as the gut or heart (Lee et al ., 2014 ). Gated acquisition Gated acquisition collects images at dedicated time points, when movement amplitude or frequency is minimal. This method is generally used to monitor the heart and lungs, because it can precisely monitor their activity and perform imaging during specific cycles of heartbeat, respiration or both. This idea supports prospective gating. Although gated acquisition helps acquire images without major artefacts, its temporal resolution is compromised and, thus, can overlook phenomena that occur outside of these 'stable' cycles. Nevertheless, this approach is relatively simple to implement and can address many biological questions. Interestingly, Paukert and Bergles showed that microdisplacements occurring in the brain were primarily caused by heartbeat. Based on this finding, they used electrocardiogram‐triggered interlaced scanning to observe dendritic spines (Jenkins et al ., 2013 ). An alternative method to prospective gating involves acquiring as many images as possible while monitoring the animal's electrocardiogram or breathing activity and then selecting frames that were recorded during the cycles of interest. The throughput is even higher when the microscope is equipped with a fast‐scanning system, such as a resonant scanner or spinning polygonal mirror. This approach, coined retrospective gating, revealed high‐resolution interactions between macrophages and dendritic cells in the mouse lung without stabilizing the organ. Of note is that only one image was selected every minutes within 5 s‐acquisition periods, which decreased temporal resolution, but created stable image sequences with good spatial resolution (Kreisel et al ., 2010 ). As a complement to gated acquisition, the heart can be paced with electrodes or infrared laser light to dictate heartbeat frequency as shown here in rabbits and mouse (Laffray et al ., 2011 ; Kelly et al ., 2018 ). However, these procedures require special equipment and change the normal function of the heart, which would not be suitable for many intravital imaging purposes. Active compensation Motion artefacts that stem from relatively high‐amplitude movements can be detected and corrected in real‐time with a number of hardware and software solutions. This concept supports active compensation, in which the microscope stage and/or objective lens move to accommodate sample displacements in 3D. Because microscope stages have high inertia, moving the objective lens is inherently faster and allows the sample to remain stationary, which can be critical when sudden movements occur. A contact‐free sensor of optical displacement has been described to actively synchronize the objective's axial position with the moving surface of the rat spinal cord, enabling stable Ca 2+ imaging in vivo (Lee et al ., 2008a ). With this approach, the circular shape of the reference laser beam becomes hemicircular as it moves away from focus, thus providing an optical means to monitor the position of the tissue with respect to that of the objective. In this case, solely adjusting the objective's axial position (perpendicular to the imaging plane) implies more steps to compensate for X‐Y displacements. One group used a piezoactuator‐driven pentagon with a high‐speed loop that provides visual feedback by tracking beads in the kidney and compensating for planar displacements (Lee et al ., 2008b ). This group also added a cantilever‐based sensor to their system to guide objective displacement in the axial direction (Schroeder et al ., 2010 ). Despite notably reduced amplitude of X‐Y movement, this approach requires contact with the sample, which lessens compensation for vertical motion and relies on charge‐coupled device camera detection. Therefore, this approach would require extensive hardware modifications to be applied to two‐photon excitation microscopy. Methods based on image analysis can also actively monitor changes in reference volumes and adjust the position of the stage and/or the objective lens in real time (Bakalar et al ., 2012 ; Lee et al ., 2014 ). Schroeder and colleagues' design is based on 2D cross‐correlation between two orthogonal imaging planes in a given volume and can correct for X‐Y displacements only with a speed limited to 2 µm/s ( Bakalar et al ., 2012 ). The authors improved the performance of their system with resonant scanners for video‐rate acquisition coupled to real‐time calculations of 3D cross‐correlation between volumes of interest over time (Lee et al ., 2014 ). They performed calculations on graphical processing units to estimate tissue motion from entire z‐stacks (i.e. in 3D), providing considerably more information for the motion compensation algorithm than 2D‐imaging planes. However, this gain in accuracy comes at the expense of increased computational time, implying that high‐resolution images might need to be scaled‐down to keep up with the large amount of data generated by the resonant scanner. This issue will undoubtedly be addressed in the near future. This innovative approach is also limited by the displacement range of the objective (limited to about 150 µm) and less‐efficient motion compensation for displacement rates greater than 200 µm/min, which might preclude its use in some organs, such as the gut or heart (Lee et al ., 2014 ). Image processing Postprocessing is generally an integral part of intravital imaging experiments. It is indispensable for most organs despite best efforts to suppress and compensate for tissue motion; however, the amount of processing required varies by organ and usually depends on the structures observed and type of data to be harvested. Some software solutions with key features for two‐photon intravital microscopy are suggested in Table 3 . Table 3 Available software and algorithms for TP‐IVM image processing Orange = algorithms requiring programming skills; Blue = Graphic User Interface available; Green = implemented feature; Red = Not implemented feature. Abbreviations: IMART, intravital microscopy artefact reduction tool; NMC‐CKF, nonlinear motion compensation algorithm using a cubature Kalman filter; SeNeCA, Search for Neural Cells Accelerated; SIMA, Sequential IMaging Analysis. John Wiley & Sons, Ltd. Commonly used operations include image filtering for noise reduction, artefact removal, features detection and image reconstruction and registration (alignment). Different experimental paradigms call for many image‐processing approaches. Of note is that image reconstruction and registration do not provide universal solutions to compensate for any type of artefact. Instead, the organs under observation, as well as the amplitude and time scale over which artefacts occur, dictate the optimal method to use. When movement amplitude is large and precludes regions of interest that will be registered between successive frames, whole‐frame removal is generally required, as often occurs when imaging the gut, heart, lung and spinal cord. Furthermore, different similarity metrics can be applied to multidimensional image sequences to automatically remove unwanted frames according to user‐defined parameters after acquisition is complete (Soulet et al ., 2013 ). Alternatively, piecewise image reconstruction can be performed offline to stitch together artefact‐free image segments from different times during acquisition (Santamaría‐Pang et al ., 2015 ). This approach reduces temporal resolution, but it also allows image collection in moving organs without the need for triggered acquisition schemes and complex instrumentation. As another option, the machine learning revolution makes it possible to create algorithms for image reconstruction predicting morphological structures. Even if it needs a specific algorithm for a specific structure recognition, the possibility of segmenting the object of interest within the surrounding noise allows a high accuracy for 3D reconstruction (Kumar et al ., 2013 ; Haft‐Javaherian et al ., 2019 ). Moreover, algorithms can be created to optimize geometric transformations applied to a target image to maximize its similarity to a reference image. For example, many image‐registration algorithms can align frames of a video sequence in which movement artefacts are present (interframe artefacts). These algorithms can be successfully applied to align images forming a time series or z‐stack, wherein each frame is aligned with the one preceding it or with a reference frame. There are a few broad approaches to image registration. One is based on feature detection and seeks to match specific sets of landmarks (e.g. points, lines, edges) from one image to another (Ishijima et al ., 2015 ; Aghayee et al ., 2017 ). Aghayee and colleagues used motion correction by cell tracking. They target the brightest neurons and track its position over time using particle finding and tracking algorithm, including deformations compensation (Lorenz et al ., 2012 ). Kumar and colleagues took advantage of feature detection and developed a general framework to correct for feature‐based motion using a vessel‐ness filter to identify blood vessels in the mouse pancreas (Ishijima et al ., 2015 ). This method is quite robust and significantly reduces movement artefacts from multichannel time series; however, it requires considerable computation time, which reduces performance for large data sets. Using this framework, motion artefacts that occurred on a global scale (i.e. affected fields of view from one frame to another in a similar fashion) were successfully corrected with rigid transformations, whereas local deformations (i.e. present in distinct sections from each frame) were corrected with nonrigid transformations. Another approach involves nonrigid registration algorithms, which correct nonuniform localized deformations well; however, care must be taken when interpreting data, especially when imaging small motile structures, such as microglial processes or dendritic spines. Indeed, these algorithms are unlikely to correct local deformations similarly and may digitally cancel, overlook or even create them altogether. Therefore, nonrigid registration algorithms might be unsuitable for some experimental paradigms (Viergever et al ., 2016 ). Alternatively, intensity‐based registration, which compares intensity patterns among images, does not require manual input to be carried out and, thus, can be easily automated (Zitová & Flusser, 2003 ; Rohde et al ., 2005 ). Common metrics used to assess similarity include the difference of squared intensity, cross‐correlation and the sum of the log of absolute differences in addition to mutual information, which is most suitable when frames to be registered are captured with different imaging modalities (Zitová & Flusser, 2003 ; Dombeck et al ., 2007 ; Loeckx et al ., 2010 ; Jenkins et al ., 2013 ; Viergever et al ., 2016 ). Given the raster‐scanning pattern used to generate images in two‐photon intravital microscopy, artefacts may occur while acquiring individual frames (intraframe artefacts). To limit the burden caused by these artefacts, different methods have been used, including the Lucas–Kanade framework, Hidden‐Markov Model and the Sequential IMaging Analysis Python package, which address artefacts on a line‐by‐line basis (Greenberg & Kerr, 2009 ; He et al ., 2013 ; Kaifosh et al ., 2014 ). Another method used to reduce intraframe artefacts is the cubature‐Kalman‐filter modelling in a nonlinear system whose constraints help the registration algorithm estimate geometrical transformations (Vercauteren et al ., 2006 ; Greenberg & Kerr, 2009 ; Kaifosh et al ., 2014 ). Moreover, Vercauteren and colleagues used the relationship between distortions and the motion that caused them to compensate for artefacts generated by a fibre‐based optical microprobe (Dunn et al ., 2014 ). In addition to the custom algorithms presented above, and commercial products, front‐end software tools that are freely available have been developed to automate artefact removal and image registration in multidimensional data sets from intravital microscopy (Bethge et al ., 2013 ; Soulet et al ., 2013 ; Santamaría‐Pang et al ., 2015 ). For comparison purposes, refer to Table 4 . Table 4 Algorithms and freeware available for image registration suitable for TP‐IVM John Wiley & Sons, Ltd. General guidelines, tips and tricks Choosing a suitable two‐photon intravital microscopy approach depends on the biological event to be investigated and the organs being imaged. Here, we discuss the most feasible approaches depending on the type of organ and motion artefact, and provide examples in which various technical solutionshave been successfully used (see Fig. 1 and corresponding Tables 1 , 2 , 3 , 4 ). Fig. 1 Intravital multiphoton procedures and considerations. Considering the numerous TP‐IVM applications, a single flowchart cannot be used to illustrate every possible scenario. Therefore, a general layout is depicted, emphasizing considerations related to the organ under study, typical artefacts encountered, restraining methods and possible requirements for hardware‐based active motion compensation, as well as software solutions. This figure can be used as a general guideline to reduce the impact of motion artefacts during intravital imaging sessions. MIP, maximum intensity projection; ROI, region of interest; TP‐IVM, two‐photon intravital microscopy. Concluding remarks Although it is virtually impossible to completely eliminate the physiological causes of movement artefacts, efficient mechanical, technical and software approaches have been developed to limit their burden and facilitate data interpretation. With the ongoing advances in molecular, optical and genetic tools that are available to access, identify and study biological mechanisms in increasingly demanding physiological environments and conditions, there is no doubt that two‐photon intravital microscopy will occupy a front‐line position in the future of biomedical research (Tomek et al ., 2013 ; Pfeiffer et al ., 2018 ; Soltanian‐Zadeh et al ., 2019 ). Of note is that technical developments related to light access, acquisition, processing speed and signal collection efficiency will increase the amount of data being collected, thereby stressing the need for robust hardware and software solutions to efficiently manipulate the imaging data. We believe that active motion compensation in real time will become increasingly popular as computational resources and calculation power expand to enable immediate diagnosis and correction of artefacts for greater experimental throughput and more direct appreciation of generated imaging data. Contributions D.S., J.L.‐P., B.A and D.D. wrote the manuscript. J.L‐P., B.A and D.S. prepared the tables and J.L‐P. and D.S. prepared the figures. Competing financial interests The authors declare no competing financial interests.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3675143/

Characterization of Putative Cholesterol Recognition/Interaction Amino Acid Consensus-Like Motif of Campylobacter jejuni Cytolethal Distending Toxin C

Cytolethal distending toxin (CDT) produced by Campylobacter jejuni comprises a heterotrimeric complex formed by CdtA, CdtB, and CdtC. Among these toxin subunits, CdtA and CdtC function as essential proteins that mediate toxin binding to cytoplasmic membranes followed by delivery of CdtB into the nucleus. The binding of CdtA/CdtC to the cell surface is mediated by cholesterol, a major component in lipid rafts. Although the putative cholesterol recognition/interaction amino acid consensus (CRAC) domain of CDT has been reported from several bacterial pathogens, the protein regions contributing to CDT binding to cholesterol in C. jejuni remain unclear. Here, we selected a potential CRAC-like region present in the CdtC from C. jejuni for analysis. Molecular modeling showed that the predicted functional domain had the shape of a hydrophobic groove, facilitating cholesterol localization to this domain. Mutation of a tyrosine residue in the CRAC-like region decreased direct binding of CdtC to cholesterol rather than toxin intermolecular interactions and led to impaired CDT intoxication. These results provide a molecular link between C. jejuni CdtC and membrane-lipid rafts through the CRAC-like region, which contributes to toxin recognition and interaction with cholesterol. Introduction Campylobacter jejuni is a Gram-negative bacterium that commonly causes diarrhea in humans worldwide [1] . C. jejuni -associated enterocolitis is typically associated with a local acute inflammatory response that involves intestinal tissue damage [2] . Several bacterial virulence factors of C. jejuni , including adhesion molecules, flagella, and cytotoxins, have been investigated for their roles in host pathogenesis [3] . Although cytolethal distending toxin (CDT) from C. jejuni has been characterized [4] , the molecular mechanisms underlying CDT involvement in C. jejuni -induced host pathogenesis requires further investigation. CDT is a bacterial genotoxin consisting of a heterotrimeric complex comprising CdtA, CdtB, and CdtC [5] . CDT holotoxin, which is produced by various important Gram-negative bacteria, has been well characterized [6] . Several studies have shown that CdtA and CdtC are essential for mediating toxin binding to the cytoplasmic membrane of target cells [5] , [7] , [8] . Upon binding to the cell membrane, CdtB is internalized into the cells and is further translocated into the nucleus [9] . Irrespective of the bacterial species, the nuclear-translocated CdtB contains type I deoxyribonuclease activity that can cause double-strand DNA breakage (DSB) followed by cell cycle arrest at G2/M [10] . These insights into the biological function of the CDT holotoxin have identified CDT as an essential factor for C. jejuni -induced pathogenesis in host cells [11] . Increasing evidence has demonstrated that CdtA and CdtC form a heterodimeric complex that enhances attachment of the toxin to cell membranes [5] , [7] , [12] . The presence of a C-terminal non-globular structure in CdtA and CdtC is important for toxin assembly and attachment to the cell membrane [8] . Structural studies have shown that CdtA and CdtC exhibit ricin-like lectin structures [10] , indicating that membrane glycoproteins contribute to the binding of CDT to cells [13] . A recent study demonstrated that both membrane carbohydrates and cholesterol play a critical role in CDT binding to cultured cells [14] . As shown by several studies, reduction in membrane cholesterol levels prevents CdtA/CdtC from binding to target cells and results in attenuated CDT intoxication [15] , [16] . In previous studies on toxin interactions with cholesterol-rich microdomains, CdtC from Aggregatibacter actinomycetemcomitans was found to contain a cholesterol recognition/interaction amino acid consensus (CRAC) region [L/V(X) 1–5 Y(X) 1–5 R/K] that is important for toxin binding and facilitating endocytosis of CdtB [17] . These lines of evidence support the hypothesis that CdtA/CdtC might harbor a unique motif required for toxin binding to cholesterol. Although putative sequences of C. jejuni CdtA/CdtC required for binding to cultured cells have been reported [7] , the exact protein regions contributing to toxin recognition and interaction with cholesterol have not yet been determined. Our recent study has shown that cholesterol provides a platform for C. jejuni CDT intoxication of cells [16] ; however, the molecular mechanism for the interaction of C. jejuni CdtA/CdtC with cholesterol remains unknown. In this study, we examined the potential CRAC-like region present in CdtC from C. jejuni and functionally assessed this candidate cholesterol-binding motif in CdtC. Mutational analysis of the CRAC-like region showed that a tyrosine residue is essential for CdtC membrane binding but not for toxin assembly. Our results further indicated that a putative CRAC-like region is present in C. jejuni CdtC, which contributes to the interaction with membrane cholesterol-rich microdomains and facilitates toxin intoxication. Materials and Methods Reagents and antibodies Antibody against proliferating cell nuclear antigen (PCNA) was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-actin mouse monoclonal antibody was purchased from Upstate Biotechnology (Lake Placid, NY). Alexa Fluor 488-conjugated anti-mouse IgG was purchased from Invitrogen (Carlsbad, CA). Antiserum against each CDT subunit was prepared as described previously [16] . All other chemicals, water-soluble cholesterol, and cholesterol depletion agent–methyl-β-cyclodextrin (MβCD) were purchased from Sigma-Aldrich (St. Louis, MO). Construction of cdtC Y81P and cdtC T163A·L164A mutants cdtC ligated pET21d [16] was utilized as the template for mutagenesis. Amino acid substitution was introduced into the cdtC gene by site-directed mutagenesis. The forward and reverse oligonucleotide primers used for amplification of cdtC Y81P were cdtC-F ( 5′-GAACTTCCTTTTGGTCCTGTGCAATTTAC-3′ ) and cdtC-R ( 5′-GTAAATTGCACAGGACC AAAAGGAAGTTC-3′ ). The oligonucleotide primers used for generation of cdtC T163A·L164A were forward: 5′-CTTTGGAATAGCCCCTTGCGCCGCAGATCCTATTTTTT-3′ and reverse: 5′-CTTTGGAA TAGCCCCTTGCGCCGCAGATCCTATTTTTT-3′ . Amplification of cdtC mutant was carried out by using the QuikChange II site-directed mutagenesis system (Stratagene, Santa Clara, CA). The mutation of cdtC was verified by DNA sequencing. Purification of CDT Subunits Each recombinant His-tagged CDT subunit was cloned and prepared as previously described [16] . Briefly, E. coli BL21-DE3 cells harboring CdtA, CdtB, CdtC or CdtC Y81P expression plasmids were induced by 0.5 mM of isopropyl β-D-thiogalactopyranoside (IPTG) at 37°C for 3 h. The expressed His-tagged CdtA, CdtB, and CdtC fusion proteins were purified by metal affinity chromatography (Clontech, Palo-Alto, CA) and assessed by SDS-PAGE and western blot. SDS-PAGE and Western Blot Analyses To test the reconstitution of CDT holotoxin, each recombinant Cdt subunit (200 nM) was prepared and incubated at 37°C for 5 min allowed to assemble followed by incubation with cells [16] . CDT holotoxin-treated cells were then washed three times with PBS and boiled in SDS-PAGE sample buffer for 5 min. The samples were resolved by 12% SDS-PAGE and transferred onto polyvinylidene difluoride membranes (Millipore, Billerica, MA). The membranes were incubated with each antiserum against each CDT subunit followed by incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies (Invitrogen). The proteins of interest were detected using the ECL Western Blotting Detection Reagents (GE Healthcare, Piscataway, NJ) and detected using X-ray film (Kodak, Rochester, NY). Structural simulation The structure-based virtual docking of cholesterol for target protein was described previously with a slight modification [18] . To build the cavity model of C. jejuni CdtC, the H. ducreyi CdtC (Protein Data Bank Code: 1SR4 [10] ) was employed as a template using homology detection tool, SWISS-MODEL [19] . The initial moiety of docked cholesterol into predicted CRAC-domain cavity was carried out using GEMDOCK [20] . Energy minimization on both the predicted CdtC model and the initial moiety were prepared by Discovery Studio v3.0 ( http://accelrys.com/products/discoverystudio/ ). To further refine the initial docked model through molecular dynamics, the final predicted docked model was retrieved using CDOCKER with CHARMm force field [21] . Structural figures were generated with the program PyMol ( http://www.pymol.org ). Dot Blot Analysis The binding activities of CdtC wt and CdtC Y81P to cholesterol were analyzed by dot blot as described previously [18] . Briefly, the polyvinylidene fluoride membranes (Millipore, Billerica, MA) were prepared, and a series concentrations of water-soluble cholesterol (0, 1.56, 3.13, 6.25, 12.5, 25, 50, 100, 200 µM) (Sigma-Aldrich) were added onto membranes at the center of grid with vacuums. The membranes were blocked by 3% BSA in PBS for 1 h followed by incubated with 200 nM CdtC wt or CdtC Y81P at room temperature for 2 h. The membranes were washed with PBS and probed with anti-CdtC antiserum and anti-mouse-HRP antibody (Santa Cruz) at room temperature for 1 h, respectively. The images were visualized by using Image Quant LAS-4000 (Fujifilm, Tokyo, Japan). The relative density of images was quantified by using UN-SCAN-IT software (Silk Scientific Corporation, Orem, UT). Cell Culture CHO-K1 cells (Chinese hamster ovary cells, CCL-61; American Type Culture Collection, Manassas, VA) and AGS cells (human gastric adenocarcinoma cells, CRL 1739) were cultured in F12 medium (HyClone, Logan, UT). COLO205 cells (human colon adenocarcinoma cells, CCL-222) were cultured in RPMI 1640 medium (Invitrogen). All culture media were supplemented with 10% complement-inactivated fetal bovine serum (HyClone, Logan, UT) and penicillin/streptomycin (Invitrogen). The cells were maintained at 37 °C in a humid atmosphere containing 5% CO 2 . Cell Binding Assay CHO-K1 cells were exposed to 200 nM CDT holotoxin or an individual CDT subunit at 4°C for 2 h. The cells were washed twice with ice-cold PBS and fixed with 1% paraformaldehyde (Sigma-Aldrich) for 30 min. The cells were washed three times, and then incubated with anti-CdtB, or anti-CdtC antisera followed by Alexa Fluor 488-conjugated anti-mouse IgG (Invitrogen). The stained cells were subjected to cell cycle analysis using an FACSCalibur flow cytometer (Becton Dickinson, San Jose, CA). The data were analyzed using Cell Quest software WinMDI (Verity Software House). All samples were examined in triplicate from three independent experiments. Immunofluorescence CHO-K1 cells were plated at a density of 5×10 4 in six-well plates and incubated for 24 h. Cells were exposed to 200 nM of CDT subunit (CdtC wt or CdtC Y81P ) or CDT holotoxin (CdtABC wt or CdtABC Y81P , 200 nM each subunit) at 11°C. After 1 h, the cells were washed and fixed in 1% paraformaldehyde (Sigma-Aldrich) for 30 min followed by permeabilized with 0.1% Triton X-100 for 30 min. Cells were incubated with and anti-CdtB, anti-CdtC antisera and probed with Alexa Fluor 488-conjugated anti-mouse IgG (Invitrogen). The prepared samples were then observed by a confocal laser-scanning microscope (Zeiss LSM 510; Carl Zeiss, Göttingen, Germany) with a 100× objective (oil immersion; aperture, 1.3). Isolation of Nuclear Fraction To explore the localization of CdtB in the nucleus of target cells, CHO-K1 cells were exposed to 200 nM CdtABC wt or CdtABC Y81P holotoxin at 37°C for 4 h. The nuclear proteins were isolated using a nuclear extraction kit (Pierce, Rockford, IL). All protein concentrations were determined by colorimetric assay using the Bio-Rad assay kit (Bio-Rad, Hercules, CA). The isolated proteins (30 µg) from the nuclear fractions were then subjected to western blot for analysis of CdtB localization. Cell Cycle Analysis Cells were treated with CdtABC wt or CdtABC Y81P holotoxin for 24 h. Cells were harvested and fixed with ice-cold 70% ethanol for 2 h and stained with 20 µg/ml propidium iodide (Sigma-Aldrich) containing 1 mg/ml RNase (Sigma-Aldrich) and 0.1% Triton X-100 for 1 h. The stained cells were analyzed with an FACSCalibur flow cytometer (Becton Dickinson, San Jose, CA). The data were collected using 10,000 cells from each sample and analyzed using Cell Quest software WinMDI (Verity Software House, Topsham, ME). All samples were examined in triplicate from three independent experiments. Statistical Analysis The Student's t -test was used to calculate the statistical significance of experimental results between two groups. A P value of less than 0.05 was considered statistically significant. Reagents and antibodies Antibody against proliferating cell nuclear antigen (PCNA) was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-actin mouse monoclonal antibody was purchased from Upstate Biotechnology (Lake Placid, NY). Alexa Fluor 488-conjugated anti-mouse IgG was purchased from Invitrogen (Carlsbad, CA). Antiserum against each CDT subunit was prepared as described previously [16] . All other chemicals, water-soluble cholesterol, and cholesterol depletion agent–methyl-β-cyclodextrin (MβCD) were purchased from Sigma-Aldrich (St. Louis, MO). Construction of cdtC Y81P and cdtC T163A·L164A mutants cdtC ligated pET21d [16] was utilized as the template for mutagenesis. Amino acid substitution was introduced into the cdtC gene by site-directed mutagenesis. The forward and reverse oligonucleotide primers used for amplification of cdtC Y81P were cdtC-F ( 5′-GAACTTCCTTTTGGTCCTGTGCAATTTAC-3′ ) and cdtC-R ( 5′-GTAAATTGCACAGGACC AAAAGGAAGTTC-3′ ). The oligonucleotide primers used for generation of cdtC T163A·L164A were forward: 5′-CTTTGGAATAGCCCCTTGCGCCGCAGATCCTATTTTTT-3′ and reverse: 5′-CTTTGGAA TAGCCCCTTGCGCCGCAGATCCTATTTTTT-3′ . Amplification of cdtC mutant was carried out by using the QuikChange II site-directed mutagenesis system (Stratagene, Santa Clara, CA). The mutation of cdtC was verified by DNA sequencing. Purification of CDT Subunits Each recombinant His-tagged CDT subunit was cloned and prepared as previously described [16] . Briefly, E. coli BL21-DE3 cells harboring CdtA, CdtB, CdtC or CdtC Y81P expression plasmids were induced by 0.5 mM of isopropyl β-D-thiogalactopyranoside (IPTG) at 37°C for 3 h. The expressed His-tagged CdtA, CdtB, and CdtC fusion proteins were purified by metal affinity chromatography (Clontech, Palo-Alto, CA) and assessed by SDS-PAGE and western blot. SDS-PAGE and Western Blot Analyses To test the reconstitution of CDT holotoxin, each recombinant Cdt subunit (200 nM) was prepared and incubated at 37°C for 5 min allowed to assemble followed by incubation with cells [16] . CDT holotoxin-treated cells were then washed three times with PBS and boiled in SDS-PAGE sample buffer for 5 min. The samples were resolved by 12% SDS-PAGE and transferred onto polyvinylidene difluoride membranes (Millipore, Billerica, MA). The membranes were incubated with each antiserum against each CDT subunit followed by incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies (Invitrogen). The proteins of interest were detected using the ECL Western Blotting Detection Reagents (GE Healthcare, Piscataway, NJ) and detected using X-ray film (Kodak, Rochester, NY). Structural simulation The structure-based virtual docking of cholesterol for target protein was described previously with a slight modification [18] . To build the cavity model of C. jejuni CdtC, the H. ducreyi CdtC (Protein Data Bank Code: 1SR4 [10] ) was employed as a template using homology detection tool, SWISS-MODEL [19] . The initial moiety of docked cholesterol into predicted CRAC-domain cavity was carried out using GEMDOCK [20] . Energy minimization on both the predicted CdtC model and the initial moiety were prepared by Discovery Studio v3.0 ( http://accelrys.com/products/discoverystudio/ ). To further refine the initial docked model through molecular dynamics, the final predicted docked model was retrieved using CDOCKER with CHARMm force field [21] . Structural figures were generated with the program PyMol ( http://www.pymol.org ). Dot Blot Analysis The binding activities of CdtC wt and CdtC Y81P to cholesterol were analyzed by dot blot as described previously [18] . Briefly, the polyvinylidene fluoride membranes (Millipore, Billerica, MA) were prepared, and a series concentrations of water-soluble cholesterol (0, 1.56, 3.13, 6.25, 12.5, 25, 50, 100, 200 µM) (Sigma-Aldrich) were added onto membranes at the center of grid with vacuums. The membranes were blocked by 3% BSA in PBS for 1 h followed by incubated with 200 nM CdtC wt or CdtC Y81P at room temperature for 2 h. The membranes were washed with PBS and probed with anti-CdtC antiserum and anti-mouse-HRP antibody (Santa Cruz) at room temperature for 1 h, respectively. The images were visualized by using Image Quant LAS-4000 (Fujifilm, Tokyo, Japan). The relative density of images was quantified by using UN-SCAN-IT software (Silk Scientific Corporation, Orem, UT). Cell Culture CHO-K1 cells (Chinese hamster ovary cells, CCL-61; American Type Culture Collection, Manassas, VA) and AGS cells (human gastric adenocarcinoma cells, CRL 1739) were cultured in F12 medium (HyClone, Logan, UT). COLO205 cells (human colon adenocarcinoma cells, CCL-222) were cultured in RPMI 1640 medium (Invitrogen). All culture media were supplemented with 10% complement-inactivated fetal bovine serum (HyClone, Logan, UT) and penicillin/streptomycin (Invitrogen). The cells were maintained at 37 °C in a humid atmosphere containing 5% CO 2 . Cell Binding Assay CHO-K1 cells were exposed to 200 nM CDT holotoxin or an individual CDT subunit at 4°C for 2 h. The cells were washed twice with ice-cold PBS and fixed with 1% paraformaldehyde (Sigma-Aldrich) for 30 min. The cells were washed three times, and then incubated with anti-CdtB, or anti-CdtC antisera followed by Alexa Fluor 488-conjugated anti-mouse IgG (Invitrogen). The stained cells were subjected to cell cycle analysis using an FACSCalibur flow cytometer (Becton Dickinson, San Jose, CA). The data were analyzed using Cell Quest software WinMDI (Verity Software House). All samples were examined in triplicate from three independent experiments. Immunofluorescence CHO-K1 cells were plated at a density of 5×10 4 in six-well plates and incubated for 24 h. Cells were exposed to 200 nM of CDT subunit (CdtC wt or CdtC Y81P ) or CDT holotoxin (CdtABC wt or CdtABC Y81P , 200 nM each subunit) at 11°C. After 1 h, the cells were washed and fixed in 1% paraformaldehyde (Sigma-Aldrich) for 30 min followed by permeabilized with 0.1% Triton X-100 for 30 min. Cells were incubated with and anti-CdtB, anti-CdtC antisera and probed with Alexa Fluor 488-conjugated anti-mouse IgG (Invitrogen). The prepared samples were then observed by a confocal laser-scanning microscope (Zeiss LSM 510; Carl Zeiss, Göttingen, Germany) with a 100× objective (oil immersion; aperture, 1.3). Isolation of Nuclear Fraction To explore the localization of CdtB in the nucleus of target cells, CHO-K1 cells were exposed to 200 nM CdtABC wt or CdtABC Y81P holotoxin at 37°C for 4 h. The nuclear proteins were isolated using a nuclear extraction kit (Pierce, Rockford, IL). All protein concentrations were determined by colorimetric assay using the Bio-Rad assay kit (Bio-Rad, Hercules, CA). The isolated proteins (30 µg) from the nuclear fractions were then subjected to western blot for analysis of CdtB localization. Cell Cycle Analysis Cells were treated with CdtABC wt or CdtABC Y81P holotoxin for 24 h. Cells were harvested and fixed with ice-cold 70% ethanol for 2 h and stained with 20 µg/ml propidium iodide (Sigma-Aldrich) containing 1 mg/ml RNase (Sigma-Aldrich) and 0.1% Triton X-100 for 1 h. The stained cells were analyzed with an FACSCalibur flow cytometer (Becton Dickinson, San Jose, CA). The data were collected using 10,000 cells from each sample and analyzed using Cell Quest software WinMDI (Verity Software House, Topsham, ME). All samples were examined in triplicate from three independent experiments. Statistical Analysis The Student's t -test was used to calculate the statistical significance of experimental results between two groups. A P value of less than 0.05 was considered statistically significant. Results Generation and Characterization of Wild-type and Mutant CDT Subunits We recently demonstrated that CDT association with CHO-K1 cells requires intact cholesterol-rich microdomains [16] . A specific conserved sequence, the CRAC motif [L/V(X) 1–5 Y(X) 1–5 R/K], may contribute to the association of proteins with cholesterol [22] . To test this, we analyzed the amino acid sequence of CdtC, which contained a putative CRAC-like motif ( 77 LPFGY 81 VQFTNPK 88 ) ( Fig. 1A ). To assess whether this CRAC-like motif is required for CdtC binding to lipid rafts and CDT intoxication of cells, we used site-directed mutagenesis to construct a single residue-substituted mutant. The tyrosine residue that plays an important role for protein binding to cholesterol was thus replaced with a proline residue (Y81P). The mutant and wild-type CDT subunits were then subjected to SDS-PAGE ( Fig. S1A ) and western blot ( Fig. S1B ) analyses. The purity and protein expression levels of CdtC Y81P were similar to those of CdtC wt . The integrity of the toxin complex was then assessed by western blot. As shown in Fig. S1C , both CdtC wt and CdtC Y81P can be assembled stably with other holotoxin elements. 10.1371/journal.pone.0066202.g001 Figure 1 Molecular modeling of the interaction of CdtC wt and cholesterol. (A) Schematic representation of a partial sequence of the CRAC-like motif in wild-type and mutant CdtC. The numbers indicate the positions of the amino acid residues. The putative CRAC-like motifs are in boxes. The amino acids in boldface indicate the residues targeted for substitution. (B) Structural model of cholesterol was in complex with putative CRAC of CdtC wt . The cholesterol was shown in stick and colored in blue. The number of amino acids shown in CdtC wt directly interacted with the cholesterol-binding sites. Oxygen atom localized in cholesterol and amino acids were shown in red. (C) Cholesterol showed hydrophobic interactions with L77, P78, F79, G80, Y81, and V82 (boldface lines in green), and formed a hydrogen bond with V82 (dash line in green). Modeling simulation was performed using PyMol, as described in the Materials and Methods. The CRAC-like Motif is Essential for CdtC Binding to Cholesterol Structure-based virtual docking was employed to assess that cholesterol binding to a CRAC sequence with 12 amino acid residues ( 77 LPFGYVQFTNPK 88 ) of CdtC wt . Docking analysis showed that the putative CRAC-like motif created a hydrophobic groove, which enabled cholesterol to localize to it ( Fig. 1B ). The best favored conformations of cholesterol were found by docking and the surface represented the same helix. In addition, cholesterol was found to be bound by hydrophobic interactions with the protein residues L77, P78, F79, G80, Y81, and V82 and to form a hydrogen bond with V82 which shown within hydrogen-bonding distance of the cholesterol oxygen (2.8 à ) ( Fig. 1C ). The results of the molecular modeling showed that cholesterol fits into the hydrophobic groove of the CRAC-like motif. To further assess whether the CRAC-like motif played a role in the CdtC-cholesterol interaction, the binding activities of CdtC wt and CdtC Y81P to cholesterol were analyzed by dot blot. As shown in Fig. 2A , the binding activity of CdtC wt to immobilized cholesterol was concentration dependent. In contrast, direct binding to cholesterol was not detected for the CdtC Y81P mutant ( Fig. 2B ). These results indicated that the CRAC-like motif mediates CdtC recognition and cholesterol binding. 10.1371/journal.pone.0066202.g002 Figure 2 Binding of CdtC to cholesterol. (A) Direct binding of wild-type CdtC (CdtC wt ) or mutant CdtC (CdtC Y81P ) at various concentrations of cholesterol were analyzed by dot blot. (B) The binding activities of CdtC to cholesterol were quantified by densitometric analysis in 3 independent experiments. *, P <0.05 was considered as statistically significant. The CRAC-like Motif is Required for the Association of CDT with the Cell Membrane We then analyzed whether the CRAC-like motif is important for the association of CDT subunits with cell membranes. CHO-K1 cells were incubated with CdtC wt or CdtC Y81P for 2 h at 4°C and were analyzed by flow cytometry for the presence of CDT subunits on the cell membrane. As shown in Fig. 3A , CdtC wt was associated with the cell membrane, and the MCF for anti-CdtC was 93.5. However, upon exposure of cells to CdtC Y81P , the MCF for anti-CdtC reduced to 23.5 ( Fig. 3B ). We further tested whether the CRAC-like motif mutant could affect the binding of holotoxin to cells. The levels of MCF for anti-CdtB were 53.5 and 20.6 when the cells were exposed to CdtABC wt and CdtABC Y81P , respectively ( Fig. 3C, D ). Notably, compared with the binding activities of CdtC wt and CdtABC wt , the binding activities of both CdtC Y81P and CdtABC Y81P to cell membranes were significantly lower ( Fig. 3E, F ). 10.1371/journal.pone.0066202.g003 Figure 3 Effects of mutating the CRAC-like region on the binding of CdtC and CDT holotoxin to cells. CHO-K1 cells were treated with 200 nM of (A) CdtC wt , (B) CdtC Y81P , (C) CdtABC wt , or (D) CdtABC Y81P at 4°C for 2 h. The cells were washed and probed with control preimmune serum (gray histograms) or antisera against CdtC (A, B) or CdtB (C, D) (white histograms), followed by staining with Alexa Fluor 488-conjugated anti-mouse IgG. The binding activity was analyzed by flow cytometry. The numbers represent the mean channel fluorescence (MCF). The quantitative data represent the mean and standard deviation values from 3 independent experiments (E, F). The asterisks indicate statistical significance (* P <0.05). We used confocal microscopy to examine whether the binding of CDT to cells was dependent on the CRAC-like motif in CdtC. The cells were treated with CdtC (CdtC wt or CdtC Y81P ) or holotoxin (CdtABC wt or CdtABC Y81P ), followed by probing with preimmune serum and antisera against CdtB or CdtC. No signal for CDT was detected in untreated cells ( Fig. 4 , first row), whereas CdtC wt (green) apparently localized to the area around the plasma membrane ( Fig. 4 , second row). In cells treated with CdtABC wt , membrane distribution of CdtB was evident ( Fig. 4 , fourth row), which was similar to that of cells treated with CdtC wt alone. However, the intensity of detectable fluorescence for CdtC and CdtB on the plasma membrane decreased when cells were treated with CdtC Y81P or CdtABC Y81P ( Fig. 4 , third and fifth rows). These results support our findings of CDT binding activity determined by flow cytometry ( Fig. 3 ), indicating that the CRAC-like motif is critical for CdtC association with cells for CDT intoxication of the cells. 10.1371/journal.pone.0066202.g004 Figure 4 The role of the CRAC-like region in the association of CdtC and holotoxin with cells. CHO-K1 cells were incubated with mock medium alone or with 200 nM of CdtC wt , CdtC Y81P , CdtABC wt , or CdtABC Y81P at 11°C for 1 h. The cells were probed with control preimmune serum (mock) or antisera against CdtC (2 nd and 3 rd rows) or CdtB (4 th and 5 th rows), followed by staining with Alexa Fluor 488-conjugated anti-mouse IgG, and then analyzed by confocal microscopy. Bar, 10 µm. Nuclear Delivery of CdtB Decreases in Cells Treated with CdtABC Y81P We then examined whether the nuclear localization of CdtB was dependent on the CRAC-like motif present in CdtC. The cells were incubated with CdtABC wt or CdtABC Y81P and subjected to western blot analysis. As shown in Fig. 5 , the nuclear localization of CdtB dramatically decreased in cells treated with CdtABC Y81P when compared to the localization in cells treated with CdtABC wt . These data suggested that CdtC associates with cell membranes through the CRAC-like motif and that this association is important for the delivery of CdtB into the nucleus. 10.1371/journal.pone.0066202.g005 Figure 5 Effects of mutations in the CRAC-like motif of CdtC on CdtB nuclear localization. CHO-K1 cells were treated with mock medium alone or with CdtABC wt or CdtABC Y81P (200 nM each subunit) at 37°C for 4 h. (A) Nuclear fractions were prepared and subjected to western blot for analysis of CdtB. Proliferating cell nuclear antigen (PCNA) was used as a loading control for the nuclear fraction of cell lysates. (B) Expression levels of CdtB protein were analyzed using scanning densitometry. The signals of CdtB/PCNA were expressed as relative values. Statistical significance was evaluated using Student's t -test. *, P <0.05 was considered as statistically significant. Mutation of the CRAC-like Motif in CdtC Attenuates CDT Intoxication of Cells To determine whether the CRAC-like motif was required for CDT intoxication of cells, we used flow cytometry to assess cell cycle distributions. In the presence of CdtABC wt , 54% of cells were arrested in G2/M ( Fig. 6C ). However, this cell cycle arrest was attenuated upon treatment of cells with CdtABC Y81P ( Fig. 6D ). The cell cycle distributions were not changed in cells exposed to CdtAC wt and CdtAC Y81P , since the treatment was not contained the toxin activity subunit–CdtB ( Fig. 6E, F ). In contrast, the cell cycle arrest was significantly reduced in cells treated with CdtBC Y81P when compared to cells treated with CdtBC wt ( Fig. 6G, H ). Pretreating cells with MβCD followed by exposure to CdtABC wt dramatically decreased the proportion of cells arrested in G2/M ( Fig. 6I ). These results supported the notion that the CRAC-like motif contributes to the association of CdtC with membrane cholesterol, leading to the intoxication of CDT in target cells. 10.1371/journal.pone.0066202.g006 Figure 6 Attenuation of G2/M arrest by CdtABC Y81P . CHO-K1 cells were incubated with (A) mock medium alone, 200 nM of (B) CdtAB, (C) CdtABC wt , (D) CdtABC Y81P , (E) CdtAC wt , (F) CdtAC Y81P , (G) CdtBC wt , or (H) CdtBC Y81P for 24 h, or (I) pretreated with 10 mM MβCD for 1 h followed by incubation with 200 nM CdtABC wt for 24 h. The cells were stained with propidium iodide, and cell cycle distribution analyzed by flow cytometry. The proportions of cells in the G0/G1, S, and G2/M phases of the cell cycle are shown at the right of each histogram. (J) The percentage of cells in G2/M were calculated and plotted as intensity histograms. The results represent mean and standard deviation values from 3 independent experiments. *, P <0.05 was considered as statistically significant. We next investigated whether mutation of the CRAC-like motif affected CDT-arrested cell cycles in other cell types. Two gastrointestinal cell lines (AGS and COLO205 cells) were used in this experiment. Cells were mock treated or treated with CDT holotoxin (CdtABC wt or CdtABC Y81P ) and analyzed for the cell cycle stage. As shown in Fig. 7 , the proportion of cells accumulating in G2/M in the 3 lines treated with CdtABC Y81P was significantly lower than that in the cells treated with CdtABC wt . These results again demonstrated that the CRAC-like motif present in CdtC is required for the association of CDT holotoxin with cell membranes as well as for the intoxication of target cells. 10.1371/journal.pone.0066202.g007 Figure 7 The role of the CRAC-like region in CDT intoxication of cells. Cells from the indicated lines (A) CHO-K1, (B) AGS, and (C) COLO205 were treated with mock medium, CdtABC wt , or CdtABC Y81P (200 nM each subunit) at 37°C for 24 h. Cell cycle distribution was assessed using flow cytometry. (D) The percentage of cells in G2/M were calculated and plotted as intensity histograms. The results represent 3 independent experiments. The statistical significance of the difference was analyzed using Student's t -test (* P <0.05). Generation and Characterization of Wild-type and Mutant CDT Subunits We recently demonstrated that CDT association with CHO-K1 cells requires intact cholesterol-rich microdomains [16] . A specific conserved sequence, the CRAC motif [L/V(X) 1–5 Y(X) 1–5 R/K], may contribute to the association of proteins with cholesterol [22] . To test this, we analyzed the amino acid sequence of CdtC, which contained a putative CRAC-like motif ( 77 LPFGY 81 VQFTNPK 88 ) ( Fig. 1A ). To assess whether this CRAC-like motif is required for CdtC binding to lipid rafts and CDT intoxication of cells, we used site-directed mutagenesis to construct a single residue-substituted mutant. The tyrosine residue that plays an important role for protein binding to cholesterol was thus replaced with a proline residue (Y81P). The mutant and wild-type CDT subunits were then subjected to SDS-PAGE ( Fig. S1A ) and western blot ( Fig. S1B ) analyses. The purity and protein expression levels of CdtC Y81P were similar to those of CdtC wt . The integrity of the toxin complex was then assessed by western blot. As shown in Fig. S1C , both CdtC wt and CdtC Y81P can be assembled stably with other holotoxin elements. 10.1371/journal.pone.0066202.g001 Figure 1 Molecular modeling of the interaction of CdtC wt and cholesterol. (A) Schematic representation of a partial sequence of the CRAC-like motif in wild-type and mutant CdtC. The numbers indicate the positions of the amino acid residues. The putative CRAC-like motifs are in boxes. The amino acids in boldface indicate the residues targeted for substitution. (B) Structural model of cholesterol was in complex with putative CRAC of CdtC wt . The cholesterol was shown in stick and colored in blue. The number of amino acids shown in CdtC wt directly interacted with the cholesterol-binding sites. Oxygen atom localized in cholesterol and amino acids were shown in red. (C) Cholesterol showed hydrophobic interactions with L77, P78, F79, G80, Y81, and V82 (boldface lines in green), and formed a hydrogen bond with V82 (dash line in green). Modeling simulation was performed using PyMol, as described in the Materials and Methods. The CRAC-like Motif is Essential for CdtC Binding to Cholesterol Structure-based virtual docking was employed to assess that cholesterol binding to a CRAC sequence with 12 amino acid residues ( 77 LPFGYVQFTNPK 88 ) of CdtC wt . Docking analysis showed that the putative CRAC-like motif created a hydrophobic groove, which enabled cholesterol to localize to it ( Fig. 1B ). The best favored conformations of cholesterol were found by docking and the surface represented the same helix. In addition, cholesterol was found to be bound by hydrophobic interactions with the protein residues L77, P78, F79, G80, Y81, and V82 and to form a hydrogen bond with V82 which shown within hydrogen-bonding distance of the cholesterol oxygen (2.8 à ) ( Fig. 1C ). The results of the molecular modeling showed that cholesterol fits into the hydrophobic groove of the CRAC-like motif. To further assess whether the CRAC-like motif played a role in the CdtC-cholesterol interaction, the binding activities of CdtC wt and CdtC Y81P to cholesterol were analyzed by dot blot. As shown in Fig. 2A , the binding activity of CdtC wt to immobilized cholesterol was concentration dependent. In contrast, direct binding to cholesterol was not detected for the CdtC Y81P mutant ( Fig. 2B ). These results indicated that the CRAC-like motif mediates CdtC recognition and cholesterol binding. 10.1371/journal.pone.0066202.g002 Figure 2 Binding of CdtC to cholesterol. (A) Direct binding of wild-type CdtC (CdtC wt ) or mutant CdtC (CdtC Y81P ) at various concentrations of cholesterol were analyzed by dot blot. (B) The binding activities of CdtC to cholesterol were quantified by densitometric analysis in 3 independent experiments. *, P <0.05 was considered as statistically significant. The CRAC-like Motif is Required for the Association of CDT with the Cell Membrane We then analyzed whether the CRAC-like motif is important for the association of CDT subunits with cell membranes. CHO-K1 cells were incubated with CdtC wt or CdtC Y81P for 2 h at 4°C and were analyzed by flow cytometry for the presence of CDT subunits on the cell membrane. As shown in Fig. 3A , CdtC wt was associated with the cell membrane, and the MCF for anti-CdtC was 93.5. However, upon exposure of cells to CdtC Y81P , the MCF for anti-CdtC reduced to 23.5 ( Fig. 3B ). We further tested whether the CRAC-like motif mutant could affect the binding of holotoxin to cells. The levels of MCF for anti-CdtB were 53.5 and 20.6 when the cells were exposed to CdtABC wt and CdtABC Y81P , respectively ( Fig. 3C, D ). Notably, compared with the binding activities of CdtC wt and CdtABC wt , the binding activities of both CdtC Y81P and CdtABC Y81P to cell membranes were significantly lower ( Fig. 3E, F ). 10.1371/journal.pone.0066202.g003 Figure 3 Effects of mutating the CRAC-like region on the binding of CdtC and CDT holotoxin to cells. CHO-K1 cells were treated with 200 nM of (A) CdtC wt , (B) CdtC Y81P , (C) CdtABC wt , or (D) CdtABC Y81P at 4°C for 2 h. The cells were washed and probed with control preimmune serum (gray histograms) or antisera against CdtC (A, B) or CdtB (C, D) (white histograms), followed by staining with Alexa Fluor 488-conjugated anti-mouse IgG. The binding activity was analyzed by flow cytometry. The numbers represent the mean channel fluorescence (MCF). The quantitative data represent the mean and standard deviation values from 3 independent experiments (E, F). The asterisks indicate statistical significance (* P <0.05). We used confocal microscopy to examine whether the binding of CDT to cells was dependent on the CRAC-like motif in CdtC. The cells were treated with CdtC (CdtC wt or CdtC Y81P ) or holotoxin (CdtABC wt or CdtABC Y81P ), followed by probing with preimmune serum and antisera against CdtB or CdtC. No signal for CDT was detected in untreated cells ( Fig. 4 , first row), whereas CdtC wt (green) apparently localized to the area around the plasma membrane ( Fig. 4 , second row). In cells treated with CdtABC wt , membrane distribution of CdtB was evident ( Fig. 4 , fourth row), which was similar to that of cells treated with CdtC wt alone. However, the intensity of detectable fluorescence for CdtC and CdtB on the plasma membrane decreased when cells were treated with CdtC Y81P or CdtABC Y81P ( Fig. 4 , third and fifth rows). These results support our findings of CDT binding activity determined by flow cytometry ( Fig. 3 ), indicating that the CRAC-like motif is critical for CdtC association with cells for CDT intoxication of the cells. 10.1371/journal.pone.0066202.g004 Figure 4 The role of the CRAC-like region in the association of CdtC and holotoxin with cells. CHO-K1 cells were incubated with mock medium alone or with 200 nM of CdtC wt , CdtC Y81P , CdtABC wt , or CdtABC Y81P at 11°C for 1 h. The cells were probed with control preimmune serum (mock) or antisera against CdtC (2 nd and 3 rd rows) or CdtB (4 th and 5 th rows), followed by staining with Alexa Fluor 488-conjugated anti-mouse IgG, and then analyzed by confocal microscopy. Bar, 10 µm. Nuclear Delivery of CdtB Decreases in Cells Treated with CdtABC Y81P We then examined whether the nuclear localization of CdtB was dependent on the CRAC-like motif present in CdtC. The cells were incubated with CdtABC wt or CdtABC Y81P and subjected to western blot analysis. As shown in Fig. 5 , the nuclear localization of CdtB dramatically decreased in cells treated with CdtABC Y81P when compared to the localization in cells treated with CdtABC wt . These data suggested that CdtC associates with cell membranes through the CRAC-like motif and that this association is important for the delivery of CdtB into the nucleus. 10.1371/journal.pone.0066202.g005 Figure 5 Effects of mutations in the CRAC-like motif of CdtC on CdtB nuclear localization. CHO-K1 cells were treated with mock medium alone or with CdtABC wt or CdtABC Y81P (200 nM each subunit) at 37°C for 4 h. (A) Nuclear fractions were prepared and subjected to western blot for analysis of CdtB. Proliferating cell nuclear antigen (PCNA) was used as a loading control for the nuclear fraction of cell lysates. (B) Expression levels of CdtB protein were analyzed using scanning densitometry. The signals of CdtB/PCNA were expressed as relative values. Statistical significance was evaluated using Student's t -test. *, P <0.05 was considered as statistically significant. Mutation of the CRAC-like Motif in CdtC Attenuates CDT Intoxication of Cells To determine whether the CRAC-like motif was required for CDT intoxication of cells, we used flow cytometry to assess cell cycle distributions. In the presence of CdtABC wt , 54% of cells were arrested in G2/M ( Fig. 6C ). However, this cell cycle arrest was attenuated upon treatment of cells with CdtABC Y81P ( Fig. 6D ). The cell cycle distributions were not changed in cells exposed to CdtAC wt and CdtAC Y81P , since the treatment was not contained the toxin activity subunit–CdtB ( Fig. 6E, F ). In contrast, the cell cycle arrest was significantly reduced in cells treated with CdtBC Y81P when compared to cells treated with CdtBC wt ( Fig. 6G, H ). Pretreating cells with MβCD followed by exposure to CdtABC wt dramatically decreased the proportion of cells arrested in G2/M ( Fig. 6I ). These results supported the notion that the CRAC-like motif contributes to the association of CdtC with membrane cholesterol, leading to the intoxication of CDT in target cells. 10.1371/journal.pone.0066202.g006 Figure 6 Attenuation of G2/M arrest by CdtABC Y81P . CHO-K1 cells were incubated with (A) mock medium alone, 200 nM of (B) CdtAB, (C) CdtABC wt , (D) CdtABC Y81P , (E) CdtAC wt , (F) CdtAC Y81P , (G) CdtBC wt , or (H) CdtBC Y81P for 24 h, or (I) pretreated with 10 mM MβCD for 1 h followed by incubation with 200 nM CdtABC wt for 24 h. The cells were stained with propidium iodide, and cell cycle distribution analyzed by flow cytometry. The proportions of cells in the G0/G1, S, and G2/M phases of the cell cycle are shown at the right of each histogram. (J) The percentage of cells in G2/M were calculated and plotted as intensity histograms. The results represent mean and standard deviation values from 3 independent experiments. *, P <0.05 was considered as statistically significant. We next investigated whether mutation of the CRAC-like motif affected CDT-arrested cell cycles in other cell types. Two gastrointestinal cell lines (AGS and COLO205 cells) were used in this experiment. Cells were mock treated or treated with CDT holotoxin (CdtABC wt or CdtABC Y81P ) and analyzed for the cell cycle stage. As shown in Fig. 7 , the proportion of cells accumulating in G2/M in the 3 lines treated with CdtABC Y81P was significantly lower than that in the cells treated with CdtABC wt . These results again demonstrated that the CRAC-like motif present in CdtC is required for the association of CDT holotoxin with cell membranes as well as for the intoxication of target cells. 10.1371/journal.pone.0066202.g007 Figure 7 The role of the CRAC-like region in CDT intoxication of cells. Cells from the indicated lines (A) CHO-K1, (B) AGS, and (C) COLO205 were treated with mock medium, CdtABC wt , or CdtABC Y81P (200 nM each subunit) at 37°C for 24 h. Cell cycle distribution was assessed using flow cytometry. (D) The percentage of cells in G2/M were calculated and plotted as intensity histograms. The results represent 3 independent experiments. The statistical significance of the difference was analyzed using Student's t -test (* P <0.05). Discussion Lipid rafts are membrane microdomains that contained mainly cholesterol, sphingolipids, and phospholipids [23] . Membrane rafts serve as a platform for several bacterial toxins binding to target cells [24] , [25] , [26] , [27] , [28] . The most relevant example is vacuolating cytotoxin A (VacA), one of the major virulence factors secreted by Helicobacter pylori , which was demonstrated to exploit cholesterol-rich microdomains for its assembly on cell membranes and delivery into target cells [29] . Similarly, CDT produced by Haemophilus ducreyi or A. actinomycetemcomitans was found to interact with lipid rafts [15] , [30] . In agreement with these observations, our recent study has shown that CDT from C. jejuni is associated with lipid rafts [16] . These lines of evidence support the hypothesis that cholesterol-rich microdomains may play a critical role in bacterial toxin assembly on cell membranes and intracellular delivery, and that these domains therefore amplify the signaling required for intoxication [31] . Indeed, not all proteins that bind cholesterol harbor CRAC-domains. For instance, the cholesterol-dependent cytolysin family of toxins contains 2 amino acids (threonine and leucine) that are responsible for interacting with cholesterol [32] . At the beginning of this investigation, we first assessed whether the threonine-leucine pair mediated the binding of CdtC to membrane cholesterol. Our results show that mutations at the threonine (T163) and leucine (L164) residues (CdtABC T163A·L164A ) did not affect CDT intoxication, as compared to treatment with CdtABC wt ( Figure S2 ), suggesting that this pair of amino acids does not mediate the CdtC-cholesterol interaction. In this study, the virtual docking simulation showed that C. jejuni CdtC contains a CRAC-like motif ( Fig. 1B ). The 12 amino acid residues LPFGYVQFTNPK created a hydrophobic groove that provided for hydrophobic interactions and hydrogen bonding with cholesterol. In addition, mutation of this domain decreased the cell-binding activity of CdtC. However, comparison with CdtC wt showed that the interaction of CdtC Y81P with CdtA and CdtB in complex formation was not altered ( Fig. S1C ). This finding is due to the fact that the CRAC-like domain does not extend to the CdtC N- and C-terminal regions that contribute to the interaction of CdtC with both CdtA and CdtB [8] . Taken together, our findings have demonstrated that the CRAC-like region involved in CdtC plays a critical role in toxin binding to membrane cholesterol and not in intermolecular interactions between toxin subunits. Our results for the dot blot analysis showed that CdtC Y81P did not bind to immobilized cholesterol ( Fig. 2A ). In addition, our data further demonstrated that cell cycle arrest was significantly lower in cells treated with CdtABC Y81P or CdtBC Y81P than in cells treated with CdtABC wt or CdtBC wt ( Fig. 6 ). These results are supported by the functional analysis of CdtC from A. actinomycetemcomitans [33] and Haemophilus parasuis [34] , indicating that CdtC contains a CRAC-like region that is important for cholesterol binding. However, the binding activity of CdtC Y81P to the cell surface did not completely abolished, as shown by our flow cytometry and confocal microscopy analyses ( Fig. 3 and 4 ), which suggests that CdtC binding to cell membranes is mediated not only by cholesterol but also by other candidate receptors. Previous studies on CDTs from E. coli and A. actinomycetemcomitans indicated a critical role of membrane carbohydrates in toxin interactions [13] , [35] . In addition, structure-based analysis has shown that CdtA and CdtC from A. actinomycetemcomitans have similar structures comprising 3 sets of beta-sheets that are homologous to the B-chain of ricin and contain lectin repeats, indicating that carbohydrates may serve as receptors for CDT [36] . However, a recent study on cell intoxication by CDTs, which were isolated from several bacterial species, demonstrated that glycolipids are not required for CDT intoxication [14] . Although these reports indicate a discrepant role for carbohydrates in membrane association of CDTs, our results are supported by several lines of evidence indicating that cholesterol is one of the candidate receptors for C. jejuni CDT interactions. Further investigations are required to identify the specific receptor(s) that mediate membrane association with C. jejuni CDT. Since the CDTs from different bacterial species have distinct binding activities, they are thought to have divergent target-cell preferences [14] . Although these CDTs originate from different pathogens, hijacking of cholesterol-rich microdomains for toxin function appears to be the universal mechanism underlying CDT action [15] , [16] , [30] . In this study, we have demonstrated that CdtC from C. jejuni contains a CRAC-like region that contributes to the CdtC interaction with cholesterol. Furthermore, the mutation of a tyrosine residue in the CRAC-like region impairs CdtC binding to and inhibits its intoxication of target cells. Elucidation of the target receptor for C. jejuni CDT would lead to better understanding of the molecular mechanisms underlying bacterial pathogenesis in host cells. The results from this study also shed light on the discovery of a novel strategy for specifically inhibiting this toxin. Supporting Information Figure S1 Characterization of wild-type and mutant CDT subunits from C. jejuni . (A) Each CDT subunit (2 µg/ml) was analyzed by SDS-PAGE. (B) Western blot analysis of each CDT subunit detected via antisera against CdtA, CdtB, or CdtC. Molecular mass markers (kDa) are shown on the left. (C) In vitro assay for the toxin assembly of CdtABC wt and CdtABC Y81P at 25°C for 1 h, determined by western blot. (TIF) Click here for additional data file. Figure S2 Threonine (T163) and leucine (L164) of CdtC are not required for CDT intoxication. CHO-K1 cells were treated with (A) mock medium alone, (B) CdtABC wt , and (C) CdtABC T163A·L164A (200 nM each subunit) for 24 h. The treated cells were stained with propidium iodide, and the cell cycle distribution was analyzed by flow cytometry. The proportions of cells in the G0/G1, S, and G2/M phases of the cell cycle are shown at the right of each histogram. (TIF) Click here for additional data file.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3073144/

Tissue expression of steroid hormone receptors is associated with differential immune responsiveness

Glucocorticoids and other steroid hormones have been used as treatments against a number of diseases, especially inflammatory conditions in which the immune system is overactive. These treatments have varying degrees of responsiveness among individuals and in different tissues (including brain); therefore, it is important to determine what could account for these differences. In this study, we evaluated expression of steroid hormone receptors in immune cells from lymphoid and non-lymphoid tissues as a possible explanation for tissue-specific differences. We analyzed leukocytes (CD45 + ) in kidney, liver, spleen, and thymus tissues from healthy mice for expression of the receptor for stress hormone (glucocorticoid - GR) as well as other steroid hormones (androgen - AR, progesterone - PR) and found that all tissues expressed these steroid hormone receptors but with varying expression patterns. To determine whether tissue-specific differences were related to immune cell composition, we examined steroid hormone receptor expression in T lymphocytes from each of these tissues and found similar patterns of expression in these cells regardless of tissue source. Because glucocorticoids can also impact brain function, we further examined expression of the stress hormone receptor in brain tissue and found GR expressed in immune cells at this site. In order to investigate the potential impact in an area of neuropathology, we utilized a mouse model of West Nile Virus (WNV). We observed pathological changes in brains of WNV-infected animals and T lymphocytes in the areas of inflammation; however, these cells did not express GR. These data indicate that tissue-specific differences in steroid hormone receptor expression by immune cells could determine responsiveness with steroid hormone treatment.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6215251/

Structural and Functional Analyses of the Shedding Protease ADAM17 in HoxB8-Immortalized Macrophages and Dendritic-like Cells

A disintegrin and metalloproteinase (ADAM) 17 has been implicated in many shedding processes. Major substrates of ADAM17 are TNF-α, IL-6R, and ligands of the epidermal growth factor receptor. The essential role of the protease is emphasized by the fact that ADAM17 deficiency is lethal in mice. To study ADAM17 function in vivo, we generated viable hypomorphic ADAM17 mice called ADAM17 ex/ex mice. Recent studies indicated regulation of proteolytic ADAM17 activity by cellular processes such as cytoplasmic phosphorylation and removal of the prodomain by furin cleavage. Maturation and thus activation of ADAM17 is not fully understood. So far, studies of ADAM17 maturation have been mainly limited to mouse embryonic fibroblasts or transfected cell lines relying on nonphysiologic stimuli such as phorbol esters, thus making interpretation of the results difficult in a physiologic context. In this article, we present a robust cell system to study ADAM17 maturation and function in primary cells of the immune system. To this end, HoxB8 conditionally immortalized macrophage precursor cell lines were derived from bone marrow of wild-type and hypomorphic ADAM17 ex/ex mice, which are devoid of measurable ADAM17 activity. ADAM17 mutants were stably expressed in macrophage precursor cells, differentiated to macrophages under different growth factor conditions (M-CSF versus GM-CSF), and analyzed for cellular localization, proteolytic activity, and podosome disassembly. Our study reveals maturation and activity of ADAM17 in a more physiological-immune cell system. We show that this cell system can be further exploited for genetic modifications of ADAM17 and for studying its function in immune cells. Introduction As a member of the a disintegrin and metalloproteinase (ADAM) protease family, ADAM17 performs ectodomain shedding of diverse transmembrane proteins. ADAM17 has been first described as key protease involved in TNF-α shedding ( 1 , 2 ). Besides TNF-α, its receptors TNFR I and TNFR II , the IL-6R, and ligands of the epidermal growth factor receptor have been added to the long list of, to date, more than 80 ADAM17 substrates ( 3 ). The important role of ADAM17 in vivo is supported by the fact that deletion of the ADAM17 gene in mice is lethal ( 4 ). To study ADAM17 function in vivo, hypomorphic ADAM17 ex/ex mice were generated, which are viable and show ∼5% residual ADAM17 expression and no measurable shedding activity ( 5 ). Because ADAM17 substrates include membrane-bound cytokines [e.g., TNF-α, cytokine receptors, and the membrane-bound chemokines fractalkine and CXCL16 ( 6 )], ADAM17 turned out to be a key regulator during inflammation. Hence, genetic deletion of ADAM17 or pharmacologic blockade in neutrophils and leukocytes mediated resistance to LPS-induced endotoxemia and protected mice from otherwise lethal septic shock ( 7 , 8 ). Because ADAM17 processes the IL-6R, ADAM17 plays a decisive role in the IL-6 transsignaling pathway as part of the immune response ( 9 ). Despite the importance of ADAM17 in processing a large range of substrates, the regulatory mechanisms leading to ADAM17 activation and substrate recognition are not fully understood. Transgenic mice overexpressing ADAM17 show no enhancement in substrate cleavage ( 10 ), pointing toward strict regulation of protease activity by posttranslational mechanisms. ADAM17 is generated as a proenzyme, and the N-terminal propeptide can act as an autoinhibitor to keep the protease in an inactive state ( 11 , 12 ). The prodomain of ADAM17 is removed by furin-like convertases at two different sites ( 13 ): one located between the propeptide and the metalloprotease domain (referred to as downstream [ds] site) and the second cleavage site found within the prodomain (referred to as upstream [us] site), cleavage of which has been described as a prerequisite to cleavage at the ds site ( 13 ). Another posttranslational modification, which has been claimed to be important for ADAM17 activity, is phosphorylation of the cytoplasmic domain ( 14 – 18 ). Consequently, treatment of cells with phorbol ester (PMA) led to an increase in ADAM17 activity ( 19 , 20 ). Under physiologic conditions, phosphorylation of ADAM17 is mediated by MAPKs ( 16 ) and polo-like kinase 2 (PLK2) at serine 794 ( 15 ). However, the importance of phosphorylation of the cytoplasmic tail for ADAM17 activity is under debate because ADAM17-deficient cell lines show normal processing of substrates after reconstitution with ADAM17 variants, in which the entire cytoplasmic region was deleted ( 21 – 23 ). In more recent studies, it was shown that ADAM17 with a short-charged membrane-proximal stretch of 5 aa of the cytoplasmic portion together with a protein tag showed ADAM17-shedding activity ( 24 , 25 ). Because there is ongoing discrepancy in the field about the role and effect of the various C-terminal (CT) deletion mutants on ADAM17 function, a recent study uses ADAM17 variants devoid of intracellular amino acids, comparing a complete deletion of all 133 cytoplasmic amino acids (ADAM17 ΔCT) with an ADAM17 mutant still exhibiting a residual stretch of six charged amino acids of the cytoplasmic portion (ADAM17 Δ700) ( 26 ). Interestingly, the ADAM17 Δ700 variant showed normal activity, whereas the complete deletion of the cytoplasmic tail (ADAM17 ΔCT) resulted in a complete loss of shedding activity ( 26 ). This is the reason why in this study we exclusively focus on the function of the ADAM17 Δ700 variant. Studies addressing the physiologic relevance of posttranslational modifications of ADAM17 have been performed mainly in mouse embryonic fibroblasts (mEF) and in human embryonic kidney (HEK) cells with stimulation induced by the phorbol ester PMA ( 13 , 15 , 24 , 26 , 27 ). ADAM17 activation by the nonphysiologic stimulator PMA and analysis of mostly cell-extrinsic substrates do not provide a natural environment for studying ADAM17 function. For instance, physiologically relevant stimuli of immune cells include LPS and zymosan, which activate the TLR-4 and TLR-2 as well as the dectin-1 pathway, respectively ( 28 – 30 ). The main goal of the current study was to establish a robust cell model to study ADAM17 regulation and function under more physiologic conditions. Because ADAM17 is a key regulator of the immune system, we chose primary cells of the innate immune system for our experiments. Retroviral transduction of bone marrow–derived macrophages and bone marrow granulocytes from wild-type (wt) and ADAM17 ex/ex mice was not efficient enough to allow for studying ADAM17 variants. Hence, we applied an approach using an estrogen-dependent HoxB8 conditional immortalization protocol generating a macrophage precursor cell (MØP) line ( 31 , 32 ). This cell system allows stable generation of proliferating hematopoietic progenitor cells, which can be further differentiated by estrogen depletion and application of M-CSF and GM-CSF ( 32 ). The generated MØP as well as the differentiated mature macrophage-like cells (M-MØ) and dendritic-like cells (GM-MØ) respond to TLR stimulation by NF-κB–induced transcription of proinflammatory genes like TNF-α, IL-1β, and IL-6. Moreover, these cells can be easily transduced by retroviral vectors. In this study, we applied this cell system to study structural alterations of ADAM17, comparing the results to overexpression studies in mEF and HEK cells. ADAM17 ex/ex -derived HoxB8 progenitor cells were transduced with cDNAs coding for ADAM17 variants, including ADAM17 Δ700, ADAM17 S794A, and the two furin cleavage–resistant mutants ADAM17 RVNG us and RVNG ds. ADAM17-reconstituted M-MØ and GM-MØ were examined for expression, cellular localization, and biological activity of the ADAM17 variants. Our data provide evidence for a divergent regulation of ADAM17 protease activity in a physiologic cellular background in which we were able to analyze the role of ADAM17 in the disassembly of podosomes, which might be involved in Ag sampling and recognition ( 33 ). Materials and Methods cDNA constructs and cloning Expression plasmids of murine ADAM17 mutants (S794A, Δ700, RVNG ds, RVNG us) were cloned via site-directed mutagenesis PCR using the mADAM17 in the pcDNA3.1 (+) vector template as described ( 15 , 26 ). All ADAM17 variants were cloned into the retroviral pMXs-IZ vector ( 34 ) using the XhoI and NotI restriction site and were myc tagged at the C terminus. All plasmids were verified by DNA sequencing (GATC Biotech). Cell culture MØP were generated from bone marrow of ADAM17 wt/wt and ADAM17 ex/ex mice as previously described ( 32 , 34 ). Briefly, CD117-enriched bone marrow cells (using CD117-biotin [BD Biosciences] and anti-biotin-MACS [Miltenyi Biotec]) were infected with an estrogen-dependent Hoxb8 pMXs retroviral vector for conditional immortalization and cultured in RPMI 1640 medium (Thermo Fisher Scientific) containing 10% heat-inactivated FCS (PAA Laboratories), 1% penicillin/streptomycin (Pen/Strep), β-estradiol (1 μM) (Sigma-Aldrich), GM-CSF (10 ng/ml) (ImmunoTools), and puromycin (20 μM/ml) (Roth). After selection in puromycin, the MØP, ADAM17-deficient cells were further transduced with retroviral vectors coding ADAM17 variants (in pMXs-IZ vector) and incubated with zeocin (100 μg/ml) for selection of stably transfected cells. ADAM17-reconstituted MØP were further differentiated to M-MØ and GM-MØ by application of M-CSF and GM-CSF, respectively, in the absence of estrogen (see below and Fig. 2A ). mEF were isolated from E13.5 ADAM17 wt/wt and ADAM17 ex/ex embryos and immortalized with the SV40 large T Ag as previously described ( 5 ). HEK293T cells were purchased from DSMZ (Braunschweig, Germany), and ADAM10- and ADAM17-deficient HEK cells (HEK double knockout [dKO]) were generated via CRISPR/Cas9 system as described ( 35 ). mEF and HEK cells were cultured in DMEM containing 10% heat-inactivated FCS and 1% Pen/Strep. Transfection and retroviral transduction HEK cells were transiently transfected with polyethylenimine (PEI) in a 1:3 DNA to PEI ratio and mEF by using X-tremeGENE HP DNA Transfection Reagent (Roche) according to the manufacturer's instructions. Retroviruses were generated in EcoPacks (Clontech Laboratories) and transfected by using TurboFect (Thermo Scientific). The supernatant of the transfected EcoPacks was concentrated with Retro-X Concentrator (Clontech Laboratories). The concentrated retrovirus was used to infect ADAM17 ex/ex MØP with the ADAM17 variants. The cells were incubated with the virus and polybrene (8 μg/ml) for 3 d and selected for stable transduction by zeocin. MØP differentiation For the differentiation in GM-MØ, the MØP were washed three times with RPMI 1640 containing 10% heat-inactivated FCS and 1% Pen/Strep and seeded at a density of 250,000 cells per well in a six-well plate in RPMI 1640 additionally supplemented with 10 ng/ml GM-CSF. Medium was changed every 2 d. On the seventh day of differentiation, the GM-MØ were stimulated with 1 μg/ml LPS ( Escherichia coli O111:B4; Sigma-Aldrich) or zymosan (25 μg/ml) for 12 h. For M-MØ differentiation, three wash steps with RPMI 1640 containing 10% heat-inactivated FCS and 1× Pen/Strep were applied before seeding at a density of 350,000 cells per well on a six-well plate in RPMI 1640 additionally supplemented with 20 ng/ml M-CSF. The medium was changed every 2 d. The stimulation with LPS (1 μg/ml) or zymosan (25 μg/ml) was done on day 5 of differentiation. Cells and the supernatant were harvested on the next day. Bright light microscopy pictures were taken during the differentiation with a compact inverted CKX41 microscope (Olympus) equipped with a UPLFLN objective (numerical aperture: 0.3). ELISA A total of 250,000 undifferentiated MØP were seeded onto a 12-well plate and stimulated 6 h later with LPS (1 μg/ml) or zymosan (25 μg/ml) overnight. The differentiated MØP were stimulated 12 h before harvesting. mEF cells were seeded at a density of 200,000 cells per well on a six-well plate. After 24 h, cells were transfected as described above. On the next day, mEF were stimulated with PMA (200 nM) for 2 h. In the cell-free supernatant IL-6, TNF-α and soluble TNFR II were measured by ELISA, according to the manufacturer's instructions. The TNFR II ELISA kit was from R&D Systems. The ELISA kits for IL-6 and TNF-α were from eBioscience. Life cell activity assay A total of 2 × 10 6 ADAM10- and ADAM17-deficient HEK cells (HEK dKO) were seeded onto a 10-cm dish and transfected the next day with PEI (1:3 DNA/PEI ratio). The following day, cells were detached from cell culture dish by trypsin and seeded onto a 96-well plate (2 × 10 5 cells per well). The next day, the medium was replaced by phenolred-free medium, and quenched fluorogenic peptide (Abz-LAQAVRSSSR-Dpa; TACE-substrate IV [no. 616407; Calbiochem]) was added to a concentration of 20 μM. Cell surface activity of ADAM17 was measured in a plate reader (Tecan Infinite 200 Pro) for 6000 s at emission 405 nm and extinction 320 nm. ADAM17 cell surface activity was determined by calculating the area under curve of the fluorescent signal over time. Flow cytometry MØP suspension cells were counted, 250,000 cells were collected, and the differentiated MØP were mechanically detached. All cells were washed three times with ice-cold PBS and blocked for 10 min in PBS containing 1% BSA and 1% Fc block (blocking buffer [BioLegend]) and incubated with the fluorescence-coupled Abs diluted 1:100 in blocking buffer for 1 h at 4°C. Afterwards, the cell suspension was washed two times with ice-cold PBS containing 1% BSA. The staining was fixed with 10% lysis/fixation solution (BioLegend) in distilled water for 15 min at room temperature. Before performing flow cytometry analysis (FACS Canto II; BD Bioscience), the cells were washed again two times with ice-cold PBS containing 1% BSA. A total of 30,000 events were gated using the forward scatter area (FSC-A) and forward scatter height (FSC-H) and only CD45 + cells were used for further analyses. All used markers were compensated before the flow cytometry analysis. All data analysis was performed using the flow cytometry analysis software FlowJo (Tree Star). The following Abs were used: brilliant violet 510 anti-mouse CD45 (clone 30-F11; Biolegend), brilliant violet 421 anti-mouse CD115 (clone AFS98; Biolegend), PE anti-mouse CD11b (clone M1/70; BD Biosciences), APC anti-mouse F4/80 (clone BM8; Biolegend), FITC anti-mouse CD11c (clone N418; Biolegend), PE/Cy7 anti-mouse CD206 (clone C068C2; Biolegend), and APC/Fire anti-mouse MHCII (clone M5/114.15.2; Biolegend). For ADAM17 cell surface staining, HEK cells were transfected as described above, harvested, and blocked in 1% BSA in PBS for 10 min at 4°C. The MØP were harvested via centrifugation (1000 × g , 5 min, 4°C), washed with PBS, and blocked (1% BSA and 1% Fc block, 10 min, 4°C). Afterwards, the cell suspension was stained with the primary Ab (1:100 diluted in 1% BSA in PBS) for 1 h at 4°C, washed with 1% BSA in PBS, and incubated with the secondary Ab (1:100 in 1% BSA in PBS) for 1 h at 4°C. Before flow cytometry analysis, the cells were washed and resuspended in 1% BSA and PBS. To stain dead cells, the MØP suspension was incubated for 5 min with 25 ng of 7-AAD (no. 420403; Biolegend) before flow cytometry analysis. The following Abs were used: primary Ab anti-ADAM17 10.1 (polyclonal Ab against the extracellular portion of murine ADAM17) and secondary Ab donkey anti-rabbit IgG Alexa Fluor 488 (no. A21206; Thermo Fisher). Immunofluorescence analysis A total of 250,000 MØP per well were seeded in a six-well plate on glass cover slips. After differentiation, cells were stimulated with LPS (1 μg/ml) or zymosan (25 μg/ml) for 12 h. HEK cells were seeded at a density of 200,000 cells per well on a six-well plate and transfected as described above on the following day. Twenty-four hours after transfection, the cells were stimulated with PMA (200 nM) for 2 h. Cells were washed with PBS, fixed with 4% PFA in PBS for 15 min at room temperature, and permeabilized with 0.3% Triton X-100 (Sigma-Aldrich) in PBS for 30 min. Cells were blocked in blocking buffer (2% BSA, 5% heat-inactivated FCS, and 0.3% Triton in PBS) for 60 min. The primary Abs were diluted in blocking buffer and incubated at 4°C overnight. Cells were washed three times with PBS containing 0.3% Triton X-100 and incubated with secondary Abs diluted 1:500 in blocking buffer. Cells were washed three times with 0.3% Triton X-100 in PBS, one time with PBS, and stained with DAPI as part of the mounting mix consisting of DABCO (Sigma-Aldrich) and MOWIOL (Merck Millipore). Stained cells were viewed and photographed with an FV1000 confocal laser scanning microscope (Olympus) equipped with a U Plan S Apo 100× oil immersion objective (numerical aperture: 1.40). Digital images were processed using FV10-ASW 4.2 Viewer (Olympus). The Pearson index was used to express colocalization of two stainings, which was calculated by an integrated tool within the FV10-ASW software. Transfected HEK cells were marked by a region of interest, and colocalization of two stainings was determined by the software, resulting in values between 0 and 1. The value 1 indicates a 100% colocalization of both markers ( 36 ). Primary Abs used are as follows: anti-PDIA6 (1:750, ab11432; Abcam), anti-myc (1:250, clone 9B11 no. 2276; Cell Signal), KDEL (1:100, clone 10C3; Enzo), anti-human ADAM17 [1:50, A300D, in-house production ( 37 )], anti-ADAM17 10.1 (1:100), and anti-ADAM17 18.2 (1:100) (polyclonal Abs generated by Pineda Antikörper-Service, Berlin, Germany). Secondary Abs were purchased from Thermo Scientific: goat-anti-mouse Alexa Fluor 488, goat-anti-rabbit Alexa Fluor 488, goat-anti-mouse Alexa Fluor 594, and goat-anti-rabbit Alexa Fluor 594. For staining of ADAM17 in M-MØ and GM-MØ, the anti-myc Ab was used because stably transfected ADAM17 constructs were myc tagged at the C terminus. Western blot analysis Differentiated MØP, mEF, and HEK cells were mechanically detached from the culture dish. MØP suspension cells were pelleted by centrifugation (1000 × g , 5 min, 4°C). All cells were washed with ice-cold PBS and lysed in lysis buffer containing 10 mM 1.10-phenanthroline (Merck Millipore). The MØP lysis buffer contained 1% Triton X-100, 150 mM NaCl, 50 mM Tris-HCL (pH 7.4), and two tablets of complete protease mixture inhibitor (Roche) in 50 ml. The lysis buffer used for the mEF and HEK cells additionally contained 1% IGEPAL (Nonidet P-40) and 2 mM EDTA. The amount of total protein was determined by BCA assay (Thermo Scientific). Aliquots of the lysates were supplemented with 5× SDS sample buffer (0.3 M Tris-HCl, pH 6.8, 10% SDS, 50% glycerol, 20% mercaptoethanol, 5% bromophenol blue) and heat inactivated at 95°C for 5 min. Equal amounts (30–40 μg total protein) were separated by SDS-PAGE and transferred to a PVDF membrane (Merck Millipore). The following Abs were used for detection: anti-ADAM17 10.1 (polyclonal Ab against the extracellular portion of murine ADAM17 [peptide: CEVKPGRHFNMAKSFPNEEK]), anti-ADAM17 18.2 (polyclonal Ab against the intracellular portion of murine ADAM17 [peptide: RLQALQPAAMMPPVSAAPKL; both Pineda Antikörper-Service]), anti-ADAM17 (polyclonal Ab against intracellular domain of ADAM17, ab39162; Abcam), anti-human ADAM17 [A300D, in-house production ( 37 )], anti-iRhom2 (RHBDF2) (ab116139; Abcam), anti-TNFR II (ab221921; Abcam), anti-pro–TNF-α [in-house production, kindly provided by B. Schröder, Technische Universität Dresden, Dresden, Germany ( 26 )], anti-β-actin (A5441; Sigma-Aldrich), and anti-ADAM10 (ab124695; Abcam). As secondary Abs, goat anti-rabbit HRP and goat anti-mouse HRP (both Dianova) were used. For protein enrichment by Con A (Sigma-Aldrich) precipitation, 1000 μl of cell lysate was incubated with 30-μl Con A beads overnight at 4°C under constant agitation. Samples were spun down for 15 min at 4000 × g and washed three times with PBS. The pellet was incubated with 100 μl of PBS and 100 μl of 5 × loading buffer (see above), heated to 95°C for 10 min, and analyzed on SDS-PAGE. Immunoprecipitation and furin assay Cells were lysed, and protein was quantified as described above. Afterwards, 1 mg of total protein was immunoprecipitated with 3 μg of anti-ADAM17 Ab (no. ab39162; Abcam) overnight at 4°C. The used Protein G Dynabeads (no. 10004D; Thermo Fisher Scientific) were blocked with 1% BSA in PBS overnight at 4°C. The blocked beads were added to the lysates and incubated for 1 h at 4°C. After three washing steps with lysis buffer, 1 × SDS sample buffer was added to the beads and heated to 65°C for 15 min. For the furin cleavage assay, the beads were incubated in furin assay buffer (100 mM HEPES, 1 mM CaCl, 1 mM 2-ME, 0.5% Triton X-100) with 1 U recombinant furin (no. P8077S; New England Biolabs) for 30 min at 30°C. Afterwards, 5 × SDS sample buffer was added and boiled. The immunoprecipitated proteins were further analyzed by Western blotting. Electron microscopy For scanning electron microscopy (SEM), cells were seeded on glass slides and differentiated to the desired cell type. Subsequently, cells were fixed in 3% glutaraldehyde in PBS for 30 min prior to three times washing in PBS. Prior to ascending ethanol steps (30, 40, 50, 70, 96, 100%), cells were fixed in 2% osmium tetroxide for 30 min. After critical point drying and sputtering with gold, samples were imaged on a Philips XL20 SEM. Analysis of podosomes MØP were differentiated by M-CSF for 6 d and stimulated with LPS for 24 h as described above. Subsequently, cells were fixed with 3.7% formaldehyde, permeabilized with 0.5% Triton X-100 and PBS (pH 7.5) for 10 min and stained for podosome markers F-actin (Alexa-Fluor-488– or Alexa-Fluor-568–phalloidin; Molecular Probes, Eugene, OR) and vinculin (mAb V9264, 1:500; Sigma-Aldrich) to highlight podosome rings. CellMask Deep Red plasma membrane stain (no. C10046; Thermo Fisher Scientific) was used to stain the cytoplasm and nuclei at a concentration of 2 μg/ml. Images of fixed samples were acquired with a confocal laser-scanning microscope (DMI 6000 with a TCS SP5 AOBS confocal point scanner; Leica) equipped with an oil-immersion HCX PL APO 63× numerical aperture 1.4–0.6 objective. Podosome number and cell area analysis were performed using an ImageJ (National Institutes of Health, Bethesda, MD) macro ( 38 ). The analysis was performed with three different donors with 30 cells for each condition ( n = 3 × 30). Protein modeling Protein modeling was performed with Swiss Model Expasy ( http://swissmodel.expasy.org/ ) and the Chimera package (University of California; www.cgl.ucsf.edu/chimera ). Data analysis and statistics All values are expressed as the mean ± SEM. For data analysis, Excel (Microsoft) and GraphPad Prism version 6.00 for Mac (GraphPad Software) were used. Differences among mean values were analyzed as indicated in the figure legends by two-tailed, unpaired Student t test or one-way ANOVA, followed by a Tukey multiple comparison test using GraphPad Prism6 software for multiple samples, when applicable. In all analyses, the null hypothesis was rejected at p < 0.05, with * p < 0.05, ** p < 0.01, *** p < 0.005, and **** p < 0.001. cDNA constructs and cloning Expression plasmids of murine ADAM17 mutants (S794A, Δ700, RVNG ds, RVNG us) were cloned via site-directed mutagenesis PCR using the mADAM17 in the pcDNA3.1 (+) vector template as described ( 15 , 26 ). All ADAM17 variants were cloned into the retroviral pMXs-IZ vector ( 34 ) using the XhoI and NotI restriction site and were myc tagged at the C terminus. All plasmids were verified by DNA sequencing (GATC Biotech). Cell culture MØP were generated from bone marrow of ADAM17 wt/wt and ADAM17 ex/ex mice as previously described ( 32 , 34 ). Briefly, CD117-enriched bone marrow cells (using CD117-biotin [BD Biosciences] and anti-biotin-MACS [Miltenyi Biotec]) were infected with an estrogen-dependent Hoxb8 pMXs retroviral vector for conditional immortalization and cultured in RPMI 1640 medium (Thermo Fisher Scientific) containing 10% heat-inactivated FCS (PAA Laboratories), 1% penicillin/streptomycin (Pen/Strep), β-estradiol (1 μM) (Sigma-Aldrich), GM-CSF (10 ng/ml) (ImmunoTools), and puromycin (20 μM/ml) (Roth). After selection in puromycin, the MØP, ADAM17-deficient cells were further transduced with retroviral vectors coding ADAM17 variants (in pMXs-IZ vector) and incubated with zeocin (100 μg/ml) for selection of stably transfected cells. ADAM17-reconstituted MØP were further differentiated to M-MØ and GM-MØ by application of M-CSF and GM-CSF, respectively, in the absence of estrogen (see below and Fig. 2A ). mEF were isolated from E13.5 ADAM17 wt/wt and ADAM17 ex/ex embryos and immortalized with the SV40 large T Ag as previously described ( 5 ). HEK293T cells were purchased from DSMZ (Braunschweig, Germany), and ADAM10- and ADAM17-deficient HEK cells (HEK double knockout [dKO]) were generated via CRISPR/Cas9 system as described ( 35 ). mEF and HEK cells were cultured in DMEM containing 10% heat-inactivated FCS and 1% Pen/Strep. Transfection and retroviral transduction HEK cells were transiently transfected with polyethylenimine (PEI) in a 1:3 DNA to PEI ratio and mEF by using X-tremeGENE HP DNA Transfection Reagent (Roche) according to the manufacturer's instructions. Retroviruses were generated in EcoPacks (Clontech Laboratories) and transfected by using TurboFect (Thermo Scientific). The supernatant of the transfected EcoPacks was concentrated with Retro-X Concentrator (Clontech Laboratories). The concentrated retrovirus was used to infect ADAM17 ex/ex MØP with the ADAM17 variants. The cells were incubated with the virus and polybrene (8 μg/ml) for 3 d and selected for stable transduction by zeocin. MØP differentiation For the differentiation in GM-MØ, the MØP were washed three times with RPMI 1640 containing 10% heat-inactivated FCS and 1% Pen/Strep and seeded at a density of 250,000 cells per well in a six-well plate in RPMI 1640 additionally supplemented with 10 ng/ml GM-CSF. Medium was changed every 2 d. On the seventh day of differentiation, the GM-MØ were stimulated with 1 μg/ml LPS ( Escherichia coli O111:B4; Sigma-Aldrich) or zymosan (25 μg/ml) for 12 h. For M-MØ differentiation, three wash steps with RPMI 1640 containing 10% heat-inactivated FCS and 1× Pen/Strep were applied before seeding at a density of 350,000 cells per well on a six-well plate in RPMI 1640 additionally supplemented with 20 ng/ml M-CSF. The medium was changed every 2 d. The stimulation with LPS (1 μg/ml) or zymosan (25 μg/ml) was done on day 5 of differentiation. Cells and the supernatant were harvested on the next day. Bright light microscopy pictures were taken during the differentiation with a compact inverted CKX41 microscope (Olympus) equipped with a UPLFLN objective (numerical aperture: 0.3). ELISA A total of 250,000 undifferentiated MØP were seeded onto a 12-well plate and stimulated 6 h later with LPS (1 μg/ml) or zymosan (25 μg/ml) overnight. The differentiated MØP were stimulated 12 h before harvesting. mEF cells were seeded at a density of 200,000 cells per well on a six-well plate. After 24 h, cells were transfected as described above. On the next day, mEF were stimulated with PMA (200 nM) for 2 h. In the cell-free supernatant IL-6, TNF-α and soluble TNFR II were measured by ELISA, according to the manufacturer's instructions. The TNFR II ELISA kit was from R&D Systems. The ELISA kits for IL-6 and TNF-α were from eBioscience. Life cell activity assay A total of 2 × 10 6 ADAM10- and ADAM17-deficient HEK cells (HEK dKO) were seeded onto a 10-cm dish and transfected the next day with PEI (1:3 DNA/PEI ratio). The following day, cells were detached from cell culture dish by trypsin and seeded onto a 96-well plate (2 × 10 5 cells per well). The next day, the medium was replaced by phenolred-free medium, and quenched fluorogenic peptide (Abz-LAQAVRSSSR-Dpa; TACE-substrate IV [no. 616407; Calbiochem]) was added to a concentration of 20 μM. Cell surface activity of ADAM17 was measured in a plate reader (Tecan Infinite 200 Pro) for 6000 s at emission 405 nm and extinction 320 nm. ADAM17 cell surface activity was determined by calculating the area under curve of the fluorescent signal over time. Flow cytometry MØP suspension cells were counted, 250,000 cells were collected, and the differentiated MØP were mechanically detached. All cells were washed three times with ice-cold PBS and blocked for 10 min in PBS containing 1% BSA and 1% Fc block (blocking buffer [BioLegend]) and incubated with the fluorescence-coupled Abs diluted 1:100 in blocking buffer for 1 h at 4°C. Afterwards, the cell suspension was washed two times with ice-cold PBS containing 1% BSA. The staining was fixed with 10% lysis/fixation solution (BioLegend) in distilled water for 15 min at room temperature. Before performing flow cytometry analysis (FACS Canto II; BD Bioscience), the cells were washed again two times with ice-cold PBS containing 1% BSA. A total of 30,000 events were gated using the forward scatter area (FSC-A) and forward scatter height (FSC-H) and only CD45 + cells were used for further analyses. All used markers were compensated before the flow cytometry analysis. All data analysis was performed using the flow cytometry analysis software FlowJo (Tree Star). The following Abs were used: brilliant violet 510 anti-mouse CD45 (clone 30-F11; Biolegend), brilliant violet 421 anti-mouse CD115 (clone AFS98; Biolegend), PE anti-mouse CD11b (clone M1/70; BD Biosciences), APC anti-mouse F4/80 (clone BM8; Biolegend), FITC anti-mouse CD11c (clone N418; Biolegend), PE/Cy7 anti-mouse CD206 (clone C068C2; Biolegend), and APC/Fire anti-mouse MHCII (clone M5/114.15.2; Biolegend). For ADAM17 cell surface staining, HEK cells were transfected as described above, harvested, and blocked in 1% BSA in PBS for 10 min at 4°C. The MØP were harvested via centrifugation (1000 × g , 5 min, 4°C), washed with PBS, and blocked (1% BSA and 1% Fc block, 10 min, 4°C). Afterwards, the cell suspension was stained with the primary Ab (1:100 diluted in 1% BSA in PBS) for 1 h at 4°C, washed with 1% BSA in PBS, and incubated with the secondary Ab (1:100 in 1% BSA in PBS) for 1 h at 4°C. Before flow cytometry analysis, the cells were washed and resuspended in 1% BSA and PBS. To stain dead cells, the MØP suspension was incubated for 5 min with 25 ng of 7-AAD (no. 420403; Biolegend) before flow cytometry analysis. The following Abs were used: primary Ab anti-ADAM17 10.1 (polyclonal Ab against the extracellular portion of murine ADAM17) and secondary Ab donkey anti-rabbit IgG Alexa Fluor 488 (no. A21206; Thermo Fisher). Immunofluorescence analysis A total of 250,000 MØP per well were seeded in a six-well plate on glass cover slips. After differentiation, cells were stimulated with LPS (1 μg/ml) or zymosan (25 μg/ml) for 12 h. HEK cells were seeded at a density of 200,000 cells per well on a six-well plate and transfected as described above on the following day. Twenty-four hours after transfection, the cells were stimulated with PMA (200 nM) for 2 h. Cells were washed with PBS, fixed with 4% PFA in PBS for 15 min at room temperature, and permeabilized with 0.3% Triton X-100 (Sigma-Aldrich) in PBS for 30 min. Cells were blocked in blocking buffer (2% BSA, 5% heat-inactivated FCS, and 0.3% Triton in PBS) for 60 min. The primary Abs were diluted in blocking buffer and incubated at 4°C overnight. Cells were washed three times with PBS containing 0.3% Triton X-100 and incubated with secondary Abs diluted 1:500 in blocking buffer. Cells were washed three times with 0.3% Triton X-100 in PBS, one time with PBS, and stained with DAPI as part of the mounting mix consisting of DABCO (Sigma-Aldrich) and MOWIOL (Merck Millipore). Stained cells were viewed and photographed with an FV1000 confocal laser scanning microscope (Olympus) equipped with a U Plan S Apo 100× oil immersion objective (numerical aperture: 1.40). Digital images were processed using FV10-ASW 4.2 Viewer (Olympus). The Pearson index was used to express colocalization of two stainings, which was calculated by an integrated tool within the FV10-ASW software. Transfected HEK cells were marked by a region of interest, and colocalization of two stainings was determined by the software, resulting in values between 0 and 1. The value 1 indicates a 100% colocalization of both markers ( 36 ). Primary Abs used are as follows: anti-PDIA6 (1:750, ab11432; Abcam), anti-myc (1:250, clone 9B11 no. 2276; Cell Signal), KDEL (1:100, clone 10C3; Enzo), anti-human ADAM17 [1:50, A300D, in-house production ( 37 )], anti-ADAM17 10.1 (1:100), and anti-ADAM17 18.2 (1:100) (polyclonal Abs generated by Pineda Antikörper-Service, Berlin, Germany). Secondary Abs were purchased from Thermo Scientific: goat-anti-mouse Alexa Fluor 488, goat-anti-rabbit Alexa Fluor 488, goat-anti-mouse Alexa Fluor 594, and goat-anti-rabbit Alexa Fluor 594. For staining of ADAM17 in M-MØ and GM-MØ, the anti-myc Ab was used because stably transfected ADAM17 constructs were myc tagged at the C terminus. Western blot analysis Differentiated MØP, mEF, and HEK cells were mechanically detached from the culture dish. MØP suspension cells were pelleted by centrifugation (1000 × g , 5 min, 4°C). All cells were washed with ice-cold PBS and lysed in lysis buffer containing 10 mM 1.10-phenanthroline (Merck Millipore). The MØP lysis buffer contained 1% Triton X-100, 150 mM NaCl, 50 mM Tris-HCL (pH 7.4), and two tablets of complete protease mixture inhibitor (Roche) in 50 ml. The lysis buffer used for the mEF and HEK cells additionally contained 1% IGEPAL (Nonidet P-40) and 2 mM EDTA. The amount of total protein was determined by BCA assay (Thermo Scientific). Aliquots of the lysates were supplemented with 5× SDS sample buffer (0.3 M Tris-HCl, pH 6.8, 10% SDS, 50% glycerol, 20% mercaptoethanol, 5% bromophenol blue) and heat inactivated at 95°C for 5 min. Equal amounts (30–40 μg total protein) were separated by SDS-PAGE and transferred to a PVDF membrane (Merck Millipore). The following Abs were used for detection: anti-ADAM17 10.1 (polyclonal Ab against the extracellular portion of murine ADAM17 [peptide: CEVKPGRHFNMAKSFPNEEK]), anti-ADAM17 18.2 (polyclonal Ab against the intracellular portion of murine ADAM17 [peptide: RLQALQPAAMMPPVSAAPKL; both Pineda Antikörper-Service]), anti-ADAM17 (polyclonal Ab against intracellular domain of ADAM17, ab39162; Abcam), anti-human ADAM17 [A300D, in-house production ( 37 )], anti-iRhom2 (RHBDF2) (ab116139; Abcam), anti-TNFR II (ab221921; Abcam), anti-pro–TNF-α [in-house production, kindly provided by B. Schröder, Technische Universität Dresden, Dresden, Germany ( 26 )], anti-β-actin (A5441; Sigma-Aldrich), and anti-ADAM10 (ab124695; Abcam). As secondary Abs, goat anti-rabbit HRP and goat anti-mouse HRP (both Dianova) were used. For protein enrichment by Con A (Sigma-Aldrich) precipitation, 1000 μl of cell lysate was incubated with 30-μl Con A beads overnight at 4°C under constant agitation. Samples were spun down for 15 min at 4000 × g and washed three times with PBS. The pellet was incubated with 100 μl of PBS and 100 μl of 5 × loading buffer (see above), heated to 95°C for 10 min, and analyzed on SDS-PAGE. Immunoprecipitation and furin assay Cells were lysed, and protein was quantified as described above. Afterwards, 1 mg of total protein was immunoprecipitated with 3 μg of anti-ADAM17 Ab (no. ab39162; Abcam) overnight at 4°C. The used Protein G Dynabeads (no. 10004D; Thermo Fisher Scientific) were blocked with 1% BSA in PBS overnight at 4°C. The blocked beads were added to the lysates and incubated for 1 h at 4°C. After three washing steps with lysis buffer, 1 × SDS sample buffer was added to the beads and heated to 65°C for 15 min. For the furin cleavage assay, the beads were incubated in furin assay buffer (100 mM HEPES, 1 mM CaCl, 1 mM 2-ME, 0.5% Triton X-100) with 1 U recombinant furin (no. P8077S; New England Biolabs) for 30 min at 30°C. Afterwards, 5 × SDS sample buffer was added and boiled. The immunoprecipitated proteins were further analyzed by Western blotting. Electron microscopy For scanning electron microscopy (SEM), cells were seeded on glass slides and differentiated to the desired cell type. Subsequently, cells were fixed in 3% glutaraldehyde in PBS for 30 min prior to three times washing in PBS. Prior to ascending ethanol steps (30, 40, 50, 70, 96, 100%), cells were fixed in 2% osmium tetroxide for 30 min. After critical point drying and sputtering with gold, samples were imaged on a Philips XL20 SEM. Analysis of podosomes MØP were differentiated by M-CSF for 6 d and stimulated with LPS for 24 h as described above. Subsequently, cells were fixed with 3.7% formaldehyde, permeabilized with 0.5% Triton X-100 and PBS (pH 7.5) for 10 min and stained for podosome markers F-actin (Alexa-Fluor-488– or Alexa-Fluor-568–phalloidin; Molecular Probes, Eugene, OR) and vinculin (mAb V9264, 1:500; Sigma-Aldrich) to highlight podosome rings. CellMask Deep Red plasma membrane stain (no. C10046; Thermo Fisher Scientific) was used to stain the cytoplasm and nuclei at a concentration of 2 μg/ml. Images of fixed samples were acquired with a confocal laser-scanning microscope (DMI 6000 with a TCS SP5 AOBS confocal point scanner; Leica) equipped with an oil-immersion HCX PL APO 63× numerical aperture 1.4–0.6 objective. Podosome number and cell area analysis were performed using an ImageJ (National Institutes of Health, Bethesda, MD) macro ( 38 ). The analysis was performed with three different donors with 30 cells for each condition ( n = 3 × 30). Protein modeling Protein modeling was performed with Swiss Model Expasy ( http://swissmodel.expasy.org/ ) and the Chimera package (University of California; www.cgl.ucsf.edu/chimera ). Data analysis and statistics All values are expressed as the mean ± SEM. For data analysis, Excel (Microsoft) and GraphPad Prism version 6.00 for Mac (GraphPad Software) were used. Differences among mean values were analyzed as indicated in the figure legends by two-tailed, unpaired Student t test or one-way ANOVA, followed by a Tukey multiple comparison test using GraphPad Prism6 software for multiple samples, when applicable. In all analyses, the null hypothesis was rejected at p < 0.05, with * p < 0.05, ** p < 0.01, *** p < 0.005, and **** p < 0.001. Results Functional and structural characterization of ADAM17 mutants in mEF and HEK cells We generated four ADAM17 mutants by introducing point mutations and a C-terminal truncation after residue 699, leaving a cytoplasmic rest of 6 aa (ADAM17 Δ700). Two furin-resistant mutants were generated by introduction of an RVNG motif in either the ds ( RVKR ) or us ( RKRD ) furin cleavage site. Furthermore, we mutated the phosphorylation site of ADAM17 by replacing Ser 794 with Ala ( Fig. 1A ). To study posttranslational processes of ADAM17, mutants were analyzed for expression, cellular localization, and proteolytic activity. Initial experiments were performed in mEF and HEK cells on an ADAM17-deficient background. mEF were obtained from hypomorphic ADAM17 ex/ex animals and transiently transfected with the respective ADAM17 mutants. Equal protein expression of all ADAM17 variants was verified by SDS-PAGE and Western blotting using C- and N-terminal ADAM17 Abs under unstimulated (−PMA) and stimulated (+PMA) conditions ( Fig. 1B ). It should be noted that the cytoplasmic deletion mutant (Δ700) can only be detected by the N-terminal ADAM17 Ab ( Fig. 1B ). Epitopes of used Abs are visualized on a schematic ADAM17 protein ( Fig. 1C ). For functional analysis, ADAM17-transfected mEF were stimulated by PMA. Cell supernatants were analyzed for endogenous TNFR II shedding by ELISA, showing an increase in soluble TNFR II levels, indicating enhanced ADAM17 activity in ADAM17-reconstituted mEF in comparison to mock-transfected cells ( Fig. 1D ). Mutants ADAM17 RVNG ds, S794A, and Δ700 exhibited higher TNFR II -shedding activities than mock transfectants, and only the RVNG us mutant showed no protease activity after PMA stimulation ( Fig. 1D ). No changes in TNFR II protein expression were observed under the experimental conditions used ( Fig. 1E , Supplemental Fig. 1A ). FIGURE 1. Functional and structural analyses of ADAM17 mutants in mEF and HEK cells. ( A ) Schematic overview of the 827-aa-long murine ADAM17 protein subdivided in different domains and ADAM17 mutants analyzed in this study. Mutations introduced in the prodomain to generate the RVNG ds and RVNG us mutants are indicated by red letters. The location of the S794A point mutation is indicated by a black cross, and the ADAM17 Δ700 mutant is truncated after aa 699. ( B ) Immunoblot of endogenous ADAM17 in wt mEF and transiently overexpressed ADAM17 in mEF exhibiting an ADAM17 knockdown (ADAM17 ex/ex ). Cells were nonstimulated (−PMA) or stimulated with PMA for 2 h. C-terminal (18.2) as well as N-terminal (10.1) Abs were used to detect ADAM17, indicating uniform expression of all ADAM17 mutants. The Δ700 can only be detected by the N-terminal Ab. β-Actin was used as loading control. ( C ) Schematic model of ADAM17 protein indicating binding sites of used Abs. The polyclonal N-terminal ADAM17 Ab 10.1 was designed against a peptide (aa 290–309) in the catalytic domain (CD). Both C-terminal Abs (ab39162 and house-made 18.2; Abcam) are polyclonal and were designed against a cytoplasmic epitope. The 18.2 Ab was designed to bind a peptide (aa 738–757) within the intracellular domain (ICD). ( D ) ELISA of endogenous TNFR II in supernatants of ADAM17 ex/ex mEF reconstituted with each respective ADAM17 mutant, showing constitutive shedding activity (black bars) and TNFR II shedding after a 2-h PMA stimulation (gray bars). The dotted line marks baseline TNFR II shedding ( n = 3 from three individual transfections). ( E ) Endogenous TNFR II protein level in ADAM17 ex/ex mEF after mock transfection and reconstitution with ADAM17 wt with and without a 2-h PMA stimulation. Signal intensity analysis of TNFR II normalized to loading control (β-actin) indicates unchanged TNFR II level. ( F ) Representative immunoblot of ADAM17 mutants transiently overexpressed in ADAM10/ADAM17 double-deficient HEK cells (HEK A10/A17 dKO) detected by a C-terminal (18.2) as well as N-terminal Ab (10.1) showing equal ADAM17 protein level. β-Actin was used as loading control. ( G ) Cell surface FACS analysis of ADAM17 overexpressed in HEK A10/A17 dKO cells stained with extracellular N-terminal Ab 10.1. Single cells (FSC-A × FSC-H) were gated and plotted for secondary Ab signal (Alexa Fluor 488). Bar graph shows median fluorescence intensity after normalization to mock transfection ( n = 3 from three individual transfections). ( H ) ADAM17 life cell surface activity assay in ADAM17-reconstituted HEK A10/17 dKO cells by fluorogenic peptide cleavage assay. Protease activity was determined by measuring fluorescence intensity of the cleaved peptide on living cells for 6000 s. The area under the curve of fluorescent signal over time was calculated and normalized to the mock-transfected control ( n = 3 from three individual transfections). ( I ) Representative Western blot of three individual experiments (anti C-terminal Ab [Abcam]) of a furin cleavage assay. Lysates of ADAM17-transfected HEK A10/A17 dKO cells were immunoprecipitated by a C-terminal Ab (Abcam) and incubated with 1 U of recombinant furin. Upper ADAM17 band (∼120 kDa) indicates full-length and nonfurin-processed ADAM17 form. After furin cleavage of the prodomain, the mature ADAM17 migrates at ∼90 kDa (black arrow). ( J ) Protein model of ADAM17 highlighting catalytic and prodomain (gray). In orange, the us and ds furin cleavage sites are indicated. The prodomain of ADAM17 and its orientation toward the catalytic domain were modeled on the basis of the previously published adamalysin crystal structure [right; PDB:3P24 ( 39 )] using Swiss Model Expasy. The red highlighted Asp194 in adamalysin is analog to the Asp190 in the prodomain of ADAM17. Both aspartates interact with a zinc ion (orange dot). Molecular imaging was performed using UCSF Chimera. For statistical analyses, a two-sided Student t test was used in (E), exhibiting no statistically significant differences. A one-way ANOVA followed by a Tukey multiple comparison test was applied in (D), (G), and (H). CD, catalytic domain; DD, disintegrin domain; ICD, intracellular domain; MPD, membrane proximal domain; PRO, prodomain; SP, signal peptide; TM, transmembrane domain. * p < 0.05, ** p < 0.01, *** p < 0.005. We further used ADAM10/ADAM17 dKO HEK cells for structural and functional studies of ADAM17 variants, in which absence of ADAM10 and ADAM17 was verified by Western blotting and immunofluorescence ( Supplemental Fig. 1B, 1C ). ADAM17 expression in reconstituted HEK A10/A17 dKO cells was verified by Western blotting ( Fig. 1F , Supplemental Fig. 1D ) and immunofluorescence ( Supplemental Fig. 1G ). For validation of ADAM17 bands on Western blots, three different Abs were used: two C-terminal (18.2; Abcam) and one N-terminal (10.1; Fig. 1F , Supplemental Fig. 1D ; see also Fig. 1C ). The mature form of overexpressed ADAM17 protein (∼90 kDa) could only be detected by Western blotting after enrichment of the protein by immunoprecipitation with a C-terminal ADAM17 Ab ( Supplemental Fig. 1E ). Further, no differences in glycosylation patterns between the different ADAM17 variants could be found as indicated by a peptide- N -glycosidase F digest ( Supplemental Fig. 1F ). Cell surface localization of all ADAM17 variants in HEK cells was verified by FACS analysis, using an N-terminal Ab ( Fig. 1G ). To further determine enzyme activity on the cell surface, a fluorogenic peptide cleavage assay was performed, indicating activity for all ADAM17 variants comparable to the wt except for the RVNG us mutant ( Fig. 1H ). To further analyze the intracellular localization of ADAM17 mutants before and after PMA stimulation in HEK A10/A17 dKO cells, costainings with an endoplasmic reticulum [ER] marker were performed ( Supplemental Fig. 1G ). Interestingly, highest ER colocalization was found for the ADAM17 Δ700 variant under nonstimulated and stimulated conditions ( Supplemental Fig. 1G, 1H ). This could indicate less efficient trafficking of this ADAM17 mutant from the ER to other cellular compartments, such as the cell surface, compared with the ADAM17 wt ( Supplemental Fig. 1G, 1H ; see also Fig. 1G ). However, cell surface localization of the proteolytically active ADAM17 variants seems to be sufficient for protease activity as indicated by the TNFR II ELISA in mEF ( Fig. 1D ) and the fluorogenic peptide cleavage assay in HEK cells ( Fig. 1H ). To confirm the described furin resistance of the ADAM17 mutants ( RVNG ds and RVNG us), their maturation was analyzed by a recombinant furin assay and Western blotting. Immunoprecipitated ADAM17 from transfected HEK A10/A17 dKO cells was incubated with recombinant furin, resulting in the immature form of ADAM17 (∼120 kDa) and, in case of cleavage of the prodomain, a mature ∼90 kDa–sized form ( Fig. 1I ). Maturation of ADAM17 protein could be observed for ADAM17 wt and surprisingly also, but to a lesser extent, the RVNG us mutant but not for the RVNG ds mutant ( Fig. 1I ). Modeling the prodomain of ADAM17 on the basis of the adamalysin crystal structure [previously published structure PDB:3P24 in Fig. 1J , right ( 39 )] shows the potential orientation and interaction of the prodomain with the catalytic domain ( Fig. 1J ). Furin cleavage sites (us: aa 57–58; ds: aa 213–214) are shown in orange ( Fig. 1J ), illustrating how the cleavage of the prodomain and, thus, maturation of ADAM17 could be mediated. The macrophage progenitor system (MØP) on ADAM17 ex/ex background Because there still is debate and controversy on posttranslational regulation processes of ADAM17, we analyzed ADAM17 maturation and activity under more physiologic conditions. We exploited a cell system of conditionally HoxB8-immortalized macrophagic progenitor cells (MØP) ( 32 ) to study ADAM17 cellular biology. From the bone marrow of wt and ADAM17 ex/ex mice, hematopoietic stem cells were enriched by selection for CD117 and infected with an estrogen-regulated HoxB8 expression vector, allowing proliferation of macrophage progenitors (MØP) in the presence of estrogen ( 32 ) ( Fig. 2A ). The progenitor cells can easily be transduced by retroviral vectors coding for the different variants of ADAM17 shown in Fig. 1A . After selection of stably transduced MØP, these were differentiated into either macrophage- or dendritic-like cells by application of M-CSF (M-MØ) or GM-CSF (GM-MØ), respectively ( Fig. 2A ). First, immortalized MØP as well as differentiated M-MØ and GM-MØ on wt or ADAM17 ex/ex background were characterized by imaging, Western blotting, and FACS analysis ( Fig. 2B–E ). Electron microscopy (EM) pictures verified characteristic morphology of undifferentiated macrophagic precursor cells, exhibiting similar size and appearance of wt and ADAM17 ex/ex -derived MØP ( Fig. 2B ). Furthermore, Con A precipitation and subsequent immunoblotting confirmed ADAM17 expression in wt but not in ADAM17 ex/ex -derived MØP ( Fig. 2C ). After differentiation of the precursor cells to mature M-MØ and GM-MØ, EM pictures revealed cell type–specific morphology of M-MØ and GM-MØ, with no differences between wt and ADAM17 ex/ex -derived cells ( Fig. 2D ). Further, M-MØ and GM-MØ showed increased expression of F4/80, CD206, CD11b, and CD11c in comparison with undifferentiated MØP ( Fig. 2E ). As expected, GM-CSF differentiated cells exhibited higher expression of CD206 and CD11c in comparison with the M-CSF differentiated cells ( Fig. 2E ) ( 32 ). Interestingly, in contrast to mEF cells, in M-MØ, ADAM17-mediated shedding of TNF-α and TNFR II could be stimulated with the immunologically relevant stimuli LPS or zymosan, indicating the usefulness of our cell system to study ADAM17 activation under physiologic conditions ( Supplemental Fig. 2A, 2B ). FIGURE 2. Establishment and characterization of MØP on wt and ADAM17 ex/ex background. ( A ) Flow diagram of protocol to establish MØP from wt and ADAM17 ex/ex mice and further differentiation to M-MØ and GM-MØ by application of M-CSF and GM-CSF. Top, Scheme of the retroviral HoxB8 expression plasmid, which is used for estrogen-dependent immortalization of CD117-positive bone marrow–derived cells. ( B ) Electron microscopic picture of HoxB8-immortalized macrophage progenitor cells gained from wt and ADAM17 ex/ex mice, indicating equal shape and size of MØP. Scale bar, 10 μm. ( C ) Representative immunoblot of Con A–enriched ADAM17, validating protein expression in wt but not in ADAM17 ex/ex MØP using an N-terminal (10.1) Ab. ( D ) EM was performed on differentiated M-MØ and GM-MØ from wt and ADAM17 ex/ex MØP. M-MØ from either genetic background exhibit characteristic round cell structure, whereas the GM-MØ show a narrower cell shape with extensions. Scale bar, 20 μm. ( E ) FACS analysis of cell surface markers of differentiated M-MØ and GM-MØ in wt and ADAM17 ex/ex background in comparison with undifferentiated MØP. Shaded histograms represent the receptor staining of differentiated M-MØ or GM-MØ. Bold lines indicate receptor signal of undifferentiated MØP. Signal of the receptor staining increases according to differentiation and is independent of ADAM17 presence. The histograms are representative of plots from three independent experiments for each marker. Using genetic modification of MØP to study ADAM17 biology To study ADAM17 function in the precursor as well as in the differentiated macrophage- and dendritic-like cells, MØP were transduced with C-terminally myc-tagged ADAM17 variants described above ( Fig. 1A ). Stable and equal expression of each ADAM17 construct in the different cell lines (MØP, M-MØ, and GM-MØ) was verified by immunoblotting with two different C-terminal, an anti-myc, and an N-terminal ADAM17 Ab ( Fig. 3A , Supplemental Fig. 2C ). Furthermore, the ability to differentiate in M-MØ or GM-MØ after transduction with an ADAM17-encoding cDNA was confirmed by light microscopy and FACS analysis ( Fig. 3B , Supplemental Fig. 2D ). M-MØ differentiation was monitored up to day 6, and GM-MØ differentiation was monitored until day 8. Both time points were chosen to ensure complete differentiation in either cell type ( Fig. 3B ). Undifferentiated MØP (day 0) were suspension cells, which became adherent during differentiation into macrophages or dendritic-like cells and grew in monolayers ( Fig. 3B ). The rate of differentiation for every ADAM17-reconstituted MØP line into M-MØ and GM-MØ was evaluated by cell morphology analyzing shape and size using light microscopy. Differentiation did not differ between ADAM17 mutants, indicating that this process is ADAM17 independent, and ADAM17 mutants do not influence macrophage development ( Fig. 3B ). This result was supported by FACS analyses showing an increase in the immune cell surface markers F4/80, CD206, CD11b, and CD11c in ADAM17-reconstituted M-MØ and GM-MØ as compared with the precursor cells on day 6 for M-MØ and day 8 for GM-MØ ( Supplemental Fig. 2D ). FIGURE 3. Using the MØP system to study ADAM17 function. ( A ) MØP with ADAM17-deficient background (ADAM17 ex/ex ) were stably reconstituted with ADAM17 variants and further differentiated to M-MØ and GM-MØ. By immunoblotting, stable protein expression of each ADAM17 variant was regularly verified using C-terminal (18.2) as well as N-terminal (10.1) Abs. The presented immunoblots are representative for at least four independent differentiations. β-Actin was used as loading control. ( B ) To ensure equal and complete differentiation of MØP to M-MØ and GM-MØ, cell morphology was monitored by light microscopy for each ADAM17-reconstituted cell line. M-CSF was applied for complete differentiation to M-MØ until day 6 (d6), and representative pictures are shown for day 3 and day 6. Application of GM-CSF had to be continued until day 8 to ensure complete differentiation to GM-MØ, and representative pictures were taken at day 4 and day 8. Undifferentiated MØP are suspension cells, becoming adherent during differentiation growing in characteristic monolayers. Scale bar, 200 μm. Analyzing function and cellular localization of ADAM17 variants in the MØP system MØP, M-MØ, and GM-MØ expressing different ADAM17 variants were analyzed for shedding of the endogenous ADAM17 substrates TNF-α and TNFR II by ELISA after stimulation with LPS and zymosan ( Fig. 4A , 4B ). To validate the assay and to ensure efficient and identical activation of ADAM17-reconstituted cell lines by LPS and zymosan, inflammatory protein-like IL-6 as well as the chemokine (C-C motif) ligand 4 were measured, showing similar ADAM17-independent cell stimulation ( Supplemental Fig. 3A, 3B ). Only the ADAM17 S794A variant exhibited increased IL6 levels after stimulation in M-MØ ( Supplemental Fig. 3A , middle panel). To confirm that soluble TNF-α and soluble TNFR II release resulted from ADAM17 activity, a specific ADAM10 inhibitor (GI254023×) and a combined ADAM10 and ADAM17 inhibitor (GW280264×) ( 40 , 41 ) were added to the MØP prior to stimulation. As shown in Supplemental Fig. 3C , the ADAM10 inhibitor GI did not significantly reduce shed TNF-α levels, whereas the ADAM10/ADAM17 inhibitor GW decreased TNF-α shedding when ADAM17 was present. Further, it was verified that substrate levels (TNFR II and TNF-α) were neither influenced by LPS stimulation nor stable ADAM17 expression in the MØP system ( Fig. 4C–E ). FIGURE 4. Analyzing function and cellular localization of ADAM17 mutants in primary immune cells. TNF-α– ( A ) and TNFR II – ( B ) shedding ELISA of cell supernatant of stably ADAM17-reconstituted ADAM17 ex/ex MØP, M-MØ, and GM-MØ after stimulation with LPS (light gray) and zymosan (dark gray). Results are shown normalized to the mock-transfected control, and dotted lines indicate baseline shedding activity ( n = 9–12, derived from three to four independent rounds of differentiation). Representative immunoblot of the ADAM17 substrates TNFR II ( C ) and pro–TNF-α ( D ) in MØP lysates without and after LPS stimulation. For detection of substrates, Abs targeted against cytoplasmic epitopes were used. ( E ) Densitometric analysis of the ∼70-kDa-sized band of TNFR II as well as the main form of pro–TNF-α (∼35 kDa) normalized to loading control β-actin indicate no differences of substrate level after LPS stimulation compared with unstimulated conditions ( n = 3). ( F and G ) Cell surface staining of ADAM17 in stably ADAM17-transfected ADAM17 ex/ex MØP by FACS analysis using an N-terminal ADAM17 Ab (10.1). (F) Gating strategy for analysis: single cells were gated (FSC-A × FSC-H), then live cells (7-AAD negative) were separated and plotted for secondary Ab signal (Alexa Fluor 488). (G) Median fluorescence intensity normalized to mock-transfected cells. Dotted line indicates baseline fluorescence ( n = 3). ( H ) Immunoblot of iRhom2 in M-MØ and GM-MØ in ADAM17-deficient background and after reconstitution with ADAM17 wt. β-Actin was used as loading control. ( I ) Representative immunofluorescence staining of iRhom2 and KDEL as ER marker in M-MØ and GM-MØ. The white box shows magnification of cell surface structures. Scale bar, 10 μm. If not indicated differently, statistical significance is shown in comparison with the respective ADAM17 wt (A and B) or mock control (G), using a one-way ANOVA, followed by a Tukey multiple comparison test in (A), (B), and (G). A two-sided Student t test was applied in (E), exhibiting no significant differences. * p < 0.05, ** p < 0.01, *** p < 0.005, **** p < 0.001. Constitutive shedding of TNF-α in all three monocytic cell lines was negligible for all ADAM17 variants in comparison with the mock-transfected control, with no ADAM17 mutant being significantly different from wt ADAM17 ( Supplemental Fig. 3D ). Interestingly, shedding of TNFR II was significantly increased for the Δ700 and S794A mutants in GM-MØ under unstimulated conditions ( Supplemental Fig. 3E , right panel). After stimulation of MØP, M-MØ, and GM-MØ with either LPS or zymosan, efficient TNF-α shedding was observed in cells expressing ADAM17 wt, S794A mutant, and Δ700 mutant ( Fig. 4A ). For the two RVNG mutants (ds and us), diminished TNF-α cleavage was found in M-MØ and GM-MØ ( Fig. 4A ), underlining the importance of the furin-mediated removal of the propeptide on the ds as well as on the us furin cleavage site for ADAM17 activation ( 13 ). Similar results for both RVNG mutants (ds and us) were obtained for a second ADAM17 substrate, TNFR II ( Fig. 4B ). Notably, TNFR II shedding was less efficiently activated by zymosan in all three cell types independent of reconstituted ADAM17 variants ( Fig. 4B ). Similar to TNF-α shedding ( Fig. 4A ), the ADAM17 Δ700 mutant and the PLK2 phosphorylation-resistant S794A mutant processed TNFR II efficiently, confirming that the C-terminal part and the PLK2 phosphorylation site are dispensable for protease activity in the analyzed immune cells ( Fig. 4B ). To further study maturation and the role of cellular localization of ADAM17, intracellular immunofluorescence studies were performed by costaining of ADAM17 and ER marker protein disulfide-isomerase in differentiated M-MØ and GM-MØ after LPS stimulation ( Supplemental Fig. 4 ). Immunofluorescent staining of M-MØ (top panel) and GM-MØ (bottom panel) cells showed characteristic cell morphology. To some extent, all ADAM17 variants (red) showed an overlap with the ER marker protein disulfide-isomerase (green) as well as cell surface localization (small magnification; Supplemental Fig. 4 ). Judging from the results, we felt that immunofluorescence studies were not sensitive enough to study cell surface localization of each ADAM17 variant in detail. Thus, cell surface detection of ADAM17 was performed in MØP by FACS analysis, as done in HEK cells ( Fig. 4F , 4G ; compare Fig. 1G ). After staining the MØP with an N-terminal ADAM17 Ab and gating for single cells and live cells (7-AAD negative), cells were plotted for secondary Ab signals (488 nm; Fig. 4F ). Interestingly, as already indicated by immunofluorescence studies ( Supplemental Fig. 4 ), the ADAM17 Δ700 mutant was found to a lesser extent in cell surface structures compared with all other ADAM17 mutants and ADAM17 wt ( Fig. 4G ). But judging from the ELISA data, surface localization of the ADAM17 Δ700 was abundant enough for sufficient TNF-α and TNFR II shedding, which was comparable to ADAM17 wt ( Fig. 4A , 4B ). To study intracellular ADAM17 regulation and transport in more detail, endogenous protein levels of the transmembrane protein iRhom2, which has been described to be required for ADAM17 ER-to-Golgi transport, cell surface stability, and protease activity ( 42 , 43 ) were analyzed ( Fig. 4H , 4I ). Immunoblots of endogenous iRhom2 revealed a similar expression of the protein in M-MØ and GM-MØ, with slightly higher protein levels in GM-MØ ( Fig. 4H ). Furthermore, immunofluorescence stainings indicated ER as well as cell surface distribution of endogenous iRhom2 in cells of the macrophage progenitor cell system (M-MØ and GM-MØ; Fig. 4I ). Overall, our results suggest that cleavage of the prodomain of ADAM17 at both furin cleavage sites was required for efficient ADAM17 activity in MØP-derived macrophage- and dendritic-like cells, whereas the cytoplasmic tail and its posttranslational modifications seemed to be dispensable for ADAM17 function, but might rather play a role in efficient cell surface trafficking, especially in immune cells. Analysis of podosomes, a physiologic readout to validate ADAM17 function in myeloid cells Podosomes are found in cells derived from the monocyte lineage, including macrophages and dendritic cells ( 44 – 47 ), and have been suggested to be involved in cell migration and invasion ( 44 , 45 , 47 ). Podosomes consist of a core of F-actin, which is typically surrounded by a ring of actin-regulatory proteins like integrins, paxillin, vinculin, and talin ( 45 ). The requirement of ADAM17 in TLR-induced disassembly of podosomes was reported ( 48 ). In this study, we use podosome number and distribution as a further physiological readout to examine ADAM17 function in myeloid cells. Moreover, the regulatory role of the cytoplasmic tail of ADAM17 was analyzed by studying the effect of the cytoplasmic deletion mutant Δ700 on TLR-induced podosome disassembly ( Fig. 5A , 5C , 5D ). To this end, ADAM17 ex/ex MØP reconstituted with ADAM17 wt, and the Δ700 mutant were differentiated to M-MØ and stimulated with the TLR-ligand LPS. Podosomes were visualized by staining F-actin with fluorescently labeled phalloidin and costained for a second podosome marker (vinculin) to highlight the presence of the podosome ring structure to distinguish them from simple actin dots ( Fig. 5A , 5B ). A cell mask staining was used to visualize cell membranes ( Fig. 5A ), to distinguish single cells for quantification, and to calculate cell area, which did not change between mock-transfected and ADAM17 wt or Δ700 variant ( Fig. 5E ). Both, number as well as density of podosomes per cell were significantly reduced when ADAM17 wt or the Δ700 mutants were reconstituted in comparison with mock-transfected cells ( Fig. 5A , 5C , 5D ). This emphasizes the significant role of ADAM17 in podosome disassembly and suggests that a truncated ADAM17 C terminus of 6 aa is sufficient for keeping the equilibrium of podosome assembly and disassembly in macrophages (M-MØ). These data identify the MØP system as a well-suited cell model to study podosome regulation, which also seems to be a robust physiological readout assay for ADAM17 function in monocytic cells. The results underline the importance of ADAM17 on podosome disassembly and thus cell motility and normal cell function of macrophages. FIGURE 5. Podosome formation as physiological readout for ADAM17 function. ( A ) Podosomes were visualized by immunofluorescence stainings in LPS-stimulated M-MØ with ADAM17-deficient background (ADAM17 ex/ex ) either expressing empty vector (mock), ADAM17 wt, or cytoplasmic deletion mutant ADAM17 Δ700. Podosomes were visualized by staining F-actin with Alexa Fluor 488–labeled phalloidin (green), and cell membranes were highlighted by using a CellMask staining (red) to clearly distinguish individual cells for quantification. Scale bar, 10 μm. ( B ) As a second marker for podosomes, vinculin, a component of the characteristic podosome ring structure, was stained using a specific primary Ab (green). F-actin staining is shown in red. Scale bar, 10 μm; insets: 1 μm. Quantification of podosome staining in mock, wt, and Δ700-reconstituted cells expressed as podosomes per cell ( C ) and podosome density (podosomes/100 μm 2 ) ( D ). ( E ) For analysis of podosome number and density, the cell area of M-MØ was determined by a cell-surrounding white line [as shown in (A) and (B)]. No statistical difference was found between the cell area of mock, ADAM17 wt, and ADAM17 Δ700–reconstituted M-MØ ( n = 3 × 30; each time, 30 cells were analyzed in three independent cell differentiations). A one-way ANOVA followed by a Tukey multiple comparison test was applied. * p < 0.05, *** p < 0.005. Functional and structural characterization of ADAM17 mutants in mEF and HEK cells We generated four ADAM17 mutants by introducing point mutations and a C-terminal truncation after residue 699, leaving a cytoplasmic rest of 6 aa (ADAM17 Δ700). Two furin-resistant mutants were generated by introduction of an RVNG motif in either the ds ( RVKR ) or us ( RKRD ) furin cleavage site. Furthermore, we mutated the phosphorylation site of ADAM17 by replacing Ser 794 with Ala ( Fig. 1A ). To study posttranslational processes of ADAM17, mutants were analyzed for expression, cellular localization, and proteolytic activity. Initial experiments were performed in mEF and HEK cells on an ADAM17-deficient background. mEF were obtained from hypomorphic ADAM17 ex/ex animals and transiently transfected with the respective ADAM17 mutants. Equal protein expression of all ADAM17 variants was verified by SDS-PAGE and Western blotting using C- and N-terminal ADAM17 Abs under unstimulated (−PMA) and stimulated (+PMA) conditions ( Fig. 1B ). It should be noted that the cytoplasmic deletion mutant (Δ700) can only be detected by the N-terminal ADAM17 Ab ( Fig. 1B ). Epitopes of used Abs are visualized on a schematic ADAM17 protein ( Fig. 1C ). For functional analysis, ADAM17-transfected mEF were stimulated by PMA. Cell supernatants were analyzed for endogenous TNFR II shedding by ELISA, showing an increase in soluble TNFR II levels, indicating enhanced ADAM17 activity in ADAM17-reconstituted mEF in comparison to mock-transfected cells ( Fig. 1D ). Mutants ADAM17 RVNG ds, S794A, and Δ700 exhibited higher TNFR II -shedding activities than mock transfectants, and only the RVNG us mutant showed no protease activity after PMA stimulation ( Fig. 1D ). No changes in TNFR II protein expression were observed under the experimental conditions used ( Fig. 1E , Supplemental Fig. 1A ). FIGURE 1. Functional and structural analyses of ADAM17 mutants in mEF and HEK cells. ( A ) Schematic overview of the 827-aa-long murine ADAM17 protein subdivided in different domains and ADAM17 mutants analyzed in this study. Mutations introduced in the prodomain to generate the RVNG ds and RVNG us mutants are indicated by red letters. The location of the S794A point mutation is indicated by a black cross, and the ADAM17 Δ700 mutant is truncated after aa 699. ( B ) Immunoblot of endogenous ADAM17 in wt mEF and transiently overexpressed ADAM17 in mEF exhibiting an ADAM17 knockdown (ADAM17 ex/ex ). Cells were nonstimulated (−PMA) or stimulated with PMA for 2 h. C-terminal (18.2) as well as N-terminal (10.1) Abs were used to detect ADAM17, indicating uniform expression of all ADAM17 mutants. The Δ700 can only be detected by the N-terminal Ab. β-Actin was used as loading control. ( C ) Schematic model of ADAM17 protein indicating binding sites of used Abs. The polyclonal N-terminal ADAM17 Ab 10.1 was designed against a peptide (aa 290–309) in the catalytic domain (CD). Both C-terminal Abs (ab39162 and house-made 18.2; Abcam) are polyclonal and were designed against a cytoplasmic epitope. The 18.2 Ab was designed to bind a peptide (aa 738–757) within the intracellular domain (ICD). ( D ) ELISA of endogenous TNFR II in supernatants of ADAM17 ex/ex mEF reconstituted with each respective ADAM17 mutant, showing constitutive shedding activity (black bars) and TNFR II shedding after a 2-h PMA stimulation (gray bars). The dotted line marks baseline TNFR II shedding ( n = 3 from three individual transfections). ( E ) Endogenous TNFR II protein level in ADAM17 ex/ex mEF after mock transfection and reconstitution with ADAM17 wt with and without a 2-h PMA stimulation. Signal intensity analysis of TNFR II normalized to loading control (β-actin) indicates unchanged TNFR II level. ( F ) Representative immunoblot of ADAM17 mutants transiently overexpressed in ADAM10/ADAM17 double-deficient HEK cells (HEK A10/A17 dKO) detected by a C-terminal (18.2) as well as N-terminal Ab (10.1) showing equal ADAM17 protein level. β-Actin was used as loading control. ( G ) Cell surface FACS analysis of ADAM17 overexpressed in HEK A10/A17 dKO cells stained with extracellular N-terminal Ab 10.1. Single cells (FSC-A × FSC-H) were gated and plotted for secondary Ab signal (Alexa Fluor 488). Bar graph shows median fluorescence intensity after normalization to mock transfection ( n = 3 from three individual transfections). ( H ) ADAM17 life cell surface activity assay in ADAM17-reconstituted HEK A10/17 dKO cells by fluorogenic peptide cleavage assay. Protease activity was determined by measuring fluorescence intensity of the cleaved peptide on living cells for 6000 s. The area under the curve of fluorescent signal over time was calculated and normalized to the mock-transfected control ( n = 3 from three individual transfections). ( I ) Representative Western blot of three individual experiments (anti C-terminal Ab [Abcam]) of a furin cleavage assay. Lysates of ADAM17-transfected HEK A10/A17 dKO cells were immunoprecipitated by a C-terminal Ab (Abcam) and incubated with 1 U of recombinant furin. Upper ADAM17 band (∼120 kDa) indicates full-length and nonfurin-processed ADAM17 form. After furin cleavage of the prodomain, the mature ADAM17 migrates at ∼90 kDa (black arrow). ( J ) Protein model of ADAM17 highlighting catalytic and prodomain (gray). In orange, the us and ds furin cleavage sites are indicated. The prodomain of ADAM17 and its orientation toward the catalytic domain were modeled on the basis of the previously published adamalysin crystal structure [right; PDB:3P24 ( 39 )] using Swiss Model Expasy. The red highlighted Asp194 in adamalysin is analog to the Asp190 in the prodomain of ADAM17. Both aspartates interact with a zinc ion (orange dot). Molecular imaging was performed using UCSF Chimera. For statistical analyses, a two-sided Student t test was used in (E), exhibiting no statistically significant differences. A one-way ANOVA followed by a Tukey multiple comparison test was applied in (D), (G), and (H). CD, catalytic domain; DD, disintegrin domain; ICD, intracellular domain; MPD, membrane proximal domain; PRO, prodomain; SP, signal peptide; TM, transmembrane domain. * p < 0.05, ** p < 0.01, *** p < 0.005. We further used ADAM10/ADAM17 dKO HEK cells for structural and functional studies of ADAM17 variants, in which absence of ADAM10 and ADAM17 was verified by Western blotting and immunofluorescence ( Supplemental Fig. 1B, 1C ). ADAM17 expression in reconstituted HEK A10/A17 dKO cells was verified by Western blotting ( Fig. 1F , Supplemental Fig. 1D ) and immunofluorescence ( Supplemental Fig. 1G ). For validation of ADAM17 bands on Western blots, three different Abs were used: two C-terminal (18.2; Abcam) and one N-terminal (10.1; Fig. 1F , Supplemental Fig. 1D ; see also Fig. 1C ). The mature form of overexpressed ADAM17 protein (∼90 kDa) could only be detected by Western blotting after enrichment of the protein by immunoprecipitation with a C-terminal ADAM17 Ab ( Supplemental Fig. 1E ). Further, no differences in glycosylation patterns between the different ADAM17 variants could be found as indicated by a peptide- N -glycosidase F digest ( Supplemental Fig. 1F ). Cell surface localization of all ADAM17 variants in HEK cells was verified by FACS analysis, using an N-terminal Ab ( Fig. 1G ). To further determine enzyme activity on the cell surface, a fluorogenic peptide cleavage assay was performed, indicating activity for all ADAM17 variants comparable to the wt except for the RVNG us mutant ( Fig. 1H ). To further analyze the intracellular localization of ADAM17 mutants before and after PMA stimulation in HEK A10/A17 dKO cells, costainings with an endoplasmic reticulum [ER] marker were performed ( Supplemental Fig. 1G ). Interestingly, highest ER colocalization was found for the ADAM17 Δ700 variant under nonstimulated and stimulated conditions ( Supplemental Fig. 1G, 1H ). This could indicate less efficient trafficking of this ADAM17 mutant from the ER to other cellular compartments, such as the cell surface, compared with the ADAM17 wt ( Supplemental Fig. 1G, 1H ; see also Fig. 1G ). However, cell surface localization of the proteolytically active ADAM17 variants seems to be sufficient for protease activity as indicated by the TNFR II ELISA in mEF ( Fig. 1D ) and the fluorogenic peptide cleavage assay in HEK cells ( Fig. 1H ). To confirm the described furin resistance of the ADAM17 mutants ( RVNG ds and RVNG us), their maturation was analyzed by a recombinant furin assay and Western blotting. Immunoprecipitated ADAM17 from transfected HEK A10/A17 dKO cells was incubated with recombinant furin, resulting in the immature form of ADAM17 (∼120 kDa) and, in case of cleavage of the prodomain, a mature ∼90 kDa–sized form ( Fig. 1I ). Maturation of ADAM17 protein could be observed for ADAM17 wt and surprisingly also, but to a lesser extent, the RVNG us mutant but not for the RVNG ds mutant ( Fig. 1I ). Modeling the prodomain of ADAM17 on the basis of the adamalysin crystal structure [previously published structure PDB:3P24 in Fig. 1J , right ( 39 )] shows the potential orientation and interaction of the prodomain with the catalytic domain ( Fig. 1J ). Furin cleavage sites (us: aa 57–58; ds: aa 213–214) are shown in orange ( Fig. 1J ), illustrating how the cleavage of the prodomain and, thus, maturation of ADAM17 could be mediated. The macrophage progenitor system (MØP) on ADAM17 ex/ex background Because there still is debate and controversy on posttranslational regulation processes of ADAM17, we analyzed ADAM17 maturation and activity under more physiologic conditions. We exploited a cell system of conditionally HoxB8-immortalized macrophagic progenitor cells (MØP) ( 32 ) to study ADAM17 cellular biology. From the bone marrow of wt and ADAM17 ex/ex mice, hematopoietic stem cells were enriched by selection for CD117 and infected with an estrogen-regulated HoxB8 expression vector, allowing proliferation of macrophage progenitors (MØP) in the presence of estrogen ( 32 ) ( Fig. 2A ). The progenitor cells can easily be transduced by retroviral vectors coding for the different variants of ADAM17 shown in Fig. 1A . After selection of stably transduced MØP, these were differentiated into either macrophage- or dendritic-like cells by application of M-CSF (M-MØ) or GM-CSF (GM-MØ), respectively ( Fig. 2A ). First, immortalized MØP as well as differentiated M-MØ and GM-MØ on wt or ADAM17 ex/ex background were characterized by imaging, Western blotting, and FACS analysis ( Fig. 2B–E ). Electron microscopy (EM) pictures verified characteristic morphology of undifferentiated macrophagic precursor cells, exhibiting similar size and appearance of wt and ADAM17 ex/ex -derived MØP ( Fig. 2B ). Furthermore, Con A precipitation and subsequent immunoblotting confirmed ADAM17 expression in wt but not in ADAM17 ex/ex -derived MØP ( Fig. 2C ). After differentiation of the precursor cells to mature M-MØ and GM-MØ, EM pictures revealed cell type–specific morphology of M-MØ and GM-MØ, with no differences between wt and ADAM17 ex/ex -derived cells ( Fig. 2D ). Further, M-MØ and GM-MØ showed increased expression of F4/80, CD206, CD11b, and CD11c in comparison with undifferentiated MØP ( Fig. 2E ). As expected, GM-CSF differentiated cells exhibited higher expression of CD206 and CD11c in comparison with the M-CSF differentiated cells ( Fig. 2E ) ( 32 ). Interestingly, in contrast to mEF cells, in M-MØ, ADAM17-mediated shedding of TNF-α and TNFR II could be stimulated with the immunologically relevant stimuli LPS or zymosan, indicating the usefulness of our cell system to study ADAM17 activation under physiologic conditions ( Supplemental Fig. 2A, 2B ). FIGURE 2. Establishment and characterization of MØP on wt and ADAM17 ex/ex background. ( A ) Flow diagram of protocol to establish MØP from wt and ADAM17 ex/ex mice and further differentiation to M-MØ and GM-MØ by application of M-CSF and GM-CSF. Top, Scheme of the retroviral HoxB8 expression plasmid, which is used for estrogen-dependent immortalization of CD117-positive bone marrow–derived cells. ( B ) Electron microscopic picture of HoxB8-immortalized macrophage progenitor cells gained from wt and ADAM17 ex/ex mice, indicating equal shape and size of MØP. Scale bar, 10 μm. ( C ) Representative immunoblot of Con A–enriched ADAM17, validating protein expression in wt but not in ADAM17 ex/ex MØP using an N-terminal (10.1) Ab. ( D ) EM was performed on differentiated M-MØ and GM-MØ from wt and ADAM17 ex/ex MØP. M-MØ from either genetic background exhibit characteristic round cell structure, whereas the GM-MØ show a narrower cell shape with extensions. Scale bar, 20 μm. ( E ) FACS analysis of cell surface markers of differentiated M-MØ and GM-MØ in wt and ADAM17 ex/ex background in comparison with undifferentiated MØP. Shaded histograms represent the receptor staining of differentiated M-MØ or GM-MØ. Bold lines indicate receptor signal of undifferentiated MØP. Signal of the receptor staining increases according to differentiation and is independent of ADAM17 presence. The histograms are representative of plots from three independent experiments for each marker. Using genetic modification of MØP to study ADAM17 biology To study ADAM17 function in the precursor as well as in the differentiated macrophage- and dendritic-like cells, MØP were transduced with C-terminally myc-tagged ADAM17 variants described above ( Fig. 1A ). Stable and equal expression of each ADAM17 construct in the different cell lines (MØP, M-MØ, and GM-MØ) was verified by immunoblotting with two different C-terminal, an anti-myc, and an N-terminal ADAM17 Ab ( Fig. 3A , Supplemental Fig. 2C ). Furthermore, the ability to differentiate in M-MØ or GM-MØ after transduction with an ADAM17-encoding cDNA was confirmed by light microscopy and FACS analysis ( Fig. 3B , Supplemental Fig. 2D ). M-MØ differentiation was monitored up to day 6, and GM-MØ differentiation was monitored until day 8. Both time points were chosen to ensure complete differentiation in either cell type ( Fig. 3B ). Undifferentiated MØP (day 0) were suspension cells, which became adherent during differentiation into macrophages or dendritic-like cells and grew in monolayers ( Fig. 3B ). The rate of differentiation for every ADAM17-reconstituted MØP line into M-MØ and GM-MØ was evaluated by cell morphology analyzing shape and size using light microscopy. Differentiation did not differ between ADAM17 mutants, indicating that this process is ADAM17 independent, and ADAM17 mutants do not influence macrophage development ( Fig. 3B ). This result was supported by FACS analyses showing an increase in the immune cell surface markers F4/80, CD206, CD11b, and CD11c in ADAM17-reconstituted M-MØ and GM-MØ as compared with the precursor cells on day 6 for M-MØ and day 8 for GM-MØ ( Supplemental Fig. 2D ). FIGURE 3. Using the MØP system to study ADAM17 function. ( A ) MØP with ADAM17-deficient background (ADAM17 ex/ex ) were stably reconstituted with ADAM17 variants and further differentiated to M-MØ and GM-MØ. By immunoblotting, stable protein expression of each ADAM17 variant was regularly verified using C-terminal (18.2) as well as N-terminal (10.1) Abs. The presented immunoblots are representative for at least four independent differentiations. β-Actin was used as loading control. ( B ) To ensure equal and complete differentiation of MØP to M-MØ and GM-MØ, cell morphology was monitored by light microscopy for each ADAM17-reconstituted cell line. M-CSF was applied for complete differentiation to M-MØ until day 6 (d6), and representative pictures are shown for day 3 and day 6. Application of GM-CSF had to be continued until day 8 to ensure complete differentiation to GM-MØ, and representative pictures were taken at day 4 and day 8. Undifferentiated MØP are suspension cells, becoming adherent during differentiation growing in characteristic monolayers. Scale bar, 200 μm. Analyzing function and cellular localization of ADAM17 variants in the MØP system MØP, M-MØ, and GM-MØ expressing different ADAM17 variants were analyzed for shedding of the endogenous ADAM17 substrates TNF-α and TNFR II by ELISA after stimulation with LPS and zymosan ( Fig. 4A , 4B ). To validate the assay and to ensure efficient and identical activation of ADAM17-reconstituted cell lines by LPS and zymosan, inflammatory protein-like IL-6 as well as the chemokine (C-C motif) ligand 4 were measured, showing similar ADAM17-independent cell stimulation ( Supplemental Fig. 3A, 3B ). Only the ADAM17 S794A variant exhibited increased IL6 levels after stimulation in M-MØ ( Supplemental Fig. 3A , middle panel). To confirm that soluble TNF-α and soluble TNFR II release resulted from ADAM17 activity, a specific ADAM10 inhibitor (GI254023×) and a combined ADAM10 and ADAM17 inhibitor (GW280264×) ( 40 , 41 ) were added to the MØP prior to stimulation. As shown in Supplemental Fig. 3C , the ADAM10 inhibitor GI did not significantly reduce shed TNF-α levels, whereas the ADAM10/ADAM17 inhibitor GW decreased TNF-α shedding when ADAM17 was present. Further, it was verified that substrate levels (TNFR II and TNF-α) were neither influenced by LPS stimulation nor stable ADAM17 expression in the MØP system ( Fig. 4C–E ). FIGURE 4. Analyzing function and cellular localization of ADAM17 mutants in primary immune cells. TNF-α– ( A ) and TNFR II – ( B ) shedding ELISA of cell supernatant of stably ADAM17-reconstituted ADAM17 ex/ex MØP, M-MØ, and GM-MØ after stimulation with LPS (light gray) and zymosan (dark gray). Results are shown normalized to the mock-transfected control, and dotted lines indicate baseline shedding activity ( n = 9–12, derived from three to four independent rounds of differentiation). Representative immunoblot of the ADAM17 substrates TNFR II ( C ) and pro–TNF-α ( D ) in MØP lysates without and after LPS stimulation. For detection of substrates, Abs targeted against cytoplasmic epitopes were used. ( E ) Densitometric analysis of the ∼70-kDa-sized band of TNFR II as well as the main form of pro–TNF-α (∼35 kDa) normalized to loading control β-actin indicate no differences of substrate level after LPS stimulation compared with unstimulated conditions ( n = 3). ( F and G ) Cell surface staining of ADAM17 in stably ADAM17-transfected ADAM17 ex/ex MØP by FACS analysis using an N-terminal ADAM17 Ab (10.1). (F) Gating strategy for analysis: single cells were gated (FSC-A × FSC-H), then live cells (7-AAD negative) were separated and plotted for secondary Ab signal (Alexa Fluor 488). (G) Median fluorescence intensity normalized to mock-transfected cells. Dotted line indicates baseline fluorescence ( n = 3). ( H ) Immunoblot of iRhom2 in M-MØ and GM-MØ in ADAM17-deficient background and after reconstitution with ADAM17 wt. β-Actin was used as loading control. ( I ) Representative immunofluorescence staining of iRhom2 and KDEL as ER marker in M-MØ and GM-MØ. The white box shows magnification of cell surface structures. Scale bar, 10 μm. If not indicated differently, statistical significance is shown in comparison with the respective ADAM17 wt (A and B) or mock control (G), using a one-way ANOVA, followed by a Tukey multiple comparison test in (A), (B), and (G). A two-sided Student t test was applied in (E), exhibiting no significant differences. * p < 0.05, ** p < 0.01, *** p < 0.005, **** p < 0.001. Constitutive shedding of TNF-α in all three monocytic cell lines was negligible for all ADAM17 variants in comparison with the mock-transfected control, with no ADAM17 mutant being significantly different from wt ADAM17 ( Supplemental Fig. 3D ). Interestingly, shedding of TNFR II was significantly increased for the Δ700 and S794A mutants in GM-MØ under unstimulated conditions ( Supplemental Fig. 3E , right panel). After stimulation of MØP, M-MØ, and GM-MØ with either LPS or zymosan, efficient TNF-α shedding was observed in cells expressing ADAM17 wt, S794A mutant, and Δ700 mutant ( Fig. 4A ). For the two RVNG mutants (ds and us), diminished TNF-α cleavage was found in M-MØ and GM-MØ ( Fig. 4A ), underlining the importance of the furin-mediated removal of the propeptide on the ds as well as on the us furin cleavage site for ADAM17 activation ( 13 ). Similar results for both RVNG mutants (ds and us) were obtained for a second ADAM17 substrate, TNFR II ( Fig. 4B ). Notably, TNFR II shedding was less efficiently activated by zymosan in all three cell types independent of reconstituted ADAM17 variants ( Fig. 4B ). Similar to TNF-α shedding ( Fig. 4A ), the ADAM17 Δ700 mutant and the PLK2 phosphorylation-resistant S794A mutant processed TNFR II efficiently, confirming that the C-terminal part and the PLK2 phosphorylation site are dispensable for protease activity in the analyzed immune cells ( Fig. 4B ). To further study maturation and the role of cellular localization of ADAM17, intracellular immunofluorescence studies were performed by costaining of ADAM17 and ER marker protein disulfide-isomerase in differentiated M-MØ and GM-MØ after LPS stimulation ( Supplemental Fig. 4 ). Immunofluorescent staining of M-MØ (top panel) and GM-MØ (bottom panel) cells showed characteristic cell morphology. To some extent, all ADAM17 variants (red) showed an overlap with the ER marker protein disulfide-isomerase (green) as well as cell surface localization (small magnification; Supplemental Fig. 4 ). Judging from the results, we felt that immunofluorescence studies were not sensitive enough to study cell surface localization of each ADAM17 variant in detail. Thus, cell surface detection of ADAM17 was performed in MØP by FACS analysis, as done in HEK cells ( Fig. 4F , 4G ; compare Fig. 1G ). After staining the MØP with an N-terminal ADAM17 Ab and gating for single cells and live cells (7-AAD negative), cells were plotted for secondary Ab signals (488 nm; Fig. 4F ). Interestingly, as already indicated by immunofluorescence studies ( Supplemental Fig. 4 ), the ADAM17 Δ700 mutant was found to a lesser extent in cell surface structures compared with all other ADAM17 mutants and ADAM17 wt ( Fig. 4G ). But judging from the ELISA data, surface localization of the ADAM17 Δ700 was abundant enough for sufficient TNF-α and TNFR II shedding, which was comparable to ADAM17 wt ( Fig. 4A , 4B ). To study intracellular ADAM17 regulation and transport in more detail, endogenous protein levels of the transmembrane protein iRhom2, which has been described to be required for ADAM17 ER-to-Golgi transport, cell surface stability, and protease activity ( 42 , 43 ) were analyzed ( Fig. 4H , 4I ). Immunoblots of endogenous iRhom2 revealed a similar expression of the protein in M-MØ and GM-MØ, with slightly higher protein levels in GM-MØ ( Fig. 4H ). Furthermore, immunofluorescence stainings indicated ER as well as cell surface distribution of endogenous iRhom2 in cells of the macrophage progenitor cell system (M-MØ and GM-MØ; Fig. 4I ). Overall, our results suggest that cleavage of the prodomain of ADAM17 at both furin cleavage sites was required for efficient ADAM17 activity in MØP-derived macrophage- and dendritic-like cells, whereas the cytoplasmic tail and its posttranslational modifications seemed to be dispensable for ADAM17 function, but might rather play a role in efficient cell surface trafficking, especially in immune cells. Analysis of podosomes, a physiologic readout to validate ADAM17 function in myeloid cells Podosomes are found in cells derived from the monocyte lineage, including macrophages and dendritic cells ( 44 – 47 ), and have been suggested to be involved in cell migration and invasion ( 44 , 45 , 47 ). Podosomes consist of a core of F-actin, which is typically surrounded by a ring of actin-regulatory proteins like integrins, paxillin, vinculin, and talin ( 45 ). The requirement of ADAM17 in TLR-induced disassembly of podosomes was reported ( 48 ). In this study, we use podosome number and distribution as a further physiological readout to examine ADAM17 function in myeloid cells. Moreover, the regulatory role of the cytoplasmic tail of ADAM17 was analyzed by studying the effect of the cytoplasmic deletion mutant Δ700 on TLR-induced podosome disassembly ( Fig. 5A , 5C , 5D ). To this end, ADAM17 ex/ex MØP reconstituted with ADAM17 wt, and the Δ700 mutant were differentiated to M-MØ and stimulated with the TLR-ligand LPS. Podosomes were visualized by staining F-actin with fluorescently labeled phalloidin and costained for a second podosome marker (vinculin) to highlight the presence of the podosome ring structure to distinguish them from simple actin dots ( Fig. 5A , 5B ). A cell mask staining was used to visualize cell membranes ( Fig. 5A ), to distinguish single cells for quantification, and to calculate cell area, which did not change between mock-transfected and ADAM17 wt or Δ700 variant ( Fig. 5E ). Both, number as well as density of podosomes per cell were significantly reduced when ADAM17 wt or the Δ700 mutants were reconstituted in comparison with mock-transfected cells ( Fig. 5A , 5C , 5D ). This emphasizes the significant role of ADAM17 in podosome disassembly and suggests that a truncated ADAM17 C terminus of 6 aa is sufficient for keeping the equilibrium of podosome assembly and disassembly in macrophages (M-MØ). These data identify the MØP system as a well-suited cell model to study podosome regulation, which also seems to be a robust physiological readout assay for ADAM17 function in monocytic cells. The results underline the importance of ADAM17 on podosome disassembly and thus cell motility and normal cell function of macrophages. FIGURE 5. Podosome formation as physiological readout for ADAM17 function. ( A ) Podosomes were visualized by immunofluorescence stainings in LPS-stimulated M-MØ with ADAM17-deficient background (ADAM17 ex/ex ) either expressing empty vector (mock), ADAM17 wt, or cytoplasmic deletion mutant ADAM17 Δ700. Podosomes were visualized by staining F-actin with Alexa Fluor 488–labeled phalloidin (green), and cell membranes were highlighted by using a CellMask staining (red) to clearly distinguish individual cells for quantification. Scale bar, 10 μm. ( B ) As a second marker for podosomes, vinculin, a component of the characteristic podosome ring structure, was stained using a specific primary Ab (green). F-actin staining is shown in red. Scale bar, 10 μm; insets: 1 μm. Quantification of podosome staining in mock, wt, and Δ700-reconstituted cells expressed as podosomes per cell ( C ) and podosome density (podosomes/100 μm 2 ) ( D ). ( E ) For analysis of podosome number and density, the cell area of M-MØ was determined by a cell-surrounding white line [as shown in (A) and (B)]. No statistical difference was found between the cell area of mock, ADAM17 wt, and ADAM17 Δ700–reconstituted M-MØ ( n = 3 × 30; each time, 30 cells were analyzed in three independent cell differentiations). A one-way ANOVA followed by a Tukey multiple comparison test was applied. * p < 0.05, *** p < 0.005. Discussion Among the ∼80 ADAM17 substrates identified so far ( 3 ), the protease has shown to be especially important for cleavage of many regulators involved in immune and inflammatory responses as well as cancer development, including membrane-tethered pro–TNF-α, both TNF-α receptors (TNFR I and TNFR II ), IL-6R, ligands of the EGF-R (e.g., TGF-α and amphiregulin), and adhesion molecules (e.g., L-selectin and ICAM-1) ( 3 , 6 ). Because inflammation is characterized by high levels of TNF-α and IL-6, and ADAM17 is essentially involved in the regulation of both cytokine pathways, it makes the metalloprotease a central player in inflammatory processes ( 6 ). Moreover, macrophages are the major cell type releasing TNF-α ( 49 ), further pointing to ADAM17 as an important regulator of inflammation in this cell type. In addition, the broad substrate repertoire of ADAM17 makes it possible that this protease is involved in pro- as well as anti-inflammatory pathways ( 6 ). For instance, TNF-α signaling can be activated by ADAM17-dependent release of the cytokine from the cell surface as well as attenuated by cleavage of TNFR, which act as antagonists of TNF-α ( 50 ). The fundamental regulatory function of ADAM17 in governing inflammatory responses requires strict regulation of protease activity. The cellular mechanisms of ADAM17 maturation and activation as well as substrate recognition are not completely understood. Numerous studies addressed the role of posttranslational modifications and their impact on ADAM17 activity. So far, most overexpression studies to analyze the biological behavior of specific ADAM17 mutants have mainly been performed in ADAM17-deficient mEF or HEK cells ( 13 , 15 , 24 , 26 , 27 ). These nonimmunological cell types have the disadvantage that heterologous substrates and nonphysiologic stimuli need to be employed to evaluate ADAM17 function. Thus, in most experimental setups, cells were cotransfected with ADAM17 substrates such as pro–TNF-α or TGF-α and various ADAM17 mutants, and shedding was induced by PMA ( 19 ). The differences of the applied cell systems and the nonphysiological conditions make interpretation as well as comparison of the different studies difficult. The continuing controversy regarding the role of posttranslational events during ADAM17 activation motivated us to establish a more physiological cell model to study ADAM17 maturation in primary cells of the immune system. We, in this study, introduced an immune cell system on an ADAM17-deficient background ( 5 ), which can be genetically modified and thus was used for expression and analysis of ADAM17 mutants. A major advantage of the HoxB8-immortalized macrophage progenitor cell line ( 31 , 32 ) is that cells can be further differentiated into macrophage- and dendritic-like cells (M-MØ and GM-MØ). Because these cells endogenously express immunologically relevant ADAM17 substrates like TNF-α and TNFR II , these substrates were used as a readout for ADAM17 activity. A further benefit of using immune cells is the possibility to work with physiologic stimuli such as the TLR agonists LPS or zymosan. We, in this study, verified the proteolytic activity of the ADAM17 Δ700 mutant, which contains six residual intracellular amino acids C-terminal of the transmembrane region, because in previous studies, a complete removal of the transmembrane region resulted in a complete inactive protease ( 26 ). The role of the cytoplasmic tail in function and regulation of the ADAM17 protease is still controversially discussed in the field. A reason for the divergent results published over the recent years could be the use of various experimental set-ups, including cell systems, choice of analyzed substrates, and the design of the ADAM17 constructs, which often contained charged C-terminal protein tags in the transmembrane region, as indicated by comparison of previous studies ( 21 – 27 , 51 ). In this study, we analyzed the activity of the ADAM17 Δ700 mutant on the cell surface of HEK as well as MØP by using a life cell activity assay. Applying the same experimental readout in both cell lines, ADAM17 Δ700 was found to a lesser extent on the cell membrane as compared with the ADAM17 wt and other ADAM17 variants in the immune cell system (MØP). This could indicate a role of the cytoplasmic tail for trafficking and thus cell surface localization in immune cells compared with nonimmune cells. However, more studies are needed to decipher a possible regulatory impact of the cytoplasmic tail of ADAM17 and to solve the discrepancy in the field about the role of its membrane-proximal portion ( 21 – 26 , 51 ). Shedding activity of the ADAM17 Δ700 mutant questions the role of intracellular phosphorylation in ADAM17 activation ( 24 – 26 ). Nonetheless, phosphorylation of ADAM17 has been regarded as a hallmark of ADAM17 activity ( 6 , 16 – 18 ). We analyzed the impact of PLK2-mediated phosphorylation at Ser 794 on ADAM17 activity because the activity of PLK2 had been predicted to be a regulator of ADAM17 activity ( 15 ). Thus, we analyzed the S794A mutant and found no interference with ADAM17 activity in mEF, HEK, and in all MØP-derived cell types. Further, this mutant was found localized at the cell membrane in all analyzed cells. Activation of ADAM17 by furin-dependent removal of the autoinhibitory propeptide seems to be complex. The proprotein convertase furin removes the prodomain within the Golgi apparatus on the ds cleavage site, which is located between the prodomain and the catalytic domain ( 11 , 12 ). A recent study identified an additional furin cleavage site within the prodomain ( 13 ). In our study, the RVNG us mutant was processed by recombinant furin, although levels of the mature (∼90 kDa) form of ADAM17 were less compared with the wt ( Fig. 1I ). In contrast, the ADAM17 RVNG ds only showed a small shift in protein size, which might indicate furin cleavage at the remaining us cleavage site ( Fig. 1I , 1J ). Both RVNG mutations showed impact on ADAM17 activity. In the mEF and HEK cell system, both furin site mutants were found at the cell surface ( Fig. 1G ), but the RVNG us mutant showed less proteolytic activity ( Fig. 1H ) as also recently described in nonimmunological cell lines ( 13 ). In contrast, both RVNG mutants showed little proteolytic activity in MØP-derived cells, although located at the cell membrane. Interestingly, it was observed earlier in COS-7 cells that inhibition of proprotein convertases had no effect on ADAM17 shedding activity ( 24 ), indicating that the role of furin processing of ADAM17 is not completely understood and will have to be studied in particular in myeloid cells. Activation of ADAM17 might be cell type–specific, depending on the expression and availability of cofactors influencing ADAM17 biology. Recently, iRhom2 has been described to be such a cofactor ( 43 , 52 , 53 ), showing to be crucially involved in ADAM17 trafficking, stability, and activation ( 42 ). Phosphorylation of the cytoplasmic tail of iRhom2 ( 42 ) might be the reason for the dispensability of the ADAM17 cytoplasmic portion upon PMA stimulation. In myeloid cells, ADAM17 is critically involved in podosome disassembly ( 48 ). In this study, we showed that the generated MØP-derived macrophage-like cells can be used to study podosome disassembly. In this physiologic cell model, substrates of ADAM17 and the disassembly mechanism involved in podosome loss can be studied, which will be important for the functional understanding of these actin-rich adhesion structures that are crucial for cell migration and matrix invasiveness of macrophages and dendritic cells ( 54 ). Supplementary Material Data Supplement Disclosures The authors have no financial conflicts of interest.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3132637/

Detection of Bacterial Endospores in Soil by Terbium Fluorescence

Spore formation is a survival mechanism of microorganisms when facing unfavorable environmental conditions resulting in "dormant" states. We investigated the occurrence of bacterial endospores in soils from various locations including grasslands (pasture, meadow), allotment gardens, and forests, as well as fluvial sediments. Bacterial spores are characterized by their high content of dipicolinic acid (DPA). In the presence of terbium, DPA forms a complex showing a distinctive photoluminescence spectrum. DPA was released from soil by microwaving or autoclaving. The addition of aluminium chloride reduced signal quenching by interfering compounds such as phosphate. The highest spore content (up to 10 9 spores per gram of dry soil) was found in grassland soils. Spore content is related to soil type, to soil depth, and to soil carbon-to-nitrogen ratio. Our study might provide a basis for the detection of "hot spots" of bacterial spores in soil. 1. Introduction The formation of spores is a survival mechanism of microorganisms when exposed to unfavorable environmental conditions (e.g., heavy metal stress, nutrient limitations) leading to a "dormant" or "resting" growth state [ 1 , 2 ]. A variety of bacteria identified in diverse habitats including soil is able to form endospores. These physiological groups include aerobic heterotrophs (e.g., Bacillus, Paenibacillus, Brevibacillus, Geobacillus, Thermoactinomyces, and Sporolactobacillus ), anaerobes ( Clostridium , Anaerobacter, and Desulfotomaculum ), microaerophiles ( Sporolactobacillus ), halophiles ( Sporohalobacter ), and phototrophs ( Heliobacterium , Heliophilum ) [ 3 , 4 ]. Bacterial spores are characterized by a series of unique chemical features which can facilitate their identification in natural environments. Besides the high content of minerals (particularly calcium), spores contain high amounts of dipicolinic acid, DPA [ 5 ]. DPA is uniquely found in bacterial spores in amounts of up to 25% of the spore dry weight and depends on the bacterial species [ 6 , 7 ]. In solution, a complex is formed in the presence of terbium which shows a very strong and distinctive fluorescence spectrum [ 8 ]. Originally, DPA was used to detect very low concentrations of terbium (III) [ 9 ]. On this basis, methods for the detection of bacterial endospores have been developed [ 10 – 13 ]: by the addition of terbium, the DPA content was determined. However, terbium-DPA fluorescence might be interfered by a series of compounds, especially when DPA has to be determined in complex samples such as sediments or soils. It has been reported that the presence of phosphorus compounds (especially ortho -phosphate) reduced terbium fluorescence by as much as 98% [ 14 ]. The addition of aluminium compounds (especially aluminium chloride, AlCl 3 ), however, ameliorated the interference caused by the quenching substances [ 14 ]. From a series of organic compounds (benzoate, tryptophan, tyrosine, phenylalanine, glucose, malate, riboflavin, NAD, and tryptone) only the latter two (especially tryptone) reduced fluorescence significantly. Carbohydrates (e.g., starch, dextrine) were reported not to interfere with the terbium fluorescence [ 15 ]. Inorganic compounds such as calcium carbonate, sodium chloride, potassium chloride, ammonium sulphate, ammonium nitrate, and sodium nitrate did not lead to a reduction of the fluorescence, but only dipotassium phosphate did [ 16 ]. The aim of this study was to adopt the fluorescence-based method to determine the spore content in soils sampled from various locations. In particular, we were interested in the differentiation between different types of soil such as grasslands (pasture, meadow), allotment gardens, and forests, as well as fluvial sediments, the relationship of soil parameters (carbon-to-nitrogen ratio) on the occurrence of bacterial spores, and the distribution of spores in relation to sampling depth. 2. Materials and Methods 2.1. Bacterial Spores Different Bacillus species ( B. megaterium, B. subtilis ) were cultivated in liquid medium containing (in g/l): glucose (3.6), ammonium chloride (2.5), magnesium sulfate (0.2), calcium chloride (0.07), iron sulfate (0.01), EDTA (0.01), potassium dihydrogen phosphate (0.6), dipotassium hydrogen phosphate (0.9), and yeast extract (1.0). Initial pH was adjusted to 7.0. Erlenmeyer flasks (250 ml) containing 100 ml of growth medium were inoculated and incubated for 10 to 15 days (150 rpm, 30°C). To initiate and stimulate sporulation, bacteria were subsequently transferred to a sporulation medium (identical composition, but without glucose and less ammonium chloride [only 1 g/l]). After additional 30 days of incubation—until vegetative cells were not present anymore after inspection by microscopy—spores were harvested by centrifugation, immediately frozen in liquid nitrogen followed by lyophilization. 2.2. Soil Samples Samples from different locations were collected using a stainless steel soil corer (15 mm in diameter), which was sterilized before each sampling. Cores with a maximum length of 25 cm were obtained, cut in sections of 5 cm, transferred to sterile screw cap Falcon tubes (20 ml), and stored on dry ice. After return to the laboratory, samples were immediately lyophylized or stored at −80°C until further processing. Sampling sites were located in the surroundings of Zurich (Switzerland): grassland soil, meadow (municipalities of Männedorf; Uerikon; Stäfa; and Dübendorf), allotment garden (University of Zurich, Irchel campus), pasture (University of Zurich, Irchel campus; municipality of Wädenswil), forest soil (municipality of Stäfa), and aquatic sediments (river Glatt in Dübendorf). Lyophilized aliquots of approximately 1 g were transferred to an Eppendorf micro test tube (2 ml) and ground (by adding a 6 mm glass bead) in TissueLyser (Retsch, Haan, Germany) for 5 × 1 min. Elemental composition (carbon, hydrogen, nitrogen) of soil was performed with a CHN-932 elemental analyzer (Leco Corp., St. Joseph, Minn, USA). Approximately 10 mg of powdered soil was used for analysis. Composition (in % of dry soil) varied between 2.2 and 15.4, 0.2 and 1.4, and 0.2 and 2.0 for total carbon, total hydrogen, and total nitrogen, respectively. Phosphate in aqueous soil extracts (250 mg soil in 5 ml sodium acetate buffer; 0.2 M, pH 5) was determined using commercially available kits (LCK 348 and 349; Hach Lange AG, Hegnau, Switzerland). 2.3. Release of DPA from Spores 10 mg of dry spore powder was resuspended in 10 ml sodium acetate buffer (0.2 M, pH 5). Spores were counted under the microscope using a Neubauer counting chamber. Soil samples were thawed and 50 mg were suspended in 0.9 ml sodium acetate buffer and 0.1 ml aluminium chloride (AlCl 3 , 0.5 M). Optimal volumetric amount and concentration of aluminium chloride was determined in preliminary experiments. Samples were microwaved (Berghof Microwave Digester MWS-1, with built-in in situ infrared temperature control) in Teflon TFM screw cap digestion vessels. Temperature and power were set to 140°C and approximately 680 W (80%), respectively. Alternatively, DPA was released from spores by autoclaving the samples in screw cap glass test tubes for 15 minutes at 121°C. The presence of spores after microwaving and autoclaving was determined by microscopy. The identical DPA extraction protocol was applied for soil samples. However, microscopy was not possible due to the presence of mineral particles interfering the observation. 2.4. Fluorescence Measurement After cooling for 30 minutes, 100 μ l of the spore suspensions were mixed with 100 μ l terbium chloride solution (TbCl 3 , 30 μ M) in white 96-well microtiter plates (in 8 replicates). Fluorescence was immediately measured using a plate reader (SpectraMax M2, Bucher Biotec, Basel, Switzerland) with the following settings: time-resolved fluorescence (delay 50 μ s, interval 1200 μ s) at an excitation wavelength of 272 nm, emission wavelength of 545 nm, and 10 endpoint readings per sample at 30°C. The number of spores in the soil samples was determined using standard addition method with spores of B. subtilis [ 17 ]. Spore content was expressed as equivalents of B. subtilis . 2.1. Bacterial Spores Different Bacillus species ( B. megaterium, B. subtilis ) were cultivated in liquid medium containing (in g/l): glucose (3.6), ammonium chloride (2.5), magnesium sulfate (0.2), calcium chloride (0.07), iron sulfate (0.01), EDTA (0.01), potassium dihydrogen phosphate (0.6), dipotassium hydrogen phosphate (0.9), and yeast extract (1.0). Initial pH was adjusted to 7.0. Erlenmeyer flasks (250 ml) containing 100 ml of growth medium were inoculated and incubated for 10 to 15 days (150 rpm, 30°C). To initiate and stimulate sporulation, bacteria were subsequently transferred to a sporulation medium (identical composition, but without glucose and less ammonium chloride [only 1 g/l]). After additional 30 days of incubation—until vegetative cells were not present anymore after inspection by microscopy—spores were harvested by centrifugation, immediately frozen in liquid nitrogen followed by lyophilization. 2.2. Soil Samples Samples from different locations were collected using a stainless steel soil corer (15 mm in diameter), which was sterilized before each sampling. Cores with a maximum length of 25 cm were obtained, cut in sections of 5 cm, transferred to sterile screw cap Falcon tubes (20 ml), and stored on dry ice. After return to the laboratory, samples were immediately lyophylized or stored at −80°C until further processing. Sampling sites were located in the surroundings of Zurich (Switzerland): grassland soil, meadow (municipalities of Männedorf; Uerikon; Stäfa; and Dübendorf), allotment garden (University of Zurich, Irchel campus), pasture (University of Zurich, Irchel campus; municipality of Wädenswil), forest soil (municipality of Stäfa), and aquatic sediments (river Glatt in Dübendorf). Lyophilized aliquots of approximately 1 g were transferred to an Eppendorf micro test tube (2 ml) and ground (by adding a 6 mm glass bead) in TissueLyser (Retsch, Haan, Germany) for 5 × 1 min. Elemental composition (carbon, hydrogen, nitrogen) of soil was performed with a CHN-932 elemental analyzer (Leco Corp., St. Joseph, Minn, USA). Approximately 10 mg of powdered soil was used for analysis. Composition (in % of dry soil) varied between 2.2 and 15.4, 0.2 and 1.4, and 0.2 and 2.0 for total carbon, total hydrogen, and total nitrogen, respectively. Phosphate in aqueous soil extracts (250 mg soil in 5 ml sodium acetate buffer; 0.2 M, pH 5) was determined using commercially available kits (LCK 348 and 349; Hach Lange AG, Hegnau, Switzerland). 2.3. Release of DPA from Spores 10 mg of dry spore powder was resuspended in 10 ml sodium acetate buffer (0.2 M, pH 5). Spores were counted under the microscope using a Neubauer counting chamber. Soil samples were thawed and 50 mg were suspended in 0.9 ml sodium acetate buffer and 0.1 ml aluminium chloride (AlCl 3 , 0.5 M). Optimal volumetric amount and concentration of aluminium chloride was determined in preliminary experiments. Samples were microwaved (Berghof Microwave Digester MWS-1, with built-in in situ infrared temperature control) in Teflon TFM screw cap digestion vessels. Temperature and power were set to 140°C and approximately 680 W (80%), respectively. Alternatively, DPA was released from spores by autoclaving the samples in screw cap glass test tubes for 15 minutes at 121°C. The presence of spores after microwaving and autoclaving was determined by microscopy. The identical DPA extraction protocol was applied for soil samples. However, microscopy was not possible due to the presence of mineral particles interfering the observation. 2.4. Fluorescence Measurement After cooling for 30 minutes, 100 μ l of the spore suspensions were mixed with 100 μ l terbium chloride solution (TbCl 3 , 30 μ M) in white 96-well microtiter plates (in 8 replicates). Fluorescence was immediately measured using a plate reader (SpectraMax M2, Bucher Biotec, Basel, Switzerland) with the following settings: time-resolved fluorescence (delay 50 μ s, interval 1200 μ s) at an excitation wavelength of 272 nm, emission wavelength of 545 nm, and 10 endpoint readings per sample at 30°C. The number of spores in the soil samples was determined using standard addition method with spores of B. subtilis [ 17 ]. Spore content was expressed as equivalents of B. subtilis . 3. Results and Discussion Microwave treatment of spore suspensions and soil samples led to a fast release of DPA ( Figure 1 ). Within two minutes, maximum release was obtained. Increased treatment time did not improve DPA mobilization. Bacterial spore content was related to soil type ( Figure 2 ). Highest spore numbers up to 4  × 10 8 spores per gram dry soil were found in agriculturally used land (meadow, pasture), less in forest soil. Fluvial sediments showed lowest spore numbers. The interference of different compounds present in soil (e.g., phosphate) might lead to quenching of the fluorescence signal. This drawback has been overcome by the addition of aluminium chloride as already shown for the determination of bacterial spores in aquatic sediments [ 18 ]. Concentration of ortho -phosphate in soil extracts (22.5 μ M) was reduced by the addition of aluminium chloride to concentrations below the detection limit (20 only low spore counts (0.5 × 10 8 spores per gram of dry soil) were detected as compared to C/N ratios <20. It has been demonstrated in pure cultures of Bacillus thuringiensis in a stirred bioreactor that low carbon-to-nitrogen ratios of 4 : 1 resulted in high spore counts [ 19 ]. In contrast however, spore formation in Streptomyces coelicolor was stimulated under nitrogen-limiting conditions [ 20 ]. In particular, C/N ratios between 50 and 100 promoted sporulation, whereas C/N rations <40 did not allow spore formation. Our results showed that in soils with extremely high C/N ratios, spore content was low. The importance of C/N ratio was stressed by Gao and coworkers regarding the sporulation of fungi, although fungal spores do not contain DPA [ 21 ]. A carbon-to-nitrogen (C/N) ratio of 20 stimulated spore formation by fungi such as Penicillium camembertii [ 22 ]. The fungus Colletotrichum coccodes produced highest spore counts at a C/N ratio of 5 to 10, whereas at a ratio of 40, spore formation was significantly lower [ 23 ]. Similarly, in Plectosporium tabacinum optimal spore formation was found when C/N rations were between 5 and 10 [ 24 ]. The distribution of spores in marine sediments (determined as DPA) showed only a low correlation with the content of total organic carbon and varied with the sediment type [ 18 ]. Highest numbers have been found in organic-rich black sediments, lowest number in sandy sediments. It was hypothesized from anthrax outbreaks, that the high numbers of Bacillus spores might be related to soils rich in organic matter that is, to a high C/N ratio [ 25 ]. These soil environmental conditions are suggested to support the presence and viability of B. anthracis spores [ 26 ]. However, we could not confirm this hypothesis. Depth distribution of spores from an area currently used as allotment garden (cultivation of flowers and vegetables) showed highest numbers in a horizon of 5 to 10 cm ( Figure 5 ). The two methods evaluated (microwaving, autoclaving) for the mobilization of DPA from bacterial spores gave similar results. However, microwaving was less time-consuming, whereas autoclaving allowed faster throughput of samples. 4. Conclusions In summary, microwave treatment of soil samples followed by the measurement of fluorescence after addition of terbium proved to be a fast and easy method to assess the content of bacterial spores. Our study might provide a basis for the detection of "hot spots" of endospores in soil.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6060349/

Updates to Clostridium difficile Spore Germination

Germination of Clostridium difficile spores is a crucial early requirement for colonization of the gastrointestinal tract. Likewise, C. difficile cannot cause disease pathologies unless its spores germinate into metabolically active, toxin-producing cells.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5491120/

Combination of poly I:C and Pam3CSK4 enhances activation of B cells in vitro and boosts antibody responses to protein vaccines in vivo

Vaccines that can rapidly induce strong and robust antibody-mediated immunity could improve protection from certain infectious diseases for which current vaccine formulations are inefficient. For indications such as anthrax and influenza, antibody production in vivo is a correlate of efficacy. Toll-like receptor (TLR) agonists are frequently studied for their role as vaccine adjuvants, largely because of their ability to enhance initiation of immune responses to antigens by activating dendritic cells. However, TLRs are also expressed on B cells and may contribute to effective B cell activation and promote differentiation into antigen-specific antibody producing plasma cells in vivo . We sought to discover an adjuvant system that could be used to augment antibody responses to influenza and anthrax vaccines. We first characterized an adjuvant system in vitro which consisted of two TLR ligands, poly I:C (TLR3) and Pam3CSK4 (TLR2), by evaluating its effects on B cell activation. Each agonist enhanced B cell activation through increased expression of surface receptors, cytokine secretion and proliferation. However, when B cells were stimulated with poly I:C and Pam3CSK4 in combination, further enhancement to cell activation was observed. Using B cells isolated from knockout mice we confirmed that poly I:C and Pam3CSK4 were signaling through TLR3 and TLR2, respectively. B cells activated with Poly I:C and Pam3CSK4 displayed enhanced capacity to stimulate allogeneic CD4 + T cell activation and differentiate into antibody-producing plasma cells in vitro . Mice vaccinated with influenza or anthrax antigens formulated with poly I:C and Pam3CSK4 in DepoVax ™ vaccine platform developed a rapid and strong antigen-specific serum antibody titer that persisted for at least 12 weeks after a single immunization. These results demonstrate that combinations of TLR adjuvants promote more effective B cell activation in vitro and can be used to augment antibody responses to vaccines in vivo . Introduction Pre-existing antibodies offer the best protection against infection, and many vaccines are available that can effectively mitigate risk of serious infection through prophylactic immunizations. However, for some indications current vaccine formulations do not provide adequate protection. Hence, there is an urgent need to develop novel vaccine technologies to meet these needs. Influenza and anthrax are two examples of indications for which improved and/ or new vaccines are required. Although these indications are quite different in terms of pathogenesis and antigen type, antibody production is the correlate of a protective immune response for both. Therefore, an optimal vaccine for these infectious diseases would induce a rapid and long lasting antibody mediated response with minimal immunizations [ 1 , 2 ]. A vaccine platform called DepoVax ™ (DPX) is a unique water-free formulation that generates strong, long lasting immune responses. DPX is a lipid-in-oil formulation that can incorporate a variety of different peptide or protein antigens. Upon injection, DPX creates a depot that facilitates active uptake by antigen-presenting cells [ 3 ]. DPX is a fully characterized formulation, and DPX formulations containing MHC class I restricted peptides have demonstrated robust cellular immune responses when tested in phase I clinical oncology indications [ 4 , 5 ]. In preclinical models, DPX formulations have shown enhanced antibody mediated immune responses and have supported an ongoing phase I trial of a DPX vaccine for respiratory syncytial virus. The ability of DPX to generate rapid and long-lasting immune responses following a single immunization makes it particularly suitable for development of influenza and anthrax vaccines [ 6 , 7 ]. The type of immune response elicited by DPX can be tailored by changing the adjuvants used in the formulation. Adjuvants influence the type and strength of immune response towards vaccination primarily through activating the innate immune response, which in turn activates the adaptive immune response [ 8 – 10 ]. Toll-like receptors (TLRs) are primarily found on innate immune cells and are common targets of adjuvants [ 11 ]; TLRs are also expressed by naïve B cells and TLR stimulation can influence their differentiation into antibody-secreting plasma cells (ASCs) [ 12 ]. Unlike T cells, B cells recognize intact soluble antigen without requiring MHC presentation [ 13 , 14 ]. The transition of naïve B cells into ASC is completed when CD4 + T cells recognize cognate peptide presented in MHC II by B cells. CD4 + T cells provide co-stimulation through CD40-CD40L interaction and cytokines to induce antigen class switching. Similar to innate antigen-presenting cells, naïve B cells also express several TLRs [ 15 , 16 ]. TLR stimulation during B cell activation influences the maturation process by skewing class switch recombination, robustness of antibody production, cytokine secretion and the function of B cells as antigen presenting cells [ 17 – 20 ]. The effects of TLR stimulation on B cells has largely been investigated using LPS (stimulating TLR4) or CpG (stimulating TLR9) since receptors are the most abundant TLRs on the surface of naïve B cells [ 12 , 21 ]. However, B cells express other TLRs at lower levels which may have different effects on B cell differentiation. Indeed, combinations of two of more TLR ligands have been reported to act synergistically to enhance activation of a variety of innate immune cells in vitro [ 22 , 23 ] and boost responses to vaccination in vivo [ 24 ]. There is limited information on the effect of stimulating B cells with multiple TLR ligands in vitro and the use of TLR agonist combinations to enhance antibody responses in vivo . TLR3 and TLR1/2 are expressed at low levels on naïve B cells, yet B cells respond to stimulation to their respective ligands, poly I:C and Pam3CSK4 in vitro [ 25 , 26 ]. Poly I:C and Pam3CSK4 have been used as adjuvants to boost antibody responses to vaccines in vivo , in some cases to a greater extent than CpG [ 27 , 28 ]. Combined poly I:C and Pam3CSK4 stimulation has been reported to synergistically enhance activation of dendritic cells and macrophages in vitro [ 29 , 30 ] which may be due in part to the different signaling pathways used by each receptor [ 8 ]. In this paper we have sought to identify a unique adjuvant system that may be used to boost antibody responses to DPX vaccines targeting influenza hemagglutinin (HA) H5N1 antigen and anthrax protective antigen (PA). Therefore, we studied whether the combination of poly I:C and Pam3CSK4 stimulation could enhance activation of B cells in vitro , and whether this combination could augment antibody production to influenza and anthrax DPX vaccines in vivo . Materials and methods Animals Female BALB/c, C57BL/6 and CD-1 mice (4–6 weeks old) were obtained from Charles River Labs (St. Constant, PQ). TLR2 knockout mice had a C57BL/6 background and were a kind gift from Dr. Jean Marshall (Dalhousie University, NS, Canada). TLR3 knockout mice (B6;129S1-Tlr3tm1Flv/J) and wild type controls (B6;129SF2/J) were obtained from Jackson Laboratories (Bar Harbor, ME). Mice were housed under filter-top conditions and provided food and water ad libitum . Institutional animal care and use guidelines set by Carleton Animal Care Facility at Dalhousie University (Halifax, NS, Canada) were followed for all experiments. All experimental procedures performed during this study were approved by the Ethics Committee at Dalhousie University and strictly followed the guidelines set by The Canadian Council on Animal Care. TLR agonists Poly I:C was obtained from Thermo Fisher Scientific (Milwaukee, WS). Pam3CSK4 was obtained from EMC Microcollections (Tuebingen, Germany). Lipopolysaccharide (LPS) and CpG 1826 were purchased from Sigma-Aldrich (St. Louis, MO). All agonists were reconstituted in water at the concentration recommended by manufacturer; for cell stimulation, appropriate dilutions were prepared in complete medium. B cell purification and culture B cells were isolated from naïve mouse spleens by negative selection (Stem Cell Technologies, Vancouver, BC). The purified B cells used for all experiments were phenotypically characterized as >95% CD19 + B220 + IgD + CD23 -/low CD5 - CD11c - , indicating that they were predominantly B-2 marginal zone B cells [ 26 ] ( S1 Fig ). Purified B cells were cultured in triplicate in 96-well plates at 10 5 cells/ well in RPMI 1640 medium supplemented with 10% fetal calf serum (FCS; HyClone, Rockford, IL), 2% penicillin-streptomycin (Gibco, Burlington, ON, Canada), 50 mM mercaptoethanol (Gibco) and 2 mM L-glutamine (Gibco). Pam3CSK4 and poly I:C dilutions were made in complete medium and added into the wells containing B cells for a final volume of 200 μL per well. T-cell-dependent B cell activation was simulated by adding purified hamster anti-mouse CD40 (2.5 μg/mL; clone HM40-3, BD Biosciences, Mississauga, ON) and purified rat anti-mouse kappa Ig (1 μg/mL; clone 187.1, BD Biosciences) to the B cell suspensions. Plates were incubated at 37°C/ 5% CO 2 . Proliferation of B cells To measure proliferation, B cells were incubated for 3 days and pulsed with 0.5 μCi of tritiated thymidine ([ 3 H]-TdR; MP Biomedical, Irvine, CA) for the last 18 hours of culture. Cells were harvested onto fiberglass filter mats (Skatron Instruments, Sterling, VA) with a Titertek Cell Harvester (Skatron Instruments). [ 3 H]-TdR uptake was assessed using a Beckman LS6000IC liquid scintillation counter (Beckman Coulter Inc., Mississauga, ON) and quantified as counts per minute (CPM). The values of each experiment were taken from the average of triplicate wells. Flow cytometry For measurement of surface receptor expression, B cells were harvested after 24 hours of culture for flow cytometry. All antibodies used for flow cytometry were purchased from eBioscience (San Diego, CA) unless otherwise stated. After blocking with purified anti-CD16/CD32 (clone 93), cells were stained using cocktails of: CD19-APC (clone 1D3), CD80-FITC (clone 16-10A1), CD86-PE (clone GL1), CD25-APC (clone PC61), MHC class II-APC (clone M5/114.15.2), CD40-PE (clone 1C10), CD69-FITC (H1.2F3), CD138-APC (clone 281–2; BD Bioscience), CD267-PE (TACI; clone ebio8F10-3). Cells were acquired using a FACSCalibur ® (BD Bioscience) and data was analyzed using WinList 7.0 software (Verity Software, Topsham, ME). Quantitation of cytokine and immunoglobulin production Cytokines were detected in B cell supernatants harvested after 24 hours. Commercially available ELISA kits from eBioscience were used for IL-2, IL-4, IL-6, IL-10, IL-12p70, IFN-γ, and CXCL10. A cytokine bead array (CBA) kit from BD Bioscience was used to measure TNF-α and IL-21 production. IgG production was measured in supernatants harvested after 4 days of culture using a commercially available ELISA kit from eBioscience. Mixed lymphocyte reaction Purified B cells were stimulated with poly I:C and/ or Pam3CSK4, as well as anti-CD40 and anti-Ig, as described above. After 24 hours, B cells were harvested and treated with 50 μg/mL of mitomycin C (Sigma) for 1 hour at 37°C. B cells were then washed three times in supplemented RPMI medium, counted and resuspended at 10 5 cells/mL, 4×10 4 cells/mL and 2×10 4 cells/mL. Each dilution of B cell was added to a 96-well plate in triplicate (100 μL/ well). CD4 + T cells were purified from spleens of BALB/c mice using negative magnetic separation (Stemcell Technologies) and resuspended at 10 6 cells/mL. T cells were added to wells containing B cells (100 μL/ well) and co-cultures were incubated for 3 days. Proliferation and flow cytometric analysis was performed after 3 days of culture as described above. CD4 + T cells were identified using anti-CD4 (clone GK1.5). Vaccines and immunization DepoVax ™ vaccines were prepared as previously described [ 31 , 32 ]. Briefly, proteins and adjuvants were solubilized in appropriate buffer and mixed with S100 lipids and cholesterol (both from Lipoid GmBH, Germany). The aqueous mixture was lyophilized to a dry cake which was reconstituted with Montanide ISA51 VG (SEPPIC, France) just prior to injection. Vaccines were administered via intramuscular vaccination 25 μL on each the right and left hind legs. Influenza vaccine was prepared with recombinant hemagglutinin protein (rHA; Protein Sciences, Meriden, CT) to deliver 0.5 μg antigen, 1 μg poly I:C, and/ or 1 μg Pam3CSK4 in a 50 μL dose. Anthrax vaccine was prepared with recombinant protective antigen (rPA; List Biologicals, Campbell, CA) to deliver 1 μg antigen, 1 μg poly I:C, and/ or 1 μg Pam3CSK4 per 50 μL dose. Antibody endpoint titration Serum samples were collected via facial venupuncture at 4, 8, 12 and 16 weeks post immunization. Endpoint antibody titers were determined by ELISA. Briefly, plates were coated overnight with1 μg/mL of either rHA or rPA antigen. After blocking with 3% gelatin, serial dilutions of the serum were applied and plates were incubated overnight at 4°C. Next day, bound antibody was detected using an alkaline phosphatase-linked Protein A (Calbiochem, Gibbstown, NJ) as the secondary reagent. Endpoint titers were defined as the reciprocal of the highest dilution above the cutoff value; cutoff values were calculated using a 95% confidence interval [ 33 ]. Statistical analysis Statistical analysis was conducted with GraphPad Prism 5 software (La Jolla, CA). Data was analysed by one-way ANOVA using Tukey post-test, two-way ANOVA with Bonferroni multiple comparisons test, or Student's t-test as indicated; *p≤0.05, **p≤0.01, ***p≤0.001. For the in vitro experiments, data is shown as the average of multiple individual B cell preparations which were assayed in groups of 1–3 in independent experiments as indicated in figure legends. Animals Female BALB/c, C57BL/6 and CD-1 mice (4–6 weeks old) were obtained from Charles River Labs (St. Constant, PQ). TLR2 knockout mice had a C57BL/6 background and were a kind gift from Dr. Jean Marshall (Dalhousie University, NS, Canada). TLR3 knockout mice (B6;129S1-Tlr3tm1Flv/J) and wild type controls (B6;129SF2/J) were obtained from Jackson Laboratories (Bar Harbor, ME). Mice were housed under filter-top conditions and provided food and water ad libitum . Institutional animal care and use guidelines set by Carleton Animal Care Facility at Dalhousie University (Halifax, NS, Canada) were followed for all experiments. All experimental procedures performed during this study were approved by the Ethics Committee at Dalhousie University and strictly followed the guidelines set by The Canadian Council on Animal Care. TLR agonists Poly I:C was obtained from Thermo Fisher Scientific (Milwaukee, WS). Pam3CSK4 was obtained from EMC Microcollections (Tuebingen, Germany). Lipopolysaccharide (LPS) and CpG 1826 were purchased from Sigma-Aldrich (St. Louis, MO). All agonists were reconstituted in water at the concentration recommended by manufacturer; for cell stimulation, appropriate dilutions were prepared in complete medium. B cell purification and culture B cells were isolated from naïve mouse spleens by negative selection (Stem Cell Technologies, Vancouver, BC). The purified B cells used for all experiments were phenotypically characterized as >95% CD19 + B220 + IgD + CD23 -/low CD5 - CD11c - , indicating that they were predominantly B-2 marginal zone B cells [ 26 ] ( S1 Fig ). Purified B cells were cultured in triplicate in 96-well plates at 10 5 cells/ well in RPMI 1640 medium supplemented with 10% fetal calf serum (FCS; HyClone, Rockford, IL), 2% penicillin-streptomycin (Gibco, Burlington, ON, Canada), 50 mM mercaptoethanol (Gibco) and 2 mM L-glutamine (Gibco). Pam3CSK4 and poly I:C dilutions were made in complete medium and added into the wells containing B cells for a final volume of 200 μL per well. T-cell-dependent B cell activation was simulated by adding purified hamster anti-mouse CD40 (2.5 μg/mL; clone HM40-3, BD Biosciences, Mississauga, ON) and purified rat anti-mouse kappa Ig (1 μg/mL; clone 187.1, BD Biosciences) to the B cell suspensions. Plates were incubated at 37°C/ 5% CO 2 . Proliferation of B cells To measure proliferation, B cells were incubated for 3 days and pulsed with 0.5 μCi of tritiated thymidine ([ 3 H]-TdR; MP Biomedical, Irvine, CA) for the last 18 hours of culture. Cells were harvested onto fiberglass filter mats (Skatron Instruments, Sterling, VA) with a Titertek Cell Harvester (Skatron Instruments). [ 3 H]-TdR uptake was assessed using a Beckman LS6000IC liquid scintillation counter (Beckman Coulter Inc., Mississauga, ON) and quantified as counts per minute (CPM). The values of each experiment were taken from the average of triplicate wells. Flow cytometry For measurement of surface receptor expression, B cells were harvested after 24 hours of culture for flow cytometry. All antibodies used for flow cytometry were purchased from eBioscience (San Diego, CA) unless otherwise stated. After blocking with purified anti-CD16/CD32 (clone 93), cells were stained using cocktails of: CD19-APC (clone 1D3), CD80-FITC (clone 16-10A1), CD86-PE (clone GL1), CD25-APC (clone PC61), MHC class II-APC (clone M5/114.15.2), CD40-PE (clone 1C10), CD69-FITC (H1.2F3), CD138-APC (clone 281–2; BD Bioscience), CD267-PE (TACI; clone ebio8F10-3). Cells were acquired using a FACSCalibur ® (BD Bioscience) and data was analyzed using WinList 7.0 software (Verity Software, Topsham, ME). Quantitation of cytokine and immunoglobulin production Cytokines were detected in B cell supernatants harvested after 24 hours. Commercially available ELISA kits from eBioscience were used for IL-2, IL-4, IL-6, IL-10, IL-12p70, IFN-γ, and CXCL10. A cytokine bead array (CBA) kit from BD Bioscience was used to measure TNF-α and IL-21 production. IgG production was measured in supernatants harvested after 4 days of culture using a commercially available ELISA kit from eBioscience. Mixed lymphocyte reaction Purified B cells were stimulated with poly I:C and/ or Pam3CSK4, as well as anti-CD40 and anti-Ig, as described above. After 24 hours, B cells were harvested and treated with 50 μg/mL of mitomycin C (Sigma) for 1 hour at 37°C. B cells were then washed three times in supplemented RPMI medium, counted and resuspended at 10 5 cells/mL, 4×10 4 cells/mL and 2×10 4 cells/mL. Each dilution of B cell was added to a 96-well plate in triplicate (100 μL/ well). CD4 + T cells were purified from spleens of BALB/c mice using negative magnetic separation (Stemcell Technologies) and resuspended at 10 6 cells/mL. T cells were added to wells containing B cells (100 μL/ well) and co-cultures were incubated for 3 days. Proliferation and flow cytometric analysis was performed after 3 days of culture as described above. CD4 + T cells were identified using anti-CD4 (clone GK1.5). Vaccines and immunization DepoVax ™ vaccines were prepared as previously described [ 31 , 32 ]. Briefly, proteins and adjuvants were solubilized in appropriate buffer and mixed with S100 lipids and cholesterol (both from Lipoid GmBH, Germany). The aqueous mixture was lyophilized to a dry cake which was reconstituted with Montanide ISA51 VG (SEPPIC, France) just prior to injection. Vaccines were administered via intramuscular vaccination 25 μL on each the right and left hind legs. Influenza vaccine was prepared with recombinant hemagglutinin protein (rHA; Protein Sciences, Meriden, CT) to deliver 0.5 μg antigen, 1 μg poly I:C, and/ or 1 μg Pam3CSK4 in a 50 μL dose. Anthrax vaccine was prepared with recombinant protective antigen (rPA; List Biologicals, Campbell, CA) to deliver 1 μg antigen, 1 μg poly I:C, and/ or 1 μg Pam3CSK4 per 50 μL dose. Antibody endpoint titration Serum samples were collected via facial venupuncture at 4, 8, 12 and 16 weeks post immunization. Endpoint antibody titers were determined by ELISA. Briefly, plates were coated overnight with1 μg/mL of either rHA or rPA antigen. After blocking with 3% gelatin, serial dilutions of the serum were applied and plates were incubated overnight at 4°C. Next day, bound antibody was detected using an alkaline phosphatase-linked Protein A (Calbiochem, Gibbstown, NJ) as the secondary reagent. Endpoint titers were defined as the reciprocal of the highest dilution above the cutoff value; cutoff values were calculated using a 95% confidence interval [ 33 ]. Statistical analysis Statistical analysis was conducted with GraphPad Prism 5 software (La Jolla, CA). Data was analysed by one-way ANOVA using Tukey post-test, two-way ANOVA with Bonferroni multiple comparisons test, or Student's t-test as indicated; *p≤0.05, **p≤0.01, ***p≤0.001. For the in vitro experiments, data is shown as the average of multiple individual B cell preparations which were assayed in groups of 1–3 in independent experiments as indicated in figure legends. Results Poly I:C and Pam3CSK4 promote B cell activation independent of T cell help Different TLR agonists have been reported to selectively increase activation markers on B cells after 24 hour stimulation [ 25 , 26 ]. In a preliminary experiment, we dosed poly I:C and Pam3CSK4 separately ( S3 Fig ) and in combination ( S4 Fig ) to identify the optimal dose of each agonist that could result in highest combined effect. Poly I:C was tested in a range of 0.1 to 50 ug/mL and Pam3CSK4 in a range of 0.01 to 10 ug/mL. Activation was assessed by expression of receptors associated with B cell activation, CD80 and CD40. We selected the dose of 25 μg/mL for poly I:C and 1 μg/mL for Pam3CSK4 as having a suboptimal effect on the induction of these receptors when used individually, but resulted in very high expression when used in combination. We then tested the agonists at these doses alone and in combination to see if they could augment expression of a more comprehensive set of markers: CD80, CD86, CD25, MHC class II, CD69 and CD40 ( Fig 1 ; S3 Fig ). We found that each agonist alone induced a unique profile of receptors in the B cells. In general, poly I:C induced lower levels of receptor expression than Pam3CSK4. The receptors CD80, CD86, CD25 and CD69 were significantly increased by combination therapy compared to each poly I:C and Pam3CSK4 treatments. CD40 and MHC class II were substantially increased by Pam3CSK4 alone treatment, which was not augmented further by combination treatment. 10.1371/journal.pone.0180073.g001 Fig 1 Poly I:C and Pam3CSK4 stimulate expression of surface receptors in B cells independent of T cell help. CD19 + B cells were purified from spleens of C57BL6 mice and stimulated with poly I:C (25 μg/mL), Pam3CSK4 (1 μg/mL), or the combination of both. Expression of surface receptor was determined after 24 hours by flow cytometry. (A) CD80 (n = 11), (B) CD86 (n = 7), (C) CD25 (n = 6), (D) MHC class II (n = 5), (E) CD69 (n = 4), (F) CD40 (n = 6). Data are shown as average ± SEM of individual B cell preparations as indicated, data is pooled from two to four independent experiments. Statistics performed by ANOVA with Tukey post-test: "#" indicates significance relative to untreated, "+" indicates significance relative to poly I:C and "*" indicates significance relative to Pam3CSK4. Poly I:C and Pam3CSK4 enhance T-cell-dependent activation of B cells TLR stimulation has been reported to synergize with T-cell-dependent B cell activation in vitro [ 15 , 16 ], and is more representative of B cell activation in vivo . Therefore, we tested the effects of poly I:C and/ or Pam3CSK4 stimulation of B cells activated in a T-cell-dependent manner by simulating T-cell-dependent activation with the addition of anti-Ig and anti-CD40 into cultures. After 24 hour incubation, we looked at the expression of surface receptors. The T-cell-dependent activation alone induced higher baseline expression of CD86, MHC class II and CD69 ( Fig 2B, 2D and 2E ), but no significant increase in CD80 or CD25 ( Fig 2A and 2C ). T-cell-dependent activation augmented the expression of CD25 and CD69 in response to stimulation with poly I:C or Pam3CSK4 alone, and overall the combination of both adjuvants resulted in expression levels similar to that achieved without T-cell-dependent activation ( Fig 1 ).T-cell-dependent T-cell-dependent The effects on CD40 expression could not be reliably detected due to the blocking effect of the anti-CD40 antibody used for T-cell-dependent activation. 10.1371/journal.pone.0180073.g002 Fig 2 Poly I:C and Pam3CSK4 enhance T-cell-dependent B cell activation. Purified B cells from C57BL/6 mice were stimulated with poly I:C (25 μg/mL), Pam3CSK4 (1 μg/mL) or a combination of both, with T-cell-dependent co-stimulation provided by anti-CD40 and anti-Ig. Expression of surface receptor was determined after 24 hours by flow cytometry. (A) CD80 (n = 8), (B) CD86 (n = 5), (C) CD25 (n = 6), (D) MHC class II (n = 7) and (E) CD69 (n = 3). Data are shown as average ± SEM of individual B cell preparations as indicated, data is pooled from two to four independent experiments. Statistics performed by ANOVA with Tukey post-test: "#" indicates significance relative to untreated, "+" indicates significance relative to poly I:C and "*" indicates significance relative to Pam3CSK4. Stimulation of B cells with TLR agonists has been reported to increase production of several cytokines [ 25 , 34 ]. We screened for production of cytokines important to B cell survival and proliferation: IL-6, IL-10, IL-21 and TNFα; Th1 cytokines: IL-12p70 and IFN-γ; and Th2 cytokine: IL-4. Since poly I:C is also known to induce CXCL10 production in other cell types [ 35 ] and CpG stimulation of B cells induces CXCL10 [ 36 ], we also tested for production of the chemokine CXCL10. We could only detect production of IL-6, TNFα and CXCL10 ( Fig 3 ) as the remaining cytokines were below the limits of detection. The combination of poly I:C and Pam3CSK4 significantly enhanced production of IL-6 and TNFα above no agonist or single agonist stimulation ( Fig 3A and 3B ). Poly I:C alone induced significant production of the chemokine CXCL10, and in combination with Pam3CSK4 this level was not increased further ( Fig 3C ). 10.1371/journal.pone.0180073.g003 Fig 3 Cytokine production enhanced by poly I:C and Pam3CSK4 stimulated B cells. Purified B cells from C57BL/6 mice were stimulated with poly I:C (25 μg/mL), Pam3CSK4 (1 μg/mL) or a combination of both, with T-cell-dependent co-stimulation provided by anti-CD40 and anti-Ig. Supernatants were harvested after 24 hours and levels of (A) IL-6 (n = 6), (B) TNF-α (n = 6) and (C) CXCL10 (n = 8) determined by ELISA or CBA analysis. Data are shown as average ± SEM of individual B cell preparations as indicated, data is pooled from three to five independent experiments. Statistics performed by ANOVA with Tukey post-test: "#" indicates significance relative to untreated, "+" indicates significance relative to poly I:C and "*" indicates significance relative to Pam3CSK4. Finally, we measured how the proliferation of B cells with T-cell-dependent activation was affected by the agonists ( Fig 4 ). After 3 days of culture, Pam3CSK4 induced a significant increase in B cell proliferation alone, which was not increased further with poly I:C combination. 10.1371/journal.pone.0180073.g004 Fig 4 Proliferative response of B cells stimulated with poly I:C and Pam3CSK. Purified B cells from C57BL/6 mice were stimulated with poly I:C (25 μg/mL), Pam3CSK4 (1 μg/mL) or a combination of both, with T-cell-dependent co-stimulation provided by anti-CD40 and anti-Ig. Proliferation was measured after 3 days by [ 3 H]-TdR uptake. Data are shown as average ± SEM of 5 individual B cell preparations, data is pooled from at two independent experiments. Statistics performed by ANOVA with Tukey post-test: "#" indicates significance relative to untreated, "+" indicates significance relative to poly I:C and "*" indicates significance relative to Pam3CSK4. TLR3 and TLR2 are essential to optimal B cell activation by poly I:C and Pam3CSK4 To confirm that poly I:C and Pam3CSK4 were mediating effects through the TLR pathway, we performed experiments using B cells isolated from TLR3 and TLR2 knockout mice. This was important in particular for poly I:C since two other receptors have been reported to recognize this agonist, MDA-5 and RIG-I [ 9 ]. It was also of interest to us to see if these effects may be due to an alternate receptor, given that TLR expression on B cells has been reported to be low [ 25 , 26 ]. Using methods already described, we evaluated stimulation of B cells isolated from TLR2-/- and TLR3-/- knockout mice with poly I:C and Pam3CSK4 ( Fig 5 ). Wild-type B cells were stimulated in parallel. 10.1371/journal.pone.0180073.g005 Fig 5 TLR3 and TLR2 Mediate B cell Response to poly I:C and Pam3CSK4 respectively. B cells purified from TLR2 -/- (A, C, E) or TLR3 -/- (B, D, F) mice were stimulated with poly I:C (25 μg/mL), Pam3CSK4 (1 μg/mL), the combination of poly I:C and Pam3CSK4, LPS (10 μg/mL) or CpG (25 μg/mL), with T-cell-dependent co-stimulation provided by anti-CD40 and anti-Ig. Relevant wild-type controls were tested in parallel (C57BL/6 for TLR2 -/- , B6;129SF2/J for TLR3 -/- ). (A, B) Expression of CD80 was measured by flow cytometry after 24 hours. (C, D) Proliferation was measured by [ 3 H]-TdR uptake after 3 days. (E, F) Production of TNF-α was measured by ELISA the supernatants harvested at 24 hours. Results are shown as the average ± SEM; TLR2 -/- is from 4 individual B cell preparations pooled from two independent experiments; TLR3 -/- is from 5 individual B cell preparations pooled from three independent experiments. Statistics compare the response of knockout B cells to corresponding wild type B cells and were calculated by 2-way ANOVA with Bonferroni post-test, *p<0.05, ***p<0.001. In comparison to wild type B cells, CD80 expression on TLR2 -/- B cells was significantly reduced by stimulation with Pam3CSK4 alone and the combination of poly I:C and Pam3CSK4 ( Fig 5A ). TLR3 -/- B cells also had significant reduction in CD80 expression after combination stimulation ( Fig 5B ). Proliferation was significantly reduced to the Pam3CSK4 alone and combination stimulation in TLR2 -/- B cells ( Fig 5C ). There were no significant differences in proliferation detected in the TLR3 -/- B cells in response to any stimulation ( Fig 5D ). TNFα production, which was significantly increased in wild type B cells after combination poly I:C and Pam3CSK4 stimulation, was reduced in both the TLR2 -/- and TLR3 -/- B cells ( Fig 5E and 5F ). Similar results were also found for other surface receptors and cytokines ( S6 and S7 Figs). For positive controls we used LPS, which signals through TLR4, and CpG, which signals through TLR9 ( Fig 5C ). TLR4 is the only other TLR besides TLR3 which utilizes the TRIF pathway, although only partially. Although both knockout B cells responded to LPS similar to the wild type, there was a significant drop in proliferation in the TLR2 -/- B cells, however TLR2 has been reported to enhance responses to LPS [ 37 ]. TLR9 is an intracellular receptor that is highly expressed by murine B cells, there were no differences in response between wild type and knockout B cells tested. Poly I:C and Pam3CSK4 enhance B cell induced activation of CD4 + T cells Stimulation of B cells with poly I:C and Pam3CSK4 resulted in increased expression of several co-stimulatory molecules involved in T cell activation, such as CD80, CD86, CD40 and MHC class II. Therefore, we tested whether poly I:C and Pam3CSK4 activated B cells could more efficiently stimulate CD4 + T cell proliferation using an allogeneic model. We found that B cells activated with poly I:C and Pam3CSK4 were most efficient at inducing CD4 + T cell proliferation ( Fig 6A ). CD4 + T cell activation was confirmed by increased levels of IL-2 detected in co-culture supernatants ( Fig 6B ) and expression of high affinity IL-2 receptor CD25 on CD4 + T cells ( Fig 6C ). 10.1371/journal.pone.0180073.g006 Fig 6 Allogeneic T cell response is increased in B cells stimulated with poly I:C and Pam3CSK4. Purified B cells from C57BL6 mice and stimulated with poly I:C (25 μg/mL), Pam3CSK4 (1 μg/mL) or a combination of both with T-cell-dependent co-stimulation provided by anti-CD40 and anti-Ig. After 24 hours, B cells were inactivated by mitomycin C, and resuspended at various concentrations. B cells were co-cultured with 100,000 CD4 + T cells isolated from BALB/c mice at ratios 1:10 (10,000 B cells), 1:25 (4,000 B cells) or 1:50 (2,000 B cells). (A) Proliferation was measured after 3 days by [ 3 H]-TdR incorporation. Separate co-cultures were setup in parallel to detect IL-2 production (B) in supernatant as well as expression of CD25 (C) on CD4 + T cells after 3 days. Results are shown as the average ± SEM of 5 individual B cell preparations from 5 independent experiments. Statistics performed by ANOVA with Tukey post-test: "#" indicates significance relative to untreated. B cell differentiation into antibody producing plasma cells is enhanced by poly I:C and Pam3CSK4 To determine if poly I:C and PamCSK4 could promote B cell differentiation into ASC, we looked for markers of plasma cell differentiation on B cells that had been stimulated for 24 hours with poly I:C and/ or Pam3CSK4. The combination treatment resulted in the highest levels of surface receptors associated with plasma cell differentiation, CD138 and TACI (CD267) ( Fig 7A and 7B ). Supernatants harvested after 4 days of culture also had high levels of IgG present in the B cells stimulated with the combination, indicating polyclonal activation of B cells. Notably, the levels of IgG in the supernatants of B cells stimulated with poly I:C and/ or Pam3CSK4 was below the limits of detection. 10.1371/journal.pone.0180073.g007 Fig 7 Antibody production and plasma cell marker expression are increased on B cells following T-cell-dependent stimulation with poly I:C and Pam3CSK4. Purified B cells from C57BL/6 mice were stimulated with poly I:C (25 μg/mL), Pam3CSK4 (1 μg/mL) or a combination of both with T-cell-dependent co-stimulation provided by anti-CD40 and anti-Ig. Surface expression of (A) CD138 (n = 7) and (B) TACI (n = 4) were determined by flow cytometry after 24 hours incubation. IgG was detected in supernatants harvested after 4 days of incubation by ELISA (n = 5). Results are shown as the average ± SEM of individual B cell preparations as indicated, data is pooled from two to three independent experiments. Statistics performed by ANOVA with Tukey post-test: "#" indicates significance relative to untreated, "+" indicates significance relative to poly I:C and "*" indicates significance relative to Pam3CSK4. Protein vaccines adjuvanted with poly I:C and Pam3CSK4 produce highest levels of antibodies in vivo Having demonstrated that poly I:C and Pam3CSK4 could augment T-cell-dependent B cell activation in vitro , resulting in enhanced function and differentiation into ASC, we sought to determine if these agonists could be used to boost antibody response to vaccination in vivo . We evaluated the poly I:C/ Pam3CSK4 adjuvant system using two different antigens: recombinant hemagglutinin (rHA) H5N1 for influenza and recombinant protective antigen (rPA) for anthrax. The vaccines were formulated using the DPX platform and contained no adjuvant, poly I:C at 1 μg dose, Pam3CSK4 at 1 μg dose or the combination of both adjuvants. The adjuvant dose was selected based on preliminary in vivo dose-response screening in which each adjuvant was separately dosed in DPX with a rPA antigen ( S8 Fig ). We choose the minimal doses that enhanced immune responses over non-adjuvanted vaccine. These experiments were conducted in the outbred CD-1 mouse strain in order to increase the translational relevancy of the findings. Naïve CD-1 mice received a single vaccination and antigen-specific antibody titers were monitored in the serum every four weeks post immunization by ELISA using Protein A for detection. Endpoint titers for each group are shown in Fig 8 . For both the influenza and anthrax vaccine, the combination of poly I:C and Pam3CSK4 resulted in rapid induction of antibodies that were at significantly higher levels than the non-adjuvanted vaccine or vaccines with containing single adjuvants. 10.1371/journal.pone.0180073.g008 Fig 8 Poly I:C and Pam3CSK4 adjuvant combination enhance antibody production by influenza and anthrax vaccines in vivo . (A) CD-1 mice (n = 8) were vaccinated once with influenza recombinant hemagglutinin antigen (rHA; 0.5ug) formulated in DepoVax with no adjuvant, poly I:C (1 μg), Pam3CSK4 (1 μg) or the combination of both. (B) CD-1 mice (n = 8) were vaccinated once with anthrax recombinant protective antigen (rPA; 1 μg) formulated in DepoVax with no adjuvant, poly I:C (1 μg), Pam3CSK4 (1 μg), or the combination of both. For A and B, mice were bled on the indicated weeks after immunization and antigen-specific antibodies detected in serum by direct ELISA. Results are shown as endpoint titre ± SEM and are each representative of two independent experiments. Statistics by 2-way ANOVA with Bonferroni post-test comparing the combination to: "#" compared to no adjuvant, "+" compared to poly I:C; "*" compared to Pam3CSK4. No differences were detected between other groups. Poly I:C and Pam3CSK4 promote B cell activation independent of T cell help Different TLR agonists have been reported to selectively increase activation markers on B cells after 24 hour stimulation [ 25 , 26 ]. In a preliminary experiment, we dosed poly I:C and Pam3CSK4 separately ( S3 Fig ) and in combination ( S4 Fig ) to identify the optimal dose of each agonist that could result in highest combined effect. Poly I:C was tested in a range of 0.1 to 50 ug/mL and Pam3CSK4 in a range of 0.01 to 10 ug/mL. Activation was assessed by expression of receptors associated with B cell activation, CD80 and CD40. We selected the dose of 25 μg/mL for poly I:C and 1 μg/mL for Pam3CSK4 as having a suboptimal effect on the induction of these receptors when used individually, but resulted in very high expression when used in combination. We then tested the agonists at these doses alone and in combination to see if they could augment expression of a more comprehensive set of markers: CD80, CD86, CD25, MHC class II, CD69 and CD40 ( Fig 1 ; S3 Fig ). We found that each agonist alone induced a unique profile of receptors in the B cells. In general, poly I:C induced lower levels of receptor expression than Pam3CSK4. The receptors CD80, CD86, CD25 and CD69 were significantly increased by combination therapy compared to each poly I:C and Pam3CSK4 treatments. CD40 and MHC class II were substantially increased by Pam3CSK4 alone treatment, which was not augmented further by combination treatment. 10.1371/journal.pone.0180073.g001 Fig 1 Poly I:C and Pam3CSK4 stimulate expression of surface receptors in B cells independent of T cell help. CD19 + B cells were purified from spleens of C57BL6 mice and stimulated with poly I:C (25 μg/mL), Pam3CSK4 (1 μg/mL), or the combination of both. Expression of surface receptor was determined after 24 hours by flow cytometry. (A) CD80 (n = 11), (B) CD86 (n = 7), (C) CD25 (n = 6), (D) MHC class II (n = 5), (E) CD69 (n = 4), (F) CD40 (n = 6). Data are shown as average ± SEM of individual B cell preparations as indicated, data is pooled from two to four independent experiments. Statistics performed by ANOVA with Tukey post-test: "#" indicates significance relative to untreated, "+" indicates significance relative to poly I:C and "*" indicates significance relative to Pam3CSK4. Poly I:C and Pam3CSK4 enhance T-cell-dependent activation of B cells TLR stimulation has been reported to synergize with T-cell-dependent B cell activation in vitro [ 15 , 16 ], and is more representative of B cell activation in vivo . Therefore, we tested the effects of poly I:C and/ or Pam3CSK4 stimulation of B cells activated in a T-cell-dependent manner by simulating T-cell-dependent activation with the addition of anti-Ig and anti-CD40 into cultures. After 24 hour incubation, we looked at the expression of surface receptors. The T-cell-dependent activation alone induced higher baseline expression of CD86, MHC class II and CD69 ( Fig 2B, 2D and 2E ), but no significant increase in CD80 or CD25 ( Fig 2A and 2C ). T-cell-dependent activation augmented the expression of CD25 and CD69 in response to stimulation with poly I:C or Pam3CSK4 alone, and overall the combination of both adjuvants resulted in expression levels similar to that achieved without T-cell-dependent activation ( Fig 1 ).T-cell-dependent T-cell-dependent The effects on CD40 expression could not be reliably detected due to the blocking effect of the anti-CD40 antibody used for T-cell-dependent activation. 10.1371/journal.pone.0180073.g002 Fig 2 Poly I:C and Pam3CSK4 enhance T-cell-dependent B cell activation. Purified B cells from C57BL/6 mice were stimulated with poly I:C (25 μg/mL), Pam3CSK4 (1 μg/mL) or a combination of both, with T-cell-dependent co-stimulation provided by anti-CD40 and anti-Ig. Expression of surface receptor was determined after 24 hours by flow cytometry. (A) CD80 (n = 8), (B) CD86 (n = 5), (C) CD25 (n = 6), (D) MHC class II (n = 7) and (E) CD69 (n = 3). Data are shown as average ± SEM of individual B cell preparations as indicated, data is pooled from two to four independent experiments. Statistics performed by ANOVA with Tukey post-test: "#" indicates significance relative to untreated, "+" indicates significance relative to poly I:C and "*" indicates significance relative to Pam3CSK4. Stimulation of B cells with TLR agonists has been reported to increase production of several cytokines [ 25 , 34 ]. We screened for production of cytokines important to B cell survival and proliferation: IL-6, IL-10, IL-21 and TNFα; Th1 cytokines: IL-12p70 and IFN-γ; and Th2 cytokine: IL-4. Since poly I:C is also known to induce CXCL10 production in other cell types [ 35 ] and CpG stimulation of B cells induces CXCL10 [ 36 ], we also tested for production of the chemokine CXCL10. We could only detect production of IL-6, TNFα and CXCL10 ( Fig 3 ) as the remaining cytokines were below the limits of detection. The combination of poly I:C and Pam3CSK4 significantly enhanced production of IL-6 and TNFα above no agonist or single agonist stimulation ( Fig 3A and 3B ). Poly I:C alone induced significant production of the chemokine CXCL10, and in combination with Pam3CSK4 this level was not increased further ( Fig 3C ). 10.1371/journal.pone.0180073.g003 Fig 3 Cytokine production enhanced by poly I:C and Pam3CSK4 stimulated B cells. Purified B cells from C57BL/6 mice were stimulated with poly I:C (25 μg/mL), Pam3CSK4 (1 μg/mL) or a combination of both, with T-cell-dependent co-stimulation provided by anti-CD40 and anti-Ig. Supernatants were harvested after 24 hours and levels of (A) IL-6 (n = 6), (B) TNF-α (n = 6) and (C) CXCL10 (n = 8) determined by ELISA or CBA analysis. Data are shown as average ± SEM of individual B cell preparations as indicated, data is pooled from three to five independent experiments. Statistics performed by ANOVA with Tukey post-test: "#" indicates significance relative to untreated, "+" indicates significance relative to poly I:C and "*" indicates significance relative to Pam3CSK4. Finally, we measured how the proliferation of B cells with T-cell-dependent activation was affected by the agonists ( Fig 4 ). After 3 days of culture, Pam3CSK4 induced a significant increase in B cell proliferation alone, which was not increased further with poly I:C combination. 10.1371/journal.pone.0180073.g004 Fig 4 Proliferative response of B cells stimulated with poly I:C and Pam3CSK. Purified B cells from C57BL/6 mice were stimulated with poly I:C (25 μg/mL), Pam3CSK4 (1 μg/mL) or a combination of both, with T-cell-dependent co-stimulation provided by anti-CD40 and anti-Ig. Proliferation was measured after 3 days by [ 3 H]-TdR uptake. Data are shown as average ± SEM of 5 individual B cell preparations, data is pooled from at two independent experiments. Statistics performed by ANOVA with Tukey post-test: "#" indicates significance relative to untreated, "+" indicates significance relative to poly I:C and "*" indicates significance relative to Pam3CSK4. TLR3 and TLR2 are essential to optimal B cell activation by poly I:C and Pam3CSK4 To confirm that poly I:C and Pam3CSK4 were mediating effects through the TLR pathway, we performed experiments using B cells isolated from TLR3 and TLR2 knockout mice. This was important in particular for poly I:C since two other receptors have been reported to recognize this agonist, MDA-5 and RIG-I [ 9 ]. It was also of interest to us to see if these effects may be due to an alternate receptor, given that TLR expression on B cells has been reported to be low [ 25 , 26 ]. Using methods already described, we evaluated stimulation of B cells isolated from TLR2-/- and TLR3-/- knockout mice with poly I:C and Pam3CSK4 ( Fig 5 ). Wild-type B cells were stimulated in parallel. 10.1371/journal.pone.0180073.g005 Fig 5 TLR3 and TLR2 Mediate B cell Response to poly I:C and Pam3CSK4 respectively. B cells purified from TLR2 -/- (A, C, E) or TLR3 -/- (B, D, F) mice were stimulated with poly I:C (25 μg/mL), Pam3CSK4 (1 μg/mL), the combination of poly I:C and Pam3CSK4, LPS (10 μg/mL) or CpG (25 μg/mL), with T-cell-dependent co-stimulation provided by anti-CD40 and anti-Ig. Relevant wild-type controls were tested in parallel (C57BL/6 for TLR2 -/- , B6;129SF2/J for TLR3 -/- ). (A, B) Expression of CD80 was measured by flow cytometry after 24 hours. (C, D) Proliferation was measured by [ 3 H]-TdR uptake after 3 days. (E, F) Production of TNF-α was measured by ELISA the supernatants harvested at 24 hours. Results are shown as the average ± SEM; TLR2 -/- is from 4 individual B cell preparations pooled from two independent experiments; TLR3 -/- is from 5 individual B cell preparations pooled from three independent experiments. Statistics compare the response of knockout B cells to corresponding wild type B cells and were calculated by 2-way ANOVA with Bonferroni post-test, *p<0.05, ***p<0.001. In comparison to wild type B cells, CD80 expression on TLR2 -/- B cells was significantly reduced by stimulation with Pam3CSK4 alone and the combination of poly I:C and Pam3CSK4 ( Fig 5A ). TLR3 -/- B cells also had significant reduction in CD80 expression after combination stimulation ( Fig 5B ). Proliferation was significantly reduced to the Pam3CSK4 alone and combination stimulation in TLR2 -/- B cells ( Fig 5C ). There were no significant differences in proliferation detected in the TLR3 -/- B cells in response to any stimulation ( Fig 5D ). TNFα production, which was significantly increased in wild type B cells after combination poly I:C and Pam3CSK4 stimulation, was reduced in both the TLR2 -/- and TLR3 -/- B cells ( Fig 5E and 5F ). Similar results were also found for other surface receptors and cytokines ( S6 and S7 Figs). For positive controls we used LPS, which signals through TLR4, and CpG, which signals through TLR9 ( Fig 5C ). TLR4 is the only other TLR besides TLR3 which utilizes the TRIF pathway, although only partially. Although both knockout B cells responded to LPS similar to the wild type, there was a significant drop in proliferation in the TLR2 -/- B cells, however TLR2 has been reported to enhance responses to LPS [ 37 ]. TLR9 is an intracellular receptor that is highly expressed by murine B cells, there were no differences in response between wild type and knockout B cells tested. Poly I:C and Pam3CSK4 enhance B cell induced activation of CD4 + T cells Stimulation of B cells with poly I:C and Pam3CSK4 resulted in increased expression of several co-stimulatory molecules involved in T cell activation, such as CD80, CD86, CD40 and MHC class II. Therefore, we tested whether poly I:C and Pam3CSK4 activated B cells could more efficiently stimulate CD4 + T cell proliferation using an allogeneic model. We found that B cells activated with poly I:C and Pam3CSK4 were most efficient at inducing CD4 + T cell proliferation ( Fig 6A ). CD4 + T cell activation was confirmed by increased levels of IL-2 detected in co-culture supernatants ( Fig 6B ) and expression of high affinity IL-2 receptor CD25 on CD4 + T cells ( Fig 6C ). 10.1371/journal.pone.0180073.g006 Fig 6 Allogeneic T cell response is increased in B cells stimulated with poly I:C and Pam3CSK4. Purified B cells from C57BL6 mice and stimulated with poly I:C (25 μg/mL), Pam3CSK4 (1 μg/mL) or a combination of both with T-cell-dependent co-stimulation provided by anti-CD40 and anti-Ig. After 24 hours, B cells were inactivated by mitomycin C, and resuspended at various concentrations. B cells were co-cultured with 100,000 CD4 + T cells isolated from BALB/c mice at ratios 1:10 (10,000 B cells), 1:25 (4,000 B cells) or 1:50 (2,000 B cells). (A) Proliferation was measured after 3 days by [ 3 H]-TdR incorporation. Separate co-cultures were setup in parallel to detect IL-2 production (B) in supernatant as well as expression of CD25 (C) on CD4 + T cells after 3 days. Results are shown as the average ± SEM of 5 individual B cell preparations from 5 independent experiments. Statistics performed by ANOVA with Tukey post-test: "#" indicates significance relative to untreated. B cell differentiation into antibody producing plasma cells is enhanced by poly I:C and Pam3CSK4 To determine if poly I:C and PamCSK4 could promote B cell differentiation into ASC, we looked for markers of plasma cell differentiation on B cells that had been stimulated for 24 hours with poly I:C and/ or Pam3CSK4. The combination treatment resulted in the highest levels of surface receptors associated with plasma cell differentiation, CD138 and TACI (CD267) ( Fig 7A and 7B ). Supernatants harvested after 4 days of culture also had high levels of IgG present in the B cells stimulated with the combination, indicating polyclonal activation of B cells. Notably, the levels of IgG in the supernatants of B cells stimulated with poly I:C and/ or Pam3CSK4 was below the limits of detection. 10.1371/journal.pone.0180073.g007 Fig 7 Antibody production and plasma cell marker expression are increased on B cells following T-cell-dependent stimulation with poly I:C and Pam3CSK4. Purified B cells from C57BL/6 mice were stimulated with poly I:C (25 μg/mL), Pam3CSK4 (1 μg/mL) or a combination of both with T-cell-dependent co-stimulation provided by anti-CD40 and anti-Ig. Surface expression of (A) CD138 (n = 7) and (B) TACI (n = 4) were determined by flow cytometry after 24 hours incubation. IgG was detected in supernatants harvested after 4 days of incubation by ELISA (n = 5). Results are shown as the average ± SEM of individual B cell preparations as indicated, data is pooled from two to three independent experiments. Statistics performed by ANOVA with Tukey post-test: "#" indicates significance relative to untreated, "+" indicates significance relative to poly I:C and "*" indicates significance relative to Pam3CSK4. Protein vaccines adjuvanted with poly I:C and Pam3CSK4 produce highest levels of antibodies in vivo Having demonstrated that poly I:C and Pam3CSK4 could augment T-cell-dependent B cell activation in vitro , resulting in enhanced function and differentiation into ASC, we sought to determine if these agonists could be used to boost antibody response to vaccination in vivo . We evaluated the poly I:C/ Pam3CSK4 adjuvant system using two different antigens: recombinant hemagglutinin (rHA) H5N1 for influenza and recombinant protective antigen (rPA) for anthrax. The vaccines were formulated using the DPX platform and contained no adjuvant, poly I:C at 1 μg dose, Pam3CSK4 at 1 μg dose or the combination of both adjuvants. The adjuvant dose was selected based on preliminary in vivo dose-response screening in which each adjuvant was separately dosed in DPX with a rPA antigen ( S8 Fig ). We choose the minimal doses that enhanced immune responses over non-adjuvanted vaccine. These experiments were conducted in the outbred CD-1 mouse strain in order to increase the translational relevancy of the findings. Naïve CD-1 mice received a single vaccination and antigen-specific antibody titers were monitored in the serum every four weeks post immunization by ELISA using Protein A for detection. Endpoint titers for each group are shown in Fig 8 . For both the influenza and anthrax vaccine, the combination of poly I:C and Pam3CSK4 resulted in rapid induction of antibodies that were at significantly higher levels than the non-adjuvanted vaccine or vaccines with containing single adjuvants. 10.1371/journal.pone.0180073.g008 Fig 8 Poly I:C and Pam3CSK4 adjuvant combination enhance antibody production by influenza and anthrax vaccines in vivo . (A) CD-1 mice (n = 8) were vaccinated once with influenza recombinant hemagglutinin antigen (rHA; 0.5ug) formulated in DepoVax with no adjuvant, poly I:C (1 μg), Pam3CSK4 (1 μg) or the combination of both. (B) CD-1 mice (n = 8) were vaccinated once with anthrax recombinant protective antigen (rPA; 1 μg) formulated in DepoVax with no adjuvant, poly I:C (1 μg), Pam3CSK4 (1 μg), or the combination of both. For A and B, mice were bled on the indicated weeks after immunization and antigen-specific antibodies detected in serum by direct ELISA. Results are shown as endpoint titre ± SEM and are each representative of two independent experiments. Statistics by 2-way ANOVA with Bonferroni post-test comparing the combination to: "#" compared to no adjuvant, "+" compared to poly I:C; "*" compared to Pam3CSK4. No differences were detected between other groups. Discussion Development of novel vaccines that can induce strong and robust antibody responses with minimal immunizations is critical to providing effective protection from influenza and anthrax infection. One aspect of this development process is the identification of adjuvant systems that can boost antibody-mediated immune responses towards vaccination. In this study we identified an adjuvant system comprised of two TLR ligands, poly I:C and Pam3CSK4, that had a potent effect on B cell activation in vitro and could enhance antigen-specific antibody production towards vaccination with DPX formulated influenza and anthrax vaccines in vivo . To our knowledge, this is the first extensive characterization of the effects of poly I:C and Pam3CSK4 stimulation on B cells and documents enhanced B cell activation in response to stimulation with a combination of these agonists. This particular combination of agonists has translational relevancy as both the receptors TLR3 and TLR2 are expressed in similar levels in both mouse and human naïve B cells [ 12 ]. B cell activation in vitro was measured by changes in surface receptor expression, cytokine production (after 24 hours) and proliferation (after 3 days). Previous reports have shown that poly I:C and Pam3CSK4 can each stimulate B cell activation at low levels [ 25 , 26 ]. Our study of various doses of both poly I:C and Pam3CSK4 clearly demonstrate that these agonists can provide activation signals to B cells when used at optimal concentrations. Each agonist resulted in a characteristic activation profile of the B cells, suggesting non-redundancy in signaling pathways. Ablation of response in TLR3 -/- and TLR2 -/- B cells demonstrated that poly I:C and Pam3CSK4 were signaling through these receptors. The additive or synergistic effect on activation may be attributed to differential signaling pathways utilized by each receptor, as TLR3 signals through the adaptor protein TRIF while TLR1/2 signals through the adaptor protein MyD88 [ 8 ]. The majority of studies on intrinsic TLR signaling on B cell activation have focused on TLR4 and TLR9 since these receptors are strongly expressed on murine B cells [ 12 ]. Our results suggest that other TLR ligands could have equally strong effects despite low expression, and that the B cell response may be regulated by requiring more than one ligand. In particular, TLR4 is not expressed on naïve human B cells as it is on murine B cells [ 12 ], our results may suggest that combining TLR4 agonists with other TLR agonists may increase the translational relevancy of these adjuvants. Stimulation with the poly I:C and Pam3CSK4 agonist combination induced increased expression of CD80 and CD86, which are co-receptors that are important for antigen-presentation to CD4 + T cells. The expression of CD80 never reached the levels of the other surface receptors evaluated, but it has been reported that expression of CD80 peaks 48–72 hours after upregulation [ 38 ]. Ppoly I:C and Pam3CSK4 stimulated B cells were most efficient at promoting allogeneic CD4 + T cell proliferation. An interesting finding was the increased expression of CD25, the high affinity IL-2 receptor, on B cells in response to poly I:C and Pam3CSK4 stimulation. While the B cells alone did not produce detectable amounts of IL-2, CD4 + T cells in co-culture with B cells did. This may indicate a mechanism through which B cell activation is amplified during a T-cell-dependent antigen response. Of note, CD25 + B cell isolated from murine splenocytes have been demonstrated by others to have enhanced ability to stimulate T cells in vitro [ 39 ]. Poly I:C and Pam3CSK4 induced a significant increase in IL-6, TNF-α and CXCL10 secretion by B cells. Both IL-6 and TNF-α are important cytokines involved in promoting B cell survival and proliferation [ 40 ]. CXCL10 is a chemokine recognized by CXCR3, an important chemokine receptor involved in migration of activated T cells during Th1-type immune responses [ 41 ]. Poly I:C is known to induce expression of CXCL10 in other cell types [ 35 ]. Expression of CXCL10 by activated B cells could therefore promote B and T cell interactions by recruiting activated T cells expressing CXCR3. Differentiation of B cells into antibody producing plasma cells in vitro was enhanced by poly I:C and Pam3CSK4 stimulation in combination with T-cell-dependent stimulation. It is noteworthy that stimulation with poly I:C alone induced CD138 expression whereas Pam3CSK4 did not, and yet Pam3CSK4 alone induced IgG secretion whereas poly I:C did not. In both cases, the combination of poly I:C and Pam3CSK4 induced the highest levels of both CD138 and IgG secretion. This could be an indication that signaling pathways induced by either TLR3 or TLR1/2 are subject to cross-regulation, and stimulation through one receptor enhances the pathways induced by the other. This type of interaction has been previously observed on macrophages stimulated sequentially with poly I:C and Pam3CSK4 [ 29 ]. B cells stimulated in T-cell-independent conditions did not produce detectable antibody secretion, indicating that the effects of these adjuvants on B cells in vivo would be unlikely to result in non-specific production of antibodies if stimulation occurred in the absence of peptide presentation to the B cells. TACI is a TNF superfamily receptor member expressed only by B lymphocytes. TACI can respond to B cell survival factors BAFF and APRIL, which are primarily produced by non-B cells in order to regulate B cell survival and differentiation during B cell maturation. Previously, TACI has been shown to be increased on murine B cells in response to TLR stimulation by LPS and CpG cells [ 42 , 43 ]. TACI itself interacts with intracellular MyD88 and converges with signaling induced by TLR and CD40 to promote class switch recombination in B cells [ 44 ]. In our studies, antibody production could only be induced in T-cell-dependent activation of B cells, and was augmented by poly I:C and Pam3CSK4 combination treatment. Although it is unlikely that TACI signaling participated in this event in vitro since BAFF and APRIL are not produced by B cells, it is interesting to speculate that increased TACI expression promoted by poly I:C and Pam3CSK4 contributes to more efficient plasma cell differentiation in vivo . TLR agonists are commonly used as vaccine adjuvants to boost antibody production; however, the contribution of direct TLR stimulation on B cells in development of antibody responses in vivo is not well understood [ 21 ]. Various studies using conditional knockouts for MyD88 or MHC class II expression have demonstrated that initiation of immune responses can be reliant on either B cells or DC, depending on the antigen type and availability [ 20 , 45 , 46 ]. Nevertheless, our in vivo studies using vaccines for influenza and anthrax clearly demonstrate that poly I:C and Pam3CSK4 adjuvant combination result in enhanced antigen-specific antibodies. However, we could not demonstrate that this effect was due to intrinsic TLR signaling on B cells in vivo . Most likely, the enhanced in vivo effect was due to TLR signaling induced by these agonists on both B cells and DCs. Further exploration of this would require extensive studies using knockout mice, this would be a valuable project to pursue. B cell interaction with antigen or adjuvant in the lymph node is facilitated by phagocytes that can transport intact antigens and adjuvants from the periphery using non-degradative intracellular compartments [ 13 , 14 ]. The vaccine delivery system used in this study, DepoVax ™ (DPX), is an oil based formulation that forms a depot in vivo. Using MRI to detect iron-labeled antigens, we have demonstrated that the immune system plays an active role in taking up the vaccine and transporting the components to the lymph node [ 3 ]. In this way, DPX is unlike other vaccine platforms which are commonly either an aqueous buffer or emulsion, both of which allow vaccine components to diffuse from the vaccine site more readily (manuscript in preparation). The active uptake mechanism of DPX ensures that vaccine components are directly delivered to immune system cells. In this work, we have shown how different adjuvants or adjuvant combinations may affect immune responses induced by the vaccine, and this may be partially attributed to direct stimulation of B cells by poly I:C and Pam3CSK4 made possible using the DPX formulation. The poly I:C and Pam3CSK4 adjuvant system induced a strong and robust antigen-specific antibody response in vivo . The doses used were the minimal dose required to elicit antigen-specific antibody responses in vivo ( S8 Fig ). Since this response could be induced with minimal doses of each adjuvant together, the adjuvant system is not only more cost effective, but also reduces the potential for an of adverse reaction to the vaccine. There is an urgent need to develop more effective vaccination strategies for both anthrax and influenza. A potential added benefit of using a dual adjuvant system is that these adjuvants, in particular poly I:C, are also known to induce potent T cell responses that can contribute to robust immunity [ 47 ]. The effect of poly I:C/ Pam3CSK4 adjuvant system on the kinetics of T cell response warrants further investigation. In conclusion, we have demonstrated that poly I:C and Pam3CSK4 stimulation of B cells results in a unique activation profile and when used in combination contribute to the optimal induction of B cell activation. Poly I:C and Pam3CSK4 together are an effective adjuvant system capable of boosting antibody responses in vivo when administered with protein antigens and could prove to be a promising new DPX-based vaccine formulation for influenza and anthrax indications. Supporting information S1 Fig Purity of B cells isolated from spleen. (A) Staining of C57BL/6 splenocyte starting populations with CD3-FITC (145-2C11), CD11c-PE (N418) and CD19-APC (1D3). (B) Staining of purified B cell populations for same markers. (C) Phenotypic analysis of purified B cells before culture staining with CD23-FITC (B3B4), B220-PE (RA3-6B2), CD5-APC (5373) and IgD-APC (11-26c). (PDF) Click here for additional data file. S2 Fig Representative histograms for isotype controls. Purified C57BL/6 CD19 + B cells were stimulated with poly I:C (25 ug/mL), Pam3CSK4 (1 ug/mL) or the combination of both adjuvants for 24 hours. B cells were then analysed by flow cytometry using isotype controls (A) Armenian Hamster IgG-FITC (eBio299Arm), (B) Rat IgG2a-PE (aBR2a), (C) Rat IgG1-APC (eBRG1). Representative of at least three independent experiments. (PDF) Click here for additional data file. S3 Fig Dosing of poly I:C and PamCSK4 in vitro on B cell response. B cells were isolated from the spleens of naïve C57BL/6 mice (n = 3) and stimulated with various concentrations of poly I:C and Pam3SK4. Expression of CD40 (A), CD80 (B) and MHC class II (C) was determined by flow cytometry after 24 hour stimulation. Dashed line indicates level of unstimulated B cells. Data are shown as average ± SEM of 3 individual B cell preparations as indicated and was collected in a single experiment(D) Proliferation of B cells was measured after 3 days incubation by [ 3 H]-TdR uptake (n = 2–7). Data shown as average ± SEM of 2–7 individual B cell preparations pooled from at least 2 separated experiments. Statics by 1-way ANOVA with Dunnett's post-test comparing each dose to unstimulated, *p<0.05, **p<0.01, ***p<0.001. (TIF) Click here for additional data file. S4 Fig Dose response to poly I:C and Pam3CSK4 combinations in vitro. B cells were isolated from the spleens of naïve C57BL/6 mice (n = 3) and stimulated with various concentrations of poly I:C and Pam3SK4 alone and in combination for 24 hours. Expression of CD80 (A), CD40 (B), MHC class II (C) was detected by flow cytometry. Secretion of IL-6 (D) was detected by ELISA, BLD: below limit of detection. Results are shown as the average of 3 individual B cell preparations and was collected in a single experiment, the selected combination of poly I:C (25 μg/mL) and Pam3CSK4 (1 μg/mL) is bolded. (PDF) Click here for additional data file. S5 Fig Representative histograms for B cell surface marker expression. Purified C57BL/6 CD19 + B cells were stimulated with poly I:C (25 ug/mL), Pam3CSK4 (1 ug/mL) or the combination of both adjuvants for 24 hours. B cells were then analysed by flow cytometry for expression of CD86, CD80, CD25, MHC class II (IA/IE), CD69 and CD40. Results from multiple experiments are summarized in Fig 1 . (PDF) Click here for additional data file. S6 Fig TLR2 knockout B cell stimulation. CD19 + B cells were purified from TLR2 -/- (n = 4) or C57BL/6 wild type (n = 4) mice and stimulated with poly I:C (25 ug/mL), Pam3CSK4 (1 ug/mL) or the combination of both adjuvants for 24 hours in (A) T-cell-independent and (B) T-cell-dependent conditions. B cells were analysed by flow cytometry for expression of CD40, CD86, MHC class II, CD25 and CD80. (C) Supernatants were analysed by ELISA for CXCL10. (TIF) Click here for additional data file. S7 Fig TLR3 knockout B cell stimulation. CD19 + B cells were purified from TLR3 -/- (n = 5) or B6;129SF2/J wild type (n = 4) mice and stimulated with poly I:C (25 ug/mL), Pam3CSK4 (1 ug/mL) or the combination of both adjuvants for 24 hours in (A) T-cell-independent and (B) T-cell-dependent conditions. B cells were analysed by flow cytometry for expression of CD40, CD86, MHC class II, CD25 and CD80. (C) Supernatants were analysed by ELISA for IL-6. (D) Supernatants were analysed by ELISA for CXCL10. (TIF) Click here for additional data file. S8 Fig Dosing of poly I:C and Pam3CSK4 in rPA vaccine. CD-1 mice were vaccinated with rPA antigen (2 ug) formulated with (A) poly I:C or (B) Pam3CSK4, at indicated doses, in DPX. Antigen-specific antibodies were detected in serum at 4 and 8 weeks post immunization. (TIF) Click here for additional data file.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7020563/

Looking for the X Factor in Bacterial Pathogenesis: Association of orfX - p47 Gene Clusters with Toxin Genes in Clostridial and Non-Clostridial Bacterial Species

The botulinum neurotoxin (BoNT) has been extensively researched over the years in regard to its structure, mode of action, and applications. Nevertheless, the biological roles of four proteins encoded from a number of BoNT gene clusters, i.e., OrfX1-3 and P47, are unknown. Here, we investigated the diversity of orfX-p47 gene clusters using in silico analytical tools. We show that the orfX-p47 cluster was not only present in the genomes of BoNT-producing bacteria but also in a substantially wider range of bacterial species across the bacterial phylogenetic tree. Remarkably, the orfX-p47 cluster was consistently located in proximity to genes coding for various toxins, suggesting that OrfX1-3 and P47 may have a conserved function related to toxinogenesis and/or pathogenesis, regardless of the toxin produced by the bacterium. Our work also led to the identification of a putative novel BoNT-like toxin gene cluster in a Bacillus isolate. This gene cluster shares striking similarities to the BoNT cluster, encoding a bont/ntnh -like gene and orfX-p47 , but also differs from it markedly, displaying additional genes putatively encoding the components of a polymorphic ABC toxin complex. These findings provide novel insights into the biological roles of OrfX1, OrfX2, OrfX3, and P47 in toxinogenesis and pathogenesis of BoNT-producing and non-producing bacteria. 1. Introduction Botulinum neurotoxin (BoNT) is mostly produced by the Gram-positive spore-forming anaerobic bacterium Clostridium botulinum . BoNT causes botulism which is a rare but deadly disease affecting both humans and animals [ 1 ]. The most well-known form of the disease is food-borne botulism, resulting from the consumption of preformed BoNT present in inappropriately preserved food products [ 2 , 3 ]. Upon ingestion and absorption into the body, BoNT, a zinc-dependent metalloprotease, blocks neurotransmission through the cleavage of key proteins within the cholinergic nerve terminals, causing flaccid paralysis [ 4 ]. BoNT is encoded from neurotoxin gene cluster(s) (NGC) which can be located either in the chromosome or within mobile genetic elements, i.e., plasmids or bacteriophages, and acquired through horizontal gene transfer [ 5 ]. The NGC typically encodes several components of the active progenitor toxin complex (PTC), as well as regulatory elements related to toxinogenesis [ 6 ]. The gene organization and content of NGCs differ substantially between botulinum neurotoxin producing strains. Two major NGC types are recognized based on the type of genes localized in direct vicinity of the BoNT-encoding gene ( bont ): the hemagglutinin ( ha ) type and the orfX-p47 -type NGCs. Both cluster types, besides carrying bont , contain five to six additional genes which encode known or putative neurotoxin-associated proteins (NAPs) and an alternative sigma factor, BotR. Regardless of the NGC type, BoNT is exclusively encoded with the non-toxic non-hemagglutinin protein (NTNH) which interlocks with BoNT to form a minimally functional PTC (M-PTC). Formation of the M-PTC shields the naturally fragile BoNT and protects it from degradation in the harsh conditions of the gastrointestinal tract [ 7 ]. The ha -type NGCs encode three hemagglutinins (HA-17, HA-33, and HA-70) which link with the M-PTC to form a large PTC (L-PTC) [ 8 , 9 ]. The L-PTC facilitates the transport of BoNT through the intestinal epithelial barrier enabling the toxin to enter circulation [ 10 , 11 , 12 ]. In contrast, the orfX-p47 type NGC does not harbor hemagglutinin genes. The ha17 , ha33 , and ha70 are replaced with genes named orfX1 , orfX2 , and orfX3 . Despite the lack of structural homology between OrfX1-3 and the HA proteins [ 13 ], an identical operon structure and location next to the bont - ntnh operon supports a hypothesis that the proteins encoded by orfX1 , orfX2 , and orfX3 might also hold a role in PTC formation or in BoNT pathogenesis. However, the roles of OrfX1, OrfX2 (PDB ID: 6EKV), and OrfX3 proteins remain to be elucidated. Along with the orfX1-3 genes, the orfX-p47 type NGCs exclusively harbor a gene ( p47 ) encoding a 47-kDa product (PDB ID: 5WIX, 6EKT) of an unknown function. OrfX1, OrfX2, OrfX3, and P47 remain poorly characterized and only few studies investigated their structures and biochemical properties. Biochemical analysis showed that OrfX1, OrfX2, and P47 attach to lipids in vitro [ 13 , 14 ]. The lipid-binding properties of OrfX2 and P47 were attributed to tubular lipid-binding (TULIP) domains so far encountered only in eukaryotic proteins [ 13 ]. Occasionally, the OrfX1, OrfX2, OrfX3, and P47 proteins were detected in association with BoNT immune-extracted from C. botulinum cultures or with commercially available purified BoNT complexes [ 15 ]. However, their presence varied significantly between samples, which suggests that specific conditions may be required to capture a putative L-PTC containing OrfX1, OrfX2, OrfX3, and/or P47. Besides the putative roles of OrfX1, OrfX2, OrfX3, and/or P47 as structural components of the L-PTC, roles in regulation of NGC expression have been proposed [ 16 ] but not experimentally supported [ 17 ]. A better understanding of P47, OrfX1, OrfX2, and OrfX3 would shed light on the mode of action of BoNTs encoded from the orfX-p47 -type NGCs. Until recently, BoNT production was merely associated with clostridial species including C. botulinum and some strains of Clostridium argentinense , Clostridium baratii , and Clostridium butyricum [ 6 , 18 , 19 ]. Rapid development of next-generation sequencing (NGS) technologies and bioinformatics tools allowed affordable genome sequencing and functional genomic analyses of clinical and environmental isolates, and resulted in in silico discovery of novel BoNT types, not only in Clostridiales but also in phylogenetically distant bacterial species [ 20 , 21 , 22 , 23 ]. Along these lines, the present study, based on bioinformatic mining of bacterial genome databases, led to the identification of genes putatively encoding OrfX1, OrfX2, OrfX3, and P47 in bacterial species belonging to Alphaproteobacteria , Bacilli , Betaproteobacteria , Cytophagia , and Gammaproteobacteria . Diverse and novel orfX1 , orfX2 , orfX3 , and p47 gene arrangements as well as the presence of a truncated form of orfX2 (tentatively called orfX-T ), previously unseen in Clostridiales , were reported in these genomes. Strikingly, these genes were consistently neighboring genes encoding non-BoNT toxins, including the crystal toxin (Cry) or vegetative insecticidal proteins (VIP). This suggests that the biological function of OrfX1, OrfX2, OrfX3, and P47 relates to toxinogenesis and/or pathogenesis, and is not BoNT-specific. Finally, we identified an orfX-p47 -containing toxin gene cluster harboring a bont/ntnh -like toxin gene in a Bacillus isolate. This cluster shares similarities with C. botulinum NGC but also possesses marked differences: besides having three orfX genes ( orfX3 and two orfX-T ) and a p47 , it contains a solitary truncated bont/ntnh -like gene lacking the canonically accompanying ntnh/bont counter partner. Instead, this cluster is juxtaposed by two open reading frames which putatively encode components of a tripartite toxin complex. This finding suggests that the truncated BoNT/NTNH-like protein detected in a Bacillus isolate may display a mode of action and host specificity different from canonical BoNTs. Together, the present work reveals new insights into the biological roles of OrfX1, OrfX2, OrfX3, and P47 in bacterial toxinogenesis and pathogenesis. 2. Results and Discussion 2.1. Phylogenetic Distribution of orfX1, orfX2, orfX3, and p47 in Bacteria The BoNT gene cluster frequently harbors orfX1 , orfX2 , orfX3 , and p47 in BoNT-producing Clostridia . While the ha operon has been exclusively linked to BoNT gene clusters of Clostridia , the orfX1-3 operon and p47 have also been found in BoNT gene clusters of phylogenetically more distant species, such as Enterococcus faecium [ 21 ]. We hypothesized that orfX1-3 and p47 may be found in a wider range of bacterial species. As OrfX2, OrfX3, and P47 belong to the Clostridium P47 superfamily (Pfam ID: PF06597), we searched for the PF06597 domain using the online web tool Annotree ( Figure 1 ). Bacterial species belonging to 23 different orders, such as Bacteroidales , Rhizobiales , or Streptomycetales , were identified as encoding Clostridium P47 superfamily proteins. As Annotree includes only representative genomes and thus omits intraspecies genomic diversity, it is possible that the analysis does not fully reflect the true phylogenetic distribution of the superfamily PF06597. Therefore, it is likely that proteins belonging to the Clostridium P47 superfamily are present even in a larger number of taxonomic orders or families. This is in contrast with the phylogenetic distribution of BoNT and BoNT-like proteins, suggesting that the biological function of Clostridium P47 superfamily proteins is not exclusively associated with BoNT. Further evidence supporting this hypothesis is discussed below. We also searched for putative homologs of Clostridium P47, OrfX1, OrfX2, and OrfX3 using protein BLAST in the 'non-redundant protein sequence (nr)' database within and beyond the order Clostridiales , to identify putative novel orfX-p47 gene clusters. In line with the Annotree search, OrfX1, OrfX2, OrfX3, and P47 homologs were detected in bacterial species belonging to various taxonomic orders ( Table 1 , Figure 1 , and Supplementary Tables S1–S5 ). In most genomes analyzed, the genes encoding OrfX1-3 and P47 homologs were arranged in clusters. While the copy number and presence of the different orfX1-3 genes varied between the identified clusters, each cluster always displayed one copy of p47 ( Figure 2 ). Additionally, most of the detected novel orfX-p47 clusters resided genes encoding components of toxins other than BoNT. In few cases, we detected orphan genes for proteins harboring the Clostridium P47 superfamily. Of note, we did perform a similar search for Clostridium hemagglutinin homologs beyond the Clostridiales order but failed to detect any significant hits, suggesting that the ha gene clusters are conserved to Clostridiales , as opposed to orfX-p47 genes. 2.2. Gene Organization and Arrangement of orfX1, orfX2, orfX3, and p47 To perceive the diversity of OrfX1, OrfX2, OrfX3, and P47 encoding loci, we compared the gene arrangement of the OrfX1-3 and P47 protein homologs retrieved from the genomes of thirteen bacterial strains ( Table 1 ). These strains originated from various environmental samples, i.e., air, insects, sediment, soil, and water samples. Most of these bacterial species are either invertebrate pathogens ( Brevibacillus laterosporus , Bacillus thuringiensis , Rickettsiella grylli , Paenibacillus larvae ) or plant pathogens ( Erwinia amylovora ). All the analyzed orfX-p47 clusters harbored one copy of p47 . We, therefore, constructed a maximum likelihood tree of all P47 proteins ( Figure 2 ) and juxtaposed the corresponding orfX-p47 gene clusters according to the tree. Phylogenetically related bacterial species harbored clusters with similar gene arrangement and order. In most Gram-positive bacteria, the overall arrangement of orfX1-3 and p47 was unidirectional and well-conserved with the following order: orfX1 , orfX2 , orfX3 , and p47 ( Figure 2 ). In contrast, in C. botulinum strains, the orientation of p47 is opposite to the orfX1 , orfX2 , and orfX3 and occasionally flanked by a regulator gene or a mobile element. The orfX-p47 gene cluster in C. botulinum strain 111 encoding BoNT/X markedly differs from the ones in other C. botulinum strains, as previously reported, and its gene arrangement rather resembles the orfX-p47 clusters found in phylogenetically more distant species, such as Bacilli , Brevibacilli , and Paenibacilli where p47 is localized downstream the orfX1-3 operon. Remarkably, there is a clear dichotomy in the architecture of the orfX-p47 gene clusters between Gram-negative and Gram-positive species: orfX1 was present within the orfX-p47 gene cluster of Gram-positive species but was absent in clusters from all analyzed Gram-negative bacteria. Considering the recently demonstrated lipid binding ability of OrfX1 in vitro, it is tempting to speculate that the biological function of OrfX1 could relate to membranes in the Gram-positive cell envelope. Gram-negative species, in turn, lacked orfX1 , and carried one or two copies of truncated orfX2 , tentatively called orfX-T , and occasionally harbored orfX3 . 2.3. P47, OrfX1, OrfX2, and OrfX3 Have a Common Origin Since P47, OrfX2, and OrfX3 all belong to the Clostridium P47 superfamily, we further examined the domain conservation and the evolutionary relationship between these proteins found in phylogenetically distinct bacterial species. We constructed a maximum likelihood tree of all OrfX1-3 and P47 proteins and collated it with the corresponding MEME (Multiple EM for Motif Elicitation) analysis which detected any shared and recurring amino acid motifs or patterns in a color-coded graphical form. All analyzed protein sequences grouped into five main clades, namely clade P47, and clades OrfX1, OrfX2, OrfX3, and OrfX-T ( Figure 3 ). OrfX1 diverged early from the other clades, retaining only a few motifs that were also present in the other clades. Remarkably, all the analyzed Gram-negative and some Gram-positive species, e.g., Bacillus sp. 2SH and Ruminoccus albus , were devoid of OrfX1 and OrfX2 but instead harbored OrfX-T, a truncated form of OrfX2. OrfX-T shares a number of unique motifs with OrfX2, but lacks the N-terminal domain of OrfX2 which appears to be associated with the presence of OrfX1. The fact that the clusters harboring orfX-T always lacked orfX1 and intact orfX2 may suggest that OrfX-T alone takes over the biological functions of both OrfX1 and OrfX2. The MEME analysis showed that P47, OrfX2, and OrfX3 share several conserved motifs inside the core domain, suggesting that the three proteins originate from a common ancestor ( Figure 3 ). There is also a high degree of conservation among three consecutive motifs located in the N-terminal domains of P47 and OrfX3 and in the core domain of OrfX2, and among other three motifs found within the core domains of each of the three proteins. P47 and OrfX3 share similar N-terminal and core domains but possess different motifs within their C-terminal domains. While the C-terminal domain of P47 appears to be unique, the one in OrfX3 shares similarities with the C-terminal domain of OrfX2. Intriguingly, the novel orfX-p47 clusters identified in B. laterosporus (VIP toxin producer), B. thuringiensis (Cry toxin producer) and P. thiaminolyticus (VIP toxin producer) encoded an extended P47 with an additional domain fused to its C-terminus. The additional domain detected in P47 was predicted to contain a ricin-type beta-trefoil lectin-like domain (Pfam ID: PF14200), involved in carbohydrate binding [ 32 ] and found for instance in HA-33 [ 33 ]. Furthermore, one of the two orfX-p47 clusters harbored by B. laterosporus carrying a gene for Cry toxin, encoded P47 fused to a domain identified as a fungal immunomodulatory protein Fve domain (Pfam ID: PF09259) that may bind cell-surface carbohydrates [ 34 ]. The identification of these two distinct types of cell-binding domains fused to P47 suggests a putative role for P47 as a cell-binding anchor (lectin-like domain or fungal immunomodulatory protein Fve domain) that could dictate target specificity of the associated toxins produced by these isolates. 2.4. Association of orfX1, orfX2, orfX3, and p47 with Toxin Genes Most of the orfX-p47 clusters were adjacent to genes encoding various well-studied insecticidal toxins like Cry toxin (classified as delta-endotoxin) or VIP (belonging to the group of binary toxins) ( Table 2 ). A significant number of putative toxin genes associated with orfX-p47 clusters encoded proteins containing different types of motifs or domains correlated with the toxic properties. Among these, we have identified proteins harboring rearrangement hotspot (RHS) repeats present in a wide range of insecticidal toxins [ 35 , 36 ], ribosome inactivating protein domains [ 37 ], or bacterial immunoglobulin-like domains, the latter shown to be associated with intestinal colonization [ 38 ] and recently detected within the BoNT-like toxin encoded by Weissella oryzae [ 39 ]. Of note, these toxins share several common features, including oral infection route, proteolytic activation, and an ability to form pores in the target host cells [ 40 , 41 , 42 , 43 ]. Interestingly, literature describing Cry and VIP toxins, or bacterial genomes encoding them, has not recognized the presence of OrfX or P47-encoding genes in the close neighborhood of the toxin genes. To the best of our knowledge, only one report shows the presence of orfX-p47 cluster in a non-BoNT-encoding genome, however this finding has not been further investigated [ 44 ]. Our BLAST analysis showed that cry and vip genes are seldom accompanied by orfX-p47 clusters. Moreover, Cry toxin is biologically effective alone, for example when used in pest control, thus OrfXs and P47 are likely not essential in Cry pathogenesis. We assume that in the absence of OrfX1-3 and P47, potency of the toxins may be lower, albeit sufficient to kill the host. Alternatively, other as-yet unidentified accessory proteins may assist in toxinogenesis. It will be important to experimentally verify if the association of Cry and VIP toxins with OrfX and P47 proteins impacts their potency. Our data mining also revealed several assumingly orphan orfX-p47 clusters neighboring genes encoding hypothetical proteins. It is not clear whether these genes co-operate with orfX1-3 and p47 , whether they encode toxic components, and whether they are coincidentally co-localized with the orfX-p47 clusters as a result of a phage or mobile element activity. 2.5. Identification of BoNT/NTNH-Like Protein in Bacillus sp. 2SH Bioinformatic analysis led to the discovery of an orfX-p47 cluster-associated gene putatively encoding a protein with similarities to BoNT and NTNH. We termed this putative protein as BoNT/NTNH-like A component (BNA) due to its close similarity to BoNT and NTNH and to its predicted function as component A of an ABC-type toxin complex (discussed below). This peculiar cluster was identified within the unclosed genome of a Bacillus sp. isolate 2SH recovered from alpine fresh water in Trento, Italy. The predicted 825 amino acid long sequence of BNA protein was compared with an extensive protein dataset consisting of various types/subtypes of BoNT. Preliminary identification of conserved protein domains revealed that BNA contains the clostridial neurotoxin zinc protease domain (Pfam ID: PF01742) characteristic for both BoNT and NTNH, and a clostridial neurotoxin translocation domain (Pfam ID: PF07952) which is identified in BoNT, NTNH and the tetanus neurotoxin (TeNT). Interestingly, the predicted BNA polypeptide chain lacked the C-terminal heavy chain domain present in both BoNT and NTNH ( Figure 4 a). The amino acid sequence alignment suggested that BNA carries several conserved motifs present in BoNT and NTNH. They include an active site-stabilizing motif RxxY [ 45 ] and a translocation motif PWISQSLN, which in BoNT is conserved as PYxGxALN and in NTNH as PWxGxALN [ 46 ]. Interestingly, BNA demonstrates the presence of two cysteines (C376 and C386) located between the BNA zinc protease and translocation domain. Homologous residues found in BoNT allow the formation of an inter-chain disulfide bond, which upon reduction enables the toxin to translocate across the target-cell membranes [ 47 , 48 ] ( Figure 4 a). The conserved active site zinc-coordinating HExxH motif present in zinc metalloproteases, including BoNT, is absent in BNA. Instead, BNA contains the amino acid sequence SKLIE, undetected in any of the known BoNTs. The SKLIE sequence does not contain two histidine residues necessary for chelating catalytic zinc ions in botulinum and tetanus zinc metalloproteases [ 49 ]. However, it remains to be experimentally determined whether the SKLIE motif is functional and in which conditions, and what would constitute the molecular target of BNA. The findings suggest that BNA cannot bind to the target cell receptors typically recognized by BoNT and most likely cannot incorporate a zinc residue that is indispensable in BoNT for proteolytic cleavage of its target proteins; however it potentially could form a translocation channel. Within the maximum likelihood phylogenetic tree, BoNT and NTNH clustered distinctly as previously reported [ 50 ]. Although the BNA amino acid sequence clustered with the NTNH clade, its early branching places it between BoNT and NTNH ( Figure 4 b). This, together with the fact that the BNA gene ( bna ) does not reside with an ntnh -like gene, may suggest that BNA represents the common ancestor of BoNT and NTNH before their likely emergence through gene duplication [ 51 ]. On the other hand, the long phylogenetic distance between BNA and BoNT or NTNH could suggest that bna constitutes a pseudogenized form of bont or ntnh and is no longer functional. Nevertheless, we have not identified any premature stop codon or frameshift mutations within the bna sequence, which opposes the pseudogene hypothesis and supports functionality. The putative BNA protein also shows striking structural resemblance to BoNT and NTNH. Its 3D structure predicted using the Phyre2 tool matched with the crystal structure of BoNT/B (PDB ID: 1S0B) indicating that BNA is likely highly similar to BoNT/B and, by extension, to other BoNTs. Another 3D structure-generating software, I-TASSER, modeled the structure of BNA utilizing NTNH/D (PDB ID: 3VUOA) as a template. The latter BNA model is the one presented and analyzed here ( Figure 4 c). This 3D model confirms the lack of C-terminal domain of the BNA heavy chain. However, the remaining light chain and N-terminal heavy chain domains share structural identity with BoNT and NTNH. Unfortunately, the degree of sequence conservation did not allow a model with a maximum confidence, therefore, a crystal structure of BNA would further validate in silico generated structural models. The bna gene is lacking the canonical bont or ntnh counterpart, which makes BNA, to our knowledge, the first described stand-alone BoNT/NTNH-like protein. Instead, bna is accompanied by two open reading frames encoding proteins containing RHS repeats ( Figure 2 , Bacillus sp. 2SH cluster). Their 3D models match with the crystal structure of TcdB2-TccC3 toxin subcomplex of Photorhabdus luminescens (PDB ID: 4O9X; 100% confidence in Phyre2). This subcomplex is involved in the formation of tripartite ABC toxin of P. luminescens which is a well-studied example of RHS-repeat containing polymorphic toxin targeting insect larvae [ 52 , 53 ]. Recent studies showed that BoNT-like proteins can exhibit insecticidal potential [ 23 ], therefore the presence of bna next to the genes putatively encoding anti-insect toxin components did not appear coincidental. Accordingly, we discuss below whether incorporation of BNA into an ABC toxin complex can be rationalized. Each component of the ABC toxin complex is responsible for performing a different task, and when assembled an active multimeric structure is generated. The A protein forms a pentameric structure to make a translocation channel in the target host cells [ 54 ]. Accordingly, in the case of Bacillus sp. 2SH putative ABC complex, BNA could act as the A component due to the fact it encompasses a BoNT translocation domain which may form translocation pores in lipid bilayers [ 55 ]. The putative self-oligomerization step would therefore stabilize BNA, explaining why BNA lacks BoNT or NTNH-like assistant. The two other ORFs localized downstream of bna appear to encode the B and C components which form a capsule-like structure protecting the cytotoxic hypervariable region inside the C-domain of the C component, which can exhibit different modes of toxicity [ 56 ]. The NCBI conserved domain detection tool showed that the C-domain of the putative C protein of Bacillus sp. 2SH consists of bacterial SNF2 helicase related to chromatin remodeling [ 57 ]. This suggests that the putative BNA-associated ABC complex toxicity could rely on rearranging the DNA of target cells. A relevant piece of evidence supporting the hypothesis of BNA being an A component of ABC toxin complex is the demonstrated interchangeability of the A component: the B and C components co-expressed with the A component of different bacterial strains can form a toxic ABC complex with variable, A component-defined host specificity [ 54 , 58 ]. Accordingly, we speculate that BNA could be utilized by the Bacillus sp. 2SH ABC toxin system as an externally acquired gatekeeper to the neuronal cells. To our knowledge, this is the first case where a BoNT/NTNH-like protein could be utilized as a component of a non-botulinum toxin system. Further experimental investigation of this putative novel toxin complex will bring more information about its target specificity, action, and evolutionary status. 2.1. Phylogenetic Distribution of orfX1, orfX2, orfX3, and p47 in Bacteria The BoNT gene cluster frequently harbors orfX1 , orfX2 , orfX3 , and p47 in BoNT-producing Clostridia . While the ha operon has been exclusively linked to BoNT gene clusters of Clostridia , the orfX1-3 operon and p47 have also been found in BoNT gene clusters of phylogenetically more distant species, such as Enterococcus faecium [ 21 ]. We hypothesized that orfX1-3 and p47 may be found in a wider range of bacterial species. As OrfX2, OrfX3, and P47 belong to the Clostridium P47 superfamily (Pfam ID: PF06597), we searched for the PF06597 domain using the online web tool Annotree ( Figure 1 ). Bacterial species belonging to 23 different orders, such as Bacteroidales , Rhizobiales , or Streptomycetales , were identified as encoding Clostridium P47 superfamily proteins. As Annotree includes only representative genomes and thus omits intraspecies genomic diversity, it is possible that the analysis does not fully reflect the true phylogenetic distribution of the superfamily PF06597. Therefore, it is likely that proteins belonging to the Clostridium P47 superfamily are present even in a larger number of taxonomic orders or families. This is in contrast with the phylogenetic distribution of BoNT and BoNT-like proteins, suggesting that the biological function of Clostridium P47 superfamily proteins is not exclusively associated with BoNT. Further evidence supporting this hypothesis is discussed below. We also searched for putative homologs of Clostridium P47, OrfX1, OrfX2, and OrfX3 using protein BLAST in the 'non-redundant protein sequence (nr)' database within and beyond the order Clostridiales , to identify putative novel orfX-p47 gene clusters. In line with the Annotree search, OrfX1, OrfX2, OrfX3, and P47 homologs were detected in bacterial species belonging to various taxonomic orders ( Table 1 , Figure 1 , and Supplementary Tables S1–S5 ). In most genomes analyzed, the genes encoding OrfX1-3 and P47 homologs were arranged in clusters. While the copy number and presence of the different orfX1-3 genes varied between the identified clusters, each cluster always displayed one copy of p47 ( Figure 2 ). Additionally, most of the detected novel orfX-p47 clusters resided genes encoding components of toxins other than BoNT. In few cases, we detected orphan genes for proteins harboring the Clostridium P47 superfamily. Of note, we did perform a similar search for Clostridium hemagglutinin homologs beyond the Clostridiales order but failed to detect any significant hits, suggesting that the ha gene clusters are conserved to Clostridiales , as opposed to orfX-p47 genes. 2.2. Gene Organization and Arrangement of orfX1, orfX2, orfX3, and p47 To perceive the diversity of OrfX1, OrfX2, OrfX3, and P47 encoding loci, we compared the gene arrangement of the OrfX1-3 and P47 protein homologs retrieved from the genomes of thirteen bacterial strains ( Table 1 ). These strains originated from various environmental samples, i.e., air, insects, sediment, soil, and water samples. Most of these bacterial species are either invertebrate pathogens ( Brevibacillus laterosporus , Bacillus thuringiensis , Rickettsiella grylli , Paenibacillus larvae ) or plant pathogens ( Erwinia amylovora ). All the analyzed orfX-p47 clusters harbored one copy of p47 . We, therefore, constructed a maximum likelihood tree of all P47 proteins ( Figure 2 ) and juxtaposed the corresponding orfX-p47 gene clusters according to the tree. Phylogenetically related bacterial species harbored clusters with similar gene arrangement and order. In most Gram-positive bacteria, the overall arrangement of orfX1-3 and p47 was unidirectional and well-conserved with the following order: orfX1 , orfX2 , orfX3 , and p47 ( Figure 2 ). In contrast, in C. botulinum strains, the orientation of p47 is opposite to the orfX1 , orfX2 , and orfX3 and occasionally flanked by a regulator gene or a mobile element. The orfX-p47 gene cluster in C. botulinum strain 111 encoding BoNT/X markedly differs from the ones in other C. botulinum strains, as previously reported, and its gene arrangement rather resembles the orfX-p47 clusters found in phylogenetically more distant species, such as Bacilli , Brevibacilli , and Paenibacilli where p47 is localized downstream the orfX1-3 operon. Remarkably, there is a clear dichotomy in the architecture of the orfX-p47 gene clusters between Gram-negative and Gram-positive species: orfX1 was present within the orfX-p47 gene cluster of Gram-positive species but was absent in clusters from all analyzed Gram-negative bacteria. Considering the recently demonstrated lipid binding ability of OrfX1 in vitro, it is tempting to speculate that the biological function of OrfX1 could relate to membranes in the Gram-positive cell envelope. Gram-negative species, in turn, lacked orfX1 , and carried one or two copies of truncated orfX2 , tentatively called orfX-T , and occasionally harbored orfX3 . 2.3. P47, OrfX1, OrfX2, and OrfX3 Have a Common Origin Since P47, OrfX2, and OrfX3 all belong to the Clostridium P47 superfamily, we further examined the domain conservation and the evolutionary relationship between these proteins found in phylogenetically distinct bacterial species. We constructed a maximum likelihood tree of all OrfX1-3 and P47 proteins and collated it with the corresponding MEME (Multiple EM for Motif Elicitation) analysis which detected any shared and recurring amino acid motifs or patterns in a color-coded graphical form. All analyzed protein sequences grouped into five main clades, namely clade P47, and clades OrfX1, OrfX2, OrfX3, and OrfX-T ( Figure 3 ). OrfX1 diverged early from the other clades, retaining only a few motifs that were also present in the other clades. Remarkably, all the analyzed Gram-negative and some Gram-positive species, e.g., Bacillus sp. 2SH and Ruminoccus albus , were devoid of OrfX1 and OrfX2 but instead harbored OrfX-T, a truncated form of OrfX2. OrfX-T shares a number of unique motifs with OrfX2, but lacks the N-terminal domain of OrfX2 which appears to be associated with the presence of OrfX1. The fact that the clusters harboring orfX-T always lacked orfX1 and intact orfX2 may suggest that OrfX-T alone takes over the biological functions of both OrfX1 and OrfX2. The MEME analysis showed that P47, OrfX2, and OrfX3 share several conserved motifs inside the core domain, suggesting that the three proteins originate from a common ancestor ( Figure 3 ). There is also a high degree of conservation among three consecutive motifs located in the N-terminal domains of P47 and OrfX3 and in the core domain of OrfX2, and among other three motifs found within the core domains of each of the three proteins. P47 and OrfX3 share similar N-terminal and core domains but possess different motifs within their C-terminal domains. While the C-terminal domain of P47 appears to be unique, the one in OrfX3 shares similarities with the C-terminal domain of OrfX2. Intriguingly, the novel orfX-p47 clusters identified in B. laterosporus (VIP toxin producer), B. thuringiensis (Cry toxin producer) and P. thiaminolyticus (VIP toxin producer) encoded an extended P47 with an additional domain fused to its C-terminus. The additional domain detected in P47 was predicted to contain a ricin-type beta-trefoil lectin-like domain (Pfam ID: PF14200), involved in carbohydrate binding [ 32 ] and found for instance in HA-33 [ 33 ]. Furthermore, one of the two orfX-p47 clusters harbored by B. laterosporus carrying a gene for Cry toxin, encoded P47 fused to a domain identified as a fungal immunomodulatory protein Fve domain (Pfam ID: PF09259) that may bind cell-surface carbohydrates [ 34 ]. The identification of these two distinct types of cell-binding domains fused to P47 suggests a putative role for P47 as a cell-binding anchor (lectin-like domain or fungal immunomodulatory protein Fve domain) that could dictate target specificity of the associated toxins produced by these isolates. 2.4. Association of orfX1, orfX2, orfX3, and p47 with Toxin Genes Most of the orfX-p47 clusters were adjacent to genes encoding various well-studied insecticidal toxins like Cry toxin (classified as delta-endotoxin) or VIP (belonging to the group of binary toxins) ( Table 2 ). A significant number of putative toxin genes associated with orfX-p47 clusters encoded proteins containing different types of motifs or domains correlated with the toxic properties. Among these, we have identified proteins harboring rearrangement hotspot (RHS) repeats present in a wide range of insecticidal toxins [ 35 , 36 ], ribosome inactivating protein domains [ 37 ], or bacterial immunoglobulin-like domains, the latter shown to be associated with intestinal colonization [ 38 ] and recently detected within the BoNT-like toxin encoded by Weissella oryzae [ 39 ]. Of note, these toxins share several common features, including oral infection route, proteolytic activation, and an ability to form pores in the target host cells [ 40 , 41 , 42 , 43 ]. Interestingly, literature describing Cry and VIP toxins, or bacterial genomes encoding them, has not recognized the presence of OrfX or P47-encoding genes in the close neighborhood of the toxin genes. To the best of our knowledge, only one report shows the presence of orfX-p47 cluster in a non-BoNT-encoding genome, however this finding has not been further investigated [ 44 ]. Our BLAST analysis showed that cry and vip genes are seldom accompanied by orfX-p47 clusters. Moreover, Cry toxin is biologically effective alone, for example when used in pest control, thus OrfXs and P47 are likely not essential in Cry pathogenesis. We assume that in the absence of OrfX1-3 and P47, potency of the toxins may be lower, albeit sufficient to kill the host. Alternatively, other as-yet unidentified accessory proteins may assist in toxinogenesis. It will be important to experimentally verify if the association of Cry and VIP toxins with OrfX and P47 proteins impacts their potency. Our data mining also revealed several assumingly orphan orfX-p47 clusters neighboring genes encoding hypothetical proteins. It is not clear whether these genes co-operate with orfX1-3 and p47 , whether they encode toxic components, and whether they are coincidentally co-localized with the orfX-p47 clusters as a result of a phage or mobile element activity. 2.5. Identification of BoNT/NTNH-Like Protein in Bacillus sp. 2SH Bioinformatic analysis led to the discovery of an orfX-p47 cluster-associated gene putatively encoding a protein with similarities to BoNT and NTNH. We termed this putative protein as BoNT/NTNH-like A component (BNA) due to its close similarity to BoNT and NTNH and to its predicted function as component A of an ABC-type toxin complex (discussed below). This peculiar cluster was identified within the unclosed genome of a Bacillus sp. isolate 2SH recovered from alpine fresh water in Trento, Italy. The predicted 825 amino acid long sequence of BNA protein was compared with an extensive protein dataset consisting of various types/subtypes of BoNT. Preliminary identification of conserved protein domains revealed that BNA contains the clostridial neurotoxin zinc protease domain (Pfam ID: PF01742) characteristic for both BoNT and NTNH, and a clostridial neurotoxin translocation domain (Pfam ID: PF07952) which is identified in BoNT, NTNH and the tetanus neurotoxin (TeNT). Interestingly, the predicted BNA polypeptide chain lacked the C-terminal heavy chain domain present in both BoNT and NTNH ( Figure 4 a). The amino acid sequence alignment suggested that BNA carries several conserved motifs present in BoNT and NTNH. They include an active site-stabilizing motif RxxY [ 45 ] and a translocation motif PWISQSLN, which in BoNT is conserved as PYxGxALN and in NTNH as PWxGxALN [ 46 ]. Interestingly, BNA demonstrates the presence of two cysteines (C376 and C386) located between the BNA zinc protease and translocation domain. Homologous residues found in BoNT allow the formation of an inter-chain disulfide bond, which upon reduction enables the toxin to translocate across the target-cell membranes [ 47 , 48 ] ( Figure 4 a). The conserved active site zinc-coordinating HExxH motif present in zinc metalloproteases, including BoNT, is absent in BNA. Instead, BNA contains the amino acid sequence SKLIE, undetected in any of the known BoNTs. The SKLIE sequence does not contain two histidine residues necessary for chelating catalytic zinc ions in botulinum and tetanus zinc metalloproteases [ 49 ]. However, it remains to be experimentally determined whether the SKLIE motif is functional and in which conditions, and what would constitute the molecular target of BNA. The findings suggest that BNA cannot bind to the target cell receptors typically recognized by BoNT and most likely cannot incorporate a zinc residue that is indispensable in BoNT for proteolytic cleavage of its target proteins; however it potentially could form a translocation channel. Within the maximum likelihood phylogenetic tree, BoNT and NTNH clustered distinctly as previously reported [ 50 ]. Although the BNA amino acid sequence clustered with the NTNH clade, its early branching places it between BoNT and NTNH ( Figure 4 b). This, together with the fact that the BNA gene ( bna ) does not reside with an ntnh -like gene, may suggest that BNA represents the common ancestor of BoNT and NTNH before their likely emergence through gene duplication [ 51 ]. On the other hand, the long phylogenetic distance between BNA and BoNT or NTNH could suggest that bna constitutes a pseudogenized form of bont or ntnh and is no longer functional. Nevertheless, we have not identified any premature stop codon or frameshift mutations within the bna sequence, which opposes the pseudogene hypothesis and supports functionality. The putative BNA protein also shows striking structural resemblance to BoNT and NTNH. Its 3D structure predicted using the Phyre2 tool matched with the crystal structure of BoNT/B (PDB ID: 1S0B) indicating that BNA is likely highly similar to BoNT/B and, by extension, to other BoNTs. Another 3D structure-generating software, I-TASSER, modeled the structure of BNA utilizing NTNH/D (PDB ID: 3VUOA) as a template. The latter BNA model is the one presented and analyzed here ( Figure 4 c). This 3D model confirms the lack of C-terminal domain of the BNA heavy chain. However, the remaining light chain and N-terminal heavy chain domains share structural identity with BoNT and NTNH. Unfortunately, the degree of sequence conservation did not allow a model with a maximum confidence, therefore, a crystal structure of BNA would further validate in silico generated structural models. The bna gene is lacking the canonical bont or ntnh counterpart, which makes BNA, to our knowledge, the first described stand-alone BoNT/NTNH-like protein. Instead, bna is accompanied by two open reading frames encoding proteins containing RHS repeats ( Figure 2 , Bacillus sp. 2SH cluster). Their 3D models match with the crystal structure of TcdB2-TccC3 toxin subcomplex of Photorhabdus luminescens (PDB ID: 4O9X; 100% confidence in Phyre2). This subcomplex is involved in the formation of tripartite ABC toxin of P. luminescens which is a well-studied example of RHS-repeat containing polymorphic toxin targeting insect larvae [ 52 , 53 ]. Recent studies showed that BoNT-like proteins can exhibit insecticidal potential [ 23 ], therefore the presence of bna next to the genes putatively encoding anti-insect toxin components did not appear coincidental. Accordingly, we discuss below whether incorporation of BNA into an ABC toxin complex can be rationalized. Each component of the ABC toxin complex is responsible for performing a different task, and when assembled an active multimeric structure is generated. The A protein forms a pentameric structure to make a translocation channel in the target host cells [ 54 ]. Accordingly, in the case of Bacillus sp. 2SH putative ABC complex, BNA could act as the A component due to the fact it encompasses a BoNT translocation domain which may form translocation pores in lipid bilayers [ 55 ]. The putative self-oligomerization step would therefore stabilize BNA, explaining why BNA lacks BoNT or NTNH-like assistant. The two other ORFs localized downstream of bna appear to encode the B and C components which form a capsule-like structure protecting the cytotoxic hypervariable region inside the C-domain of the C component, which can exhibit different modes of toxicity [ 56 ]. The NCBI conserved domain detection tool showed that the C-domain of the putative C protein of Bacillus sp. 2SH consists of bacterial SNF2 helicase related to chromatin remodeling [ 57 ]. This suggests that the putative BNA-associated ABC complex toxicity could rely on rearranging the DNA of target cells. A relevant piece of evidence supporting the hypothesis of BNA being an A component of ABC toxin complex is the demonstrated interchangeability of the A component: the B and C components co-expressed with the A component of different bacterial strains can form a toxic ABC complex with variable, A component-defined host specificity [ 54 , 58 ]. Accordingly, we speculate that BNA could be utilized by the Bacillus sp. 2SH ABC toxin system as an externally acquired gatekeeper to the neuronal cells. To our knowledge, this is the first case where a BoNT/NTNH-like protein could be utilized as a component of a non-botulinum toxin system. Further experimental investigation of this putative novel toxin complex will bring more information about its target specificity, action, and evolutionary status. 3. Conclusions Here we showed that the orfX-p47 gene cluster, so far exclusively associated with botulinum neurotoxins, is widely distributed across the bacterial phylogenetic tree, reaching far beyond the distribution of the bont-ntnh gene pair. The orfX-p47 cluster showed large diversity in gene arrangement and gene content, which to great extent is parallel with the phylogenetic relationships among the bacteria harboring these gene clusters. Phylogenetic analysis of OrfX1-3 and P47 protein sequences suggested that the four proteins originate from a common ancestor and evolved through the acquisition or loss of functional domains. The Gram-negative bacteria possessing orfX-p47 clusters harbored atypical orfX genes in comparison to their putative Gram-positive orthologs. Different cell envelope architecture between Gram-negative and Gram-positive bacteria and evidence of OrfX1, OrfX2, and P47 binding to lipids in vitro [ 13 , 14 ] may suggest a role for OrfX proteins related to the bacterial cell envelope. This hypothesis awaits further testing by studying of the localization of the OrfX proteins within the bacterial cells. Remarkably, the orfX-p47 clusters were consistently associated with genes encoding various types of oral insecticidal toxins, i.e., delta-endotoxins (Cry toxin), binary toxins (VIP toxin) or ABC toxins. This suggests that the biological role of OrfX1-3 and P47 is not specific to BoNT. These proteins rather play a general role in oral toxinogenesis and pathogenesis of bacteria. Although the mode of action of OrfX proteins is unknown, recent work in Paraclostridium bifermentans subsp. malaysia showed that the co-expression of OrfX proteins may increase the oral toxicity of the mosquitocidal toxin PMP1 [ 23 ]. The relatively high degree of conservation among OrfX and P47 ( Figure 3 and Tables S1–S5 ) as opposed to the diversity of associated toxins (in terms of size, structure, mode of action) suggests that the OrfX and P47 proteins indirectly assist in toxin production, release, or trafficking. Further research on the insecticidal toxin-related OrfX and P47 proteins may lead to novel interventions in pest control. Identification of the orfX-p47 cluster in Bacillus sp. 2SH led to the discovery of a bont/ntnh -like toxin gene ( bna ) predicted to encode a protein with partial homology to BoNT and NTNH. This BoNT/NTNH-like A component appears unique since it is assumingly devoid of a canonical assistant protein. Instead, the bna gene is located upstream of two genes predicted to encode the B and C components of an ABC toxin complex [ 59 ]. We suggest that BNA constitutes the A component of an ABC toxin complex and therefore may have a role in determining host specificity of the toxin complex. To our knowledge, this is the first piece of evidence supporting incorporation of a BoNT/NTNH-like protein into a toxin complex different from the botulinum neurotoxin complex. Moreover, this is the first report of a bont/ntnh homolog in Bacillus sp. These data suggest that BoNT or NTNH proteins/homologs could be interchanged between different toxin complexes. This finding constitutes a basis for further studies on the interchangeability of BoNT domains with other toxins, and provides further evidence on the possible function of OrfX and P47 proteins in oral toxicity of BoNT and other bacterial toxins. 4. Materials and Methods 4.1. Sequence Database Mining and Conserved Domain Analysis Amino acid sequences of OrfX1 (WP_003369622.1), OrfX2 (WP_003371659.1), OrfX3 (WP_003372464.1), and P47 (WP_003374133.1) of C. botulinum strain Beluga were used as the query sequence to perform a protein–protein BLAST (blastp) search [ 60 ] against the NCBI 'non-redundant protein sequence (nr)' database (July 26, 2019). Homolog searches were also performed by excluding Clostridia (taxid: 186801), in order to filter out all OrfX1, OrfX2, OrfX3, and P47 protein sequences present in BoNT-producing Clostridia genomes. All BLAST (blastp) analyses were performed using default settings (scoring parameters: BLOSUM62 matrix; gap costs: existence 11 and extension1; expected threshold 10, word size 6) ( Supplementary Tables S1–S5 ). Conserved domains of protein homologs were further scanned and checked using the NCBI Conserved Domain Search using defined options (database: CDD v3.17 – 52,910 PSSMs, expected value threshold 0.01) [ 61 , 62 , 63 , 64 ]. Genome sequences and amino acid sequences of relevant protein homologs were retrieved from the NCBI database. OrfX1, OrfX2, OrfX3, and P47 protein sequences from C. botulinum (strains Beluga, Kyoto-F, Langeland, Mauritius, CDC_297, 111, Mfbjulcb3), C. baratii (strain Sullivan), E. faecium (strain 3G1_DIV0629), and P. bifermentans subsp. malaysia (strain Pbm) were included in the OrfX-P47 dataset for further comparative analysis. Search using the protein family corresponding to Clostridium P47 superfamily (Pfam ID: PF06597) was performed in Annotree [ 65 ]. Protein sequences from the genes located in the vicinity of the orfX gene cluster were further investigated for domain conservation (Pfam 32.0 search) [ 66 ], sequence homology (BLAST) [ 60 ], and structural homology (Phyre2) [ 67 ]. Phylogenetic distributions of the different bacterial genomes analyzed in this work were positioned within the tree generated by Annotree (taxonomic order level) [ 65 ]. The list of genome sequences analyzed in this study is shown in Table 1 . 4.2. Comparative Sequence Analysis, Motif-Based Sequence Analysis, and Phylogenetic Tree Analysis Sequence alignment based on ClustalW algorithm (gap opening penalty 10, gap extension penalty 0.1, protein weight matrix BLOSUM) [ 68 ] was computed in MEGA7 [ 69 ]. Sequence alignment was used to perform principal component analysis (PCA) in JalView v2.10.5 [ 70 ]. Maximum-likelihood phylogenetic trees were generated in MEGA7 with custom options (nearest-neighbor-interchange, Jones-Thornton model, 250 iterations) [ 69 ]. Putative motifs conserved among protein sequences were further analyzed using the MEME suite v.5.0.5 (MEME tool, classic mode, site distribution set to zero or one per sequence, number of motifs set to 30) [ 71 ]. 4.3. Sequence Analysis of the Putative Toxin Gene Cluster in Bacillus sp. 2SH and Structural Modeling Protein BLAST search, Pfam 32.0 search, and Phyre2 analysis of the gene (tentatively named BoNT/NTNH-like A component, BNA, WP_137842862.1) downstream p47 in Bacillus sp. 2SH revealed sequence similarities with NTNH type B of C. botulinum (BAQ12789.1, 30.03% identity, E value 8 × 10 −104 , query cover 90%) and other NTNH homologs. The amino acid sequence of BNA was subsequently aligned with a dataset of BoNT, NTNH, and TeNT amino acid sequences retrieved from public databases. Sequence alignment was performed based on ClustalW algorithm (gap opening penalty 10, gap extension penalty 0.1, protein weight matrix BLOSUM) [ 68 ] in MEGA7 [ 69 ]. Maximum-likelihood phylogenetic trees were generated in MEGA7 as described above [ 69 ]. Initial protein homology modeling was performed using Phyre2 (intensive modeling mode) [ 67 ] and SWISS MODEL [ 72 , 73 , 74 , 75 , 76 , 77 ]. Structural modeling was also carried out using I-TASSER v5.1 [ 78 , 79 ]. BNA model and relevant structural templates were visualized using PyMOL [ 80 ]. 4.1. Sequence Database Mining and Conserved Domain Analysis Amino acid sequences of OrfX1 (WP_003369622.1), OrfX2 (WP_003371659.1), OrfX3 (WP_003372464.1), and P47 (WP_003374133.1) of C. botulinum strain Beluga were used as the query sequence to perform a protein–protein BLAST (blastp) search [ 60 ] against the NCBI 'non-redundant protein sequence (nr)' database (July 26, 2019). Homolog searches were also performed by excluding Clostridia (taxid: 186801), in order to filter out all OrfX1, OrfX2, OrfX3, and P47 protein sequences present in BoNT-producing Clostridia genomes. All BLAST (blastp) analyses were performed using default settings (scoring parameters: BLOSUM62 matrix; gap costs: existence 11 and extension1; expected threshold 10, word size 6) ( Supplementary Tables S1–S5 ). Conserved domains of protein homologs were further scanned and checked using the NCBI Conserved Domain Search using defined options (database: CDD v3.17 – 52,910 PSSMs, expected value threshold 0.01) [ 61 , 62 , 63 , 64 ]. Genome sequences and amino acid sequences of relevant protein homologs were retrieved from the NCBI database. OrfX1, OrfX2, OrfX3, and P47 protein sequences from C. botulinum (strains Beluga, Kyoto-F, Langeland, Mauritius, CDC_297, 111, Mfbjulcb3), C. baratii (strain Sullivan), E. faecium (strain 3G1_DIV0629), and P. bifermentans subsp. malaysia (strain Pbm) were included in the OrfX-P47 dataset for further comparative analysis. Search using the protein family corresponding to Clostridium P47 superfamily (Pfam ID: PF06597) was performed in Annotree [ 65 ]. Protein sequences from the genes located in the vicinity of the orfX gene cluster were further investigated for domain conservation (Pfam 32.0 search) [ 66 ], sequence homology (BLAST) [ 60 ], and structural homology (Phyre2) [ 67 ]. Phylogenetic distributions of the different bacterial genomes analyzed in this work were positioned within the tree generated by Annotree (taxonomic order level) [ 65 ]. The list of genome sequences analyzed in this study is shown in Table 1 . 4.2. Comparative Sequence Analysis, Motif-Based Sequence Analysis, and Phylogenetic Tree Analysis Sequence alignment based on ClustalW algorithm (gap opening penalty 10, gap extension penalty 0.1, protein weight matrix BLOSUM) [ 68 ] was computed in MEGA7 [ 69 ]. Sequence alignment was used to perform principal component analysis (PCA) in JalView v2.10.5 [ 70 ]. Maximum-likelihood phylogenetic trees were generated in MEGA7 with custom options (nearest-neighbor-interchange, Jones-Thornton model, 250 iterations) [ 69 ]. Putative motifs conserved among protein sequences were further analyzed using the MEME suite v.5.0.5 (MEME tool, classic mode, site distribution set to zero or one per sequence, number of motifs set to 30) [ 71 ]. 4.3. Sequence Analysis of the Putative Toxin Gene Cluster in Bacillus sp. 2SH and Structural Modeling Protein BLAST search, Pfam 32.0 search, and Phyre2 analysis of the gene (tentatively named BoNT/NTNH-like A component, BNA, WP_137842862.1) downstream p47 in Bacillus sp. 2SH revealed sequence similarities with NTNH type B of C. botulinum (BAQ12789.1, 30.03% identity, E value 8 × 10 −104 , query cover 90%) and other NTNH homologs. The amino acid sequence of BNA was subsequently aligned with a dataset of BoNT, NTNH, and TeNT amino acid sequences retrieved from public databases. Sequence alignment was performed based on ClustalW algorithm (gap opening penalty 10, gap extension penalty 0.1, protein weight matrix BLOSUM) [ 68 ] in MEGA7 [ 69 ]. Maximum-likelihood phylogenetic trees were generated in MEGA7 as described above [ 69 ]. Initial protein homology modeling was performed using Phyre2 (intensive modeling mode) [ 67 ] and SWISS MODEL [ 72 , 73 , 74 , 75 , 76 , 77 ]. Structural modeling was also carried out using I-TASSER v5.1 [ 78 , 79 ]. BNA model and relevant structural templates were visualized using PyMOL [ 80 ].

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2044511/

Mannheimia haemolytica Leukotoxin Binds to Lipid Rafts in Bovine Lymphoblastoid Cells and Is Internalized in a Dynamin-2- and Clathrin-Dependent Manner ▿

Mannheimia haemolytica is the principal bacterial pathogen of the bovine respiratory disease complex. Its most important virulence factor is a leukotoxin (LKT), which is a member of the RTX family of exotoxins produced by many gram-negative bacteria. Previous studies demonstrated that LKT binds to the β 2 -integrin LFA-1 (CD11a/CD18) on bovine leukocytes, resulting in cell death. In this study, we demonstrated that depletion of lipid rafts significantly decreases LKT-induced bovine lymphoblastoid cell (BL-3) death. After binding to BL-3 cells, some of the LKT relocated to lipid rafts in an LFA-1-independent manner. We hypothesized that after binding to LFA-1 on BL-3 cells, LKT moves to lipid rafts and clathrin-coated pits via a dynamic process that results in LKT internalization and cytotoxicity. Knocking down dynamin-2 by small interfering RNA reduced both LKT internalization and cytotoxicity. Similarly, expression of dominant negative Eps15 protein expression, which is required for clathrin coat formation, reduced LKT internalization and LKT-mediated cytotoxicity to BL-3 cells. Finally, we demonstrated that inhibiting actin polymerization reduced both LKT internalization and LKT-mediated cytotoxicity. These results suggest that both lipid rafts and clathrin-mediated mechanisms are important for LKT internalization and cytotoxicity in BL-3 cells and illustrate the complex nature of LKT internalization by the cytoskeletal network.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5435337/

Effect of dietary restriction and subsequent re-alimentation on the transcriptional profile of bovine ruminal epithelium

Compensatory growth (CG) is utilised worldwide in beef production systems as a management approach to reduce feed costs. However the underlying biology regulating the expression of CG remains to be fully elucidated. The objective of this study was to examine the effect of dietary restriction and subsequent re-alimentation induced CG on the global gene expression profile of ruminal epithelial papillae. Holstein Friesian bulls (n = 60) were assigned to one of two groups: restricted feed allowance (RES; n = 30) for 125 days (Period 1) followed by ad libitum access to feed for 55 days (Period 2) or (ii) ad libitum access to feed throughout (ADLIB; n = 30). At the end of each period, 15 animals from each treatment were slaughtered and rumen papillae harvested. mRNA was isolated from all papillae samples collected. cDNA libraries were then prepared and sequenced. Resultant reads were subsequently analysed bioinformatically and differentially expressed genes (DEGs) are defined as having a Benjamini-Hochberg P value of 30 ng/μl. After quality control procedures, individual RNA-seq libraries were pooled based on their respective sample-specific-6bp adaptors and sequenced at 100 bp/sequence on an Illumina HiSeq 2000 generating single-end reads. Read alignment and abundance calculation Preliminary quality control analysis was carried out using FASTQC software (version 0.10.0). FASTX-Toolkit (v0.0.13) was then used to trim 3' adaptor sequences. Trimmed reads were then subsequently aligned to the UMD3.1 Bos Taurus genome assembly using Tophat (v2.0.9) and Bowtie2 ultra-fast short read alignment software (v2.1.0). The software package HTSeq (v0.5.4p5) ( http://pypi.python.org/pypi/HTSeq ) was employed to calculate the abundance of mRNAs for all annotated genes from the ENSEMBL v74 annotation of the bovine genome. The number of read counts mapping to each annotated gene from HTSeq was then collated into a single file and subsequently used to identify differentially expressed genes (DEGs). Identification of DEGs The R (v2.14.1) based Bioconductor package, EdgeR (v3.4.1), which uses a negative binomial distribution model to account for both biological and technical variability, was employed to identify statistically significant DEGs. Genes with low read counts across all libraries were excluded from subsequent analysis. The analysis was undertaken using moderated tagwise dispersions. DEGs were defined as having a Benjamini-Hochberg corrected P value of 30 ng/μl. After quality control procedures, individual RNA-seq libraries were pooled based on their respective sample-specific-6bp adaptors and sequenced at 100 bp/sequence on an Illumina HiSeq 2000 generating single-end reads. Read alignment and abundance calculation Preliminary quality control analysis was carried out using FASTQC software (version 0.10.0). FASTX-Toolkit (v0.0.13) was then used to trim 3' adaptor sequences. Trimmed reads were then subsequently aligned to the UMD3.1 Bos Taurus genome assembly using Tophat (v2.0.9) and Bowtie2 ultra-fast short read alignment software (v2.1.0). The software package HTSeq (v0.5.4p5) ( http://pypi.python.org/pypi/HTSeq ) was employed to calculate the abundance of mRNAs for all annotated genes from the ENSEMBL v74 annotation of the bovine genome. The number of read counts mapping to each annotated gene from HTSeq was then collated into a single file and subsequently used to identify differentially expressed genes (DEGs). Identification of DEGs The R (v2.14.1) based Bioconductor package, EdgeR (v3.4.1), which uses a negative binomial distribution model to account for both biological and technical variability, was employed to identify statistically significant DEGs. Genes with low read counts across all libraries were excluded from subsequent analysis. The analysis was undertaken using moderated tagwise dispersions. DEGs were defined as having a Benjamini-Hochberg corrected P value of 0.05) [ 11 ]. Read mapping and differential gene expression Approximately 86% of sequencing reads (after trimming) were aligned to the bovine genome and 73% of those that aligned were mapped to the gene space. The bovine reference genome (UMD3.1) contains 26,740 gene transcripts. At the end of dietary restriction in Period 1, the number of genes that had mapped reads was 12,634, whereas following 55 days of re-alimentation in Period 2, 12,711 genes had reads mapping to them. Using the bioconductor package EdgeR, 64 genes were identified as differentially expressed between RES and ADLIB animals at the end of Period 1. These were manifested as up-regulation of 40 and down-regulation of 24 genes in RES animals compared to ADLIB treatment. Further details of these genes are provided in S1 Table . Following 55 days of subsequent ad libitum feeding only one gene was differentially expressed between RES and ADLIB. BNBD10 , a beta defensin gene, was down-regulated in RES animals compared to ADLIB animals at the end of Period 2. Additionally, when the data were examined within the RES treatment, 411 genes were identified as differentially expressed between periods 1 and 2. From this latter analysis 226 genes were down-regulated and 185 up-regulated in RES during Period 2 compared with Period 1. Further details of these genes are provided in S2 Table . Data pertaining to the ADLIB group across time resulted in differential expression of 5 genes, these included MAOB and DNAJC6 which both had greater expression at the end of Period 2 compared to Period 1 and NQO2 , ADAMTSL3 and CCL22 , which had lower expression at the end of Period 2 compared to Period 1. These RNA-seq data have been deposited in the NCBI's Gene Expression Omnibus [ 31 ] and are accessible through GEO Series accession number GSE89162. Pathway analysis DEGs were analysed and separated according to their biological functions using IPA software. At the end of Period 1, genes involved in processes including cellular signalling and interaction, protein synthesis and gene expression were differentially expressed. The direction of fold change for DEGs within these processes indicated an overall down-regulation of these cellular functions in rumen papillae in response to dietary restriction. During CG of RES papillae in Period 2 genes coding for proteins involved in cellular survival/organisation and protein folding were differentially expressed compared with dietary restriction in Period 1. The direction of fold change of these genes suggested an up-regulation of these processes during CG. Further details of the genes involved in these processes are outlined in Tables 1 , 2 and 3 ( Table 1 : gene expression and protein folding; Table 2 : cellular interactions and organisation; Table 3 : immune response). Details of functional processes affected by dietary restriction and subsequent re-alimentation induced CG are presented in Figs 1 and 2 respectively. 10.1371/journal.pone.0177852.t001 Table 1 Genes involved in gene expression and protein folding functions found to be differentially expressed in ruminal papillae following: A period of (i) dietary restriction and (ii) re-alimentation induced compensatory growth. Gene ID Gene name Fold change 1 Dietary restriction CRYAB Crystallin, alpha B -1.708 HSPB8 Heat shock 22kDa protein 8 -1.466 HSPH1 Heat shock 105kDa/110kDa protein 1 -1.628 SATB1 SATB homeobox 1 2.354 ZC3H12A Zinc finger CCCH-type containing 12A 1.748 Compensatory growth AHSA1 AHA1, activator of heat shock 90kDa protein ATPase homolog 1 (yeast) 1.491 DNAJB4 DnaJ (Hsp40) homolog, subfamily B, member 4 1.304 HSPA8 Heat shock 70kDa protein 8 1.543 HSPB8 Heat shock 22kDa protein 8 1.739 HSPD1 Heat shock 60kDa protein 1 (chaperonin) 1.306 MDN1 MDN1, midasin homolog (yeast) 1.386 CCT2 Chaperonin containing TCP1, subunit 2 (beta) 1.262 HSP90AA1 Heat shock protein 90kDa alpha (cytosolic), class A member 1 1.697 HSP90AB1 Heat shock protein 90kDa alpha (cytosolic), class B member 1 1.284 HSPE1 Heat shock 10kDa protein 1 1.375 PPID Peptidylprolyl isomerase D 1.392 STIP1 Stress-induced phosphoprotein 1 1.436 EMG1 EMG1 N1-specific pseudouridine methyltransferase 1.267 FOXN1 Forkhead box N1 2.581 FOXP4 Forkhead box P4 1.403 INTS3 Integrator complex subunit 3 1.28 KHDRBS3 KH domain containing, RNA binding, signal transduction associated 3 1.488 PRCC Papillary renal cell carcinoma (translocation-associated) 1.253 EIF4G2 Eukaryotic translation initiation factor 4 gamma, 2 1.250 EIF4G3 , Eukaryotic translation initiation factor 4 gamma, 3 1.251 ELL2 Elongation factor, RNA polymerase II, 2 1.423 HIST1H2AC Histone cluster 1, H2ac 2.538 HIST1H2BD Histone cluster 1, H2bd 1.595 HIST1H2BN Histone cluster 1, H2bn 1.618 HISTH2BO Histone cluster 1, H2bo 1.331 HIST2H4A Histone cluster 2, H4a 1.634 KAT2A K(lysine) acetyltransferase 2A 1.250 1 Fold changes are as follows: (i) dietary restriction: up or down in restricted fed animals compared with ad libitum control animals during dietary restriction at the end of Period 1; (ii) compensatory growth: up or down in restricted Period 2 animals compared to restricted Period 1 animals during compensatory growth. 10.1371/journal.pone.0177852.t002 Table 2 Genes involved in cellular interactions and organisation differentially expressed in rumen papillae following a period of (i) dietary restriction and (ii) re-alimentation induced compensatory growth. Gene ID Gene name Fold change 1 Dietary restriction CDH2 Cadherin 2, type 1, N-cadherin (neuronal) -2.895 DSG1 Desmoglein 1 -4.632 Compensatory growth ANTXR1 Anthrax toxin receptor 1 1.347 CEP97 Centrosomal protein 97kDa 1.461 FAT4 FAT atypical cadherin 4 1.609 PCDH12 Protocadherin 12 1.715 PCDH7 Protocadherin 7 1.666 ITGA8 Integrin, alpha 8 1.746 NRG1 Neuregulin 1 1.512 RELN Reelin 1.647 SMAGP Small cell adhesion glycoprotein 1.303 THBS4 Thrombospondin 4 1.818 SLC1A5 Solute carrier family 1 (neutral amino acid transporter), member 5 1.342 SLC22A17 Solute carrier family 22, member 17 2.296 SLC25A15 Solute carrier family 25 (mitochondrial carrier; ornithine transporter) member 15 1.276 SLC25A26 Solute carrier family 25 (S-adenosylmethionine carrier), member 26 1.251 SLC30A6 Solute carrier family 30 (zinc transporter), member 6 1.295 SLC4A7 Solute carrier family 4, sodium bicarbonate cotransporter, member 7 1.397 SLC6A9 Solute carrier family 6 (neurotransmitter transporter, glycine), member 9 1.349 SLC9A1 Solute carrier family 9, subfamily A (NHE1, cation proton antiporter 1), member 1 1.318 CACNA1G Calcium channel, voltage-dependent, T type, alpha 1G subunit 1.975 KCNC4 Potassium channel, voltage gated Shaw related subfamily C, member 4 1.492 1 Fold changes are as follows: (i) dietary restriction: up or down in restricted fed animals compared with ad libitum control animals during dietary restriction at the end of Period 1; (ii) compensatory growth: up or down in restricted Period 2 animals compared to restricted Period 1 animals during compensatory growth. 10.1371/journal.pone.0177852.t003 Table 3 Genes involved in immune response differentially expressed in rumen papillae following a period of (i) dietary restriction and (ii) re-alimentation induced compensatory growth. Gene ID Gene name Fold change 1 Dietary restriction IL17A Interleukin 17A 3.707 LBP Lipopolysaccharide binding protein 1.74 Compensatory growth BDKRB1 Bradykinin receptor B1 -1.971 CHI3L1 Chitinase 3-like 1 (cartilage glycoprotein-39) -3.03 HPGD Hydroxyprostaglandin dehydrogenase 15-(NAD) -1.674 LTA4H Leukotriene A4 hydrolase -1.276 C5AR2 Complement component 5a receptor 2 -1.605 CD59 CD59 molecule, complement regulatory protein -1.458 CCL19 Chemokine (C-C motif) ligand 19 -4.237 CCL20 Chemokine (C-C motif) ligand 20 -2.965 CXCL12 Chemokine (C-X-C motif) ligand 12 -2.189 CXCL17 Chemokine (C-X-C motif) ligand 17 -2.691 CXCL2 Chemokine (C-X-C motif) ligand 2 -2.957 CXCR4 Chemokine (C-X-C motif) receptor 4 -1.646 LYZ Lysozyme -3.144 SGSH N-sulfoglucosamine sulfohydrolase -1.261 CYBA Cytochrome b-245, alpha polypeptide -1.418 1 Fold changes are as follows: (i) dietary restriction: up or down in restricted fed animals compared with ad libitum control animals during dietary restriction at the end of Period 1; (ii) compensatory growth: up or down in restricted Period 2 animals compared to restricted Period 1 animals during compensatory growth. 10.1371/journal.pone.0177852.g001 Fig 1 Classification of differentially expressed genes according to molecular and cellular function, most significantly affected by restricted feeding in rumen papillae at the end of Period 1. The bars indicate the likelihood [-log(P value)] that the specific molecular and cellular function was affected by restricted feeding compared with other functions represented in the list of differentially expressed genes. 10.1371/journal.pone.0177852.g002 Fig 2 Classification of differentially expressed genes according to molecular and cellular function, most significantly affected by re-alimentation and compensatory growth in rumen papillae. The bars indicate the likelihood [-log(P value)] that the specific molecular and cellular function was affected by re-alimentation induced CG compared with other functions represented in the list of differentially expressed genes. Using IPA software, a total of five networks were identified for DEGs at the end of Period 1 ( S3 Table ), with 25 networks identified in rumen papillae of animals undergoing CG (RES Period 2 v RES Period 1; S4 Table ). Network 6 was of particular interest in rumen papillae undergoing CG. This network consisted of genes involved in carbohydrate metabolism, small molecule biochemistry and cellular assembly and organisation and details are presented in Fig 3 . 10.1371/journal.pone.0177852.g003 Fig 3 Metabolism and cellular assembly/organisation network in rumen papillae following re-alimentation induced compensatory growth (Network 6: Carbohydrate metabolism, small molecule biochemistry and cellular assembly and organisation). The network is displayed graphically as nodes (genes). The node colour intensity indicates the expression of genes; with red representing up-regulation and green, down-regulation in animals following a period of re-alimentation induced compensatory growth relative to following a period of dietary restriction. Animal performance The animal performance data pertaining to samples utilised in this study are presented and discussed in detail by Keogh et al. [ 11 ]. Briefly, following a period of 125 days of dietary restriction, there was a 161 kg difference in bodyweight between RES (mean (SEM); 442 (6.67) kg) and ADLIB (603 (7.21) kg) animals (P 0.05) [ 11 ]. Read mapping and differential gene expression Approximately 86% of sequencing reads (after trimming) were aligned to the bovine genome and 73% of those that aligned were mapped to the gene space. The bovine reference genome (UMD3.1) contains 26,740 gene transcripts. At the end of dietary restriction in Period 1, the number of genes that had mapped reads was 12,634, whereas following 55 days of re-alimentation in Period 2, 12,711 genes had reads mapping to them. Using the bioconductor package EdgeR, 64 genes were identified as differentially expressed between RES and ADLIB animals at the end of Period 1. These were manifested as up-regulation of 40 and down-regulation of 24 genes in RES animals compared to ADLIB treatment. Further details of these genes are provided in S1 Table . Following 55 days of subsequent ad libitum feeding only one gene was differentially expressed between RES and ADLIB. BNBD10 , a beta defensin gene, was down-regulated in RES animals compared to ADLIB animals at the end of Period 2. Additionally, when the data were examined within the RES treatment, 411 genes were identified as differentially expressed between periods 1 and 2. From this latter analysis 226 genes were down-regulated and 185 up-regulated in RES during Period 2 compared with Period 1. Further details of these genes are provided in S2 Table . Data pertaining to the ADLIB group across time resulted in differential expression of 5 genes, these included MAOB and DNAJC6 which both had greater expression at the end of Period 2 compared to Period 1 and NQO2 , ADAMTSL3 and CCL22 , which had lower expression at the end of Period 2 compared to Period 1. These RNA-seq data have been deposited in the NCBI's Gene Expression Omnibus [ 31 ] and are accessible through GEO Series accession number GSE89162. Pathway analysis DEGs were analysed and separated according to their biological functions using IPA software. At the end of Period 1, genes involved in processes including cellular signalling and interaction, protein synthesis and gene expression were differentially expressed. The direction of fold change for DEGs within these processes indicated an overall down-regulation of these cellular functions in rumen papillae in response to dietary restriction. During CG of RES papillae in Period 2 genes coding for proteins involved in cellular survival/organisation and protein folding were differentially expressed compared with dietary restriction in Period 1. The direction of fold change of these genes suggested an up-regulation of these processes during CG. Further details of the genes involved in these processes are outlined in Tables 1 , 2 and 3 ( Table 1 : gene expression and protein folding; Table 2 : cellular interactions and organisation; Table 3 : immune response). Details of functional processes affected by dietary restriction and subsequent re-alimentation induced CG are presented in Figs 1 and 2 respectively. 10.1371/journal.pone.0177852.t001 Table 1 Genes involved in gene expression and protein folding functions found to be differentially expressed in ruminal papillae following: A period of (i) dietary restriction and (ii) re-alimentation induced compensatory growth. Gene ID Gene name Fold change 1 Dietary restriction CRYAB Crystallin, alpha B -1.708 HSPB8 Heat shock 22kDa protein 8 -1.466 HSPH1 Heat shock 105kDa/110kDa protein 1 -1.628 SATB1 SATB homeobox 1 2.354 ZC3H12A Zinc finger CCCH-type containing 12A 1.748 Compensatory growth AHSA1 AHA1, activator of heat shock 90kDa protein ATPase homolog 1 (yeast) 1.491 DNAJB4 DnaJ (Hsp40) homolog, subfamily B, member 4 1.304 HSPA8 Heat shock 70kDa protein 8 1.543 HSPB8 Heat shock 22kDa protein 8 1.739 HSPD1 Heat shock 60kDa protein 1 (chaperonin) 1.306 MDN1 MDN1, midasin homolog (yeast) 1.386 CCT2 Chaperonin containing TCP1, subunit 2 (beta) 1.262 HSP90AA1 Heat shock protein 90kDa alpha (cytosolic), class A member 1 1.697 HSP90AB1 Heat shock protein 90kDa alpha (cytosolic), class B member 1 1.284 HSPE1 Heat shock 10kDa protein 1 1.375 PPID Peptidylprolyl isomerase D 1.392 STIP1 Stress-induced phosphoprotein 1 1.436 EMG1 EMG1 N1-specific pseudouridine methyltransferase 1.267 FOXN1 Forkhead box N1 2.581 FOXP4 Forkhead box P4 1.403 INTS3 Integrator complex subunit 3 1.28 KHDRBS3 KH domain containing, RNA binding, signal transduction associated 3 1.488 PRCC Papillary renal cell carcinoma (translocation-associated) 1.253 EIF4G2 Eukaryotic translation initiation factor 4 gamma, 2 1.250 EIF4G3 , Eukaryotic translation initiation factor 4 gamma, 3 1.251 ELL2 Elongation factor, RNA polymerase II, 2 1.423 HIST1H2AC Histone cluster 1, H2ac 2.538 HIST1H2BD Histone cluster 1, H2bd 1.595 HIST1H2BN Histone cluster 1, H2bn 1.618 HISTH2BO Histone cluster 1, H2bo 1.331 HIST2H4A Histone cluster 2, H4a 1.634 KAT2A K(lysine) acetyltransferase 2A 1.250 1 Fold changes are as follows: (i) dietary restriction: up or down in restricted fed animals compared with ad libitum control animals during dietary restriction at the end of Period 1; (ii) compensatory growth: up or down in restricted Period 2 animals compared to restricted Period 1 animals during compensatory growth. 10.1371/journal.pone.0177852.t002 Table 2 Genes involved in cellular interactions and organisation differentially expressed in rumen papillae following a period of (i) dietary restriction and (ii) re-alimentation induced compensatory growth. Gene ID Gene name Fold change 1 Dietary restriction CDH2 Cadherin 2, type 1, N-cadherin (neuronal) -2.895 DSG1 Desmoglein 1 -4.632 Compensatory growth ANTXR1 Anthrax toxin receptor 1 1.347 CEP97 Centrosomal protein 97kDa 1.461 FAT4 FAT atypical cadherin 4 1.609 PCDH12 Protocadherin 12 1.715 PCDH7 Protocadherin 7 1.666 ITGA8 Integrin, alpha 8 1.746 NRG1 Neuregulin 1 1.512 RELN Reelin 1.647 SMAGP Small cell adhesion glycoprotein 1.303 THBS4 Thrombospondin 4 1.818 SLC1A5 Solute carrier family 1 (neutral amino acid transporter), member 5 1.342 SLC22A17 Solute carrier family 22, member 17 2.296 SLC25A15 Solute carrier family 25 (mitochondrial carrier; ornithine transporter) member 15 1.276 SLC25A26 Solute carrier family 25 (S-adenosylmethionine carrier), member 26 1.251 SLC30A6 Solute carrier family 30 (zinc transporter), member 6 1.295 SLC4A7 Solute carrier family 4, sodium bicarbonate cotransporter, member 7 1.397 SLC6A9 Solute carrier family 6 (neurotransmitter transporter, glycine), member 9 1.349 SLC9A1 Solute carrier family 9, subfamily A (NHE1, cation proton antiporter 1), member 1 1.318 CACNA1G Calcium channel, voltage-dependent, T type, alpha 1G subunit 1.975 KCNC4 Potassium channel, voltage gated Shaw related subfamily C, member 4 1.492 1 Fold changes are as follows: (i) dietary restriction: up or down in restricted fed animals compared with ad libitum control animals during dietary restriction at the end of Period 1; (ii) compensatory growth: up or down in restricted Period 2 animals compared to restricted Period 1 animals during compensatory growth. 10.1371/journal.pone.0177852.t003 Table 3 Genes involved in immune response differentially expressed in rumen papillae following a period of (i) dietary restriction and (ii) re-alimentation induced compensatory growth. Gene ID Gene name Fold change 1 Dietary restriction IL17A Interleukin 17A 3.707 LBP Lipopolysaccharide binding protein 1.74 Compensatory growth BDKRB1 Bradykinin receptor B1 -1.971 CHI3L1 Chitinase 3-like 1 (cartilage glycoprotein-39) -3.03 HPGD Hydroxyprostaglandin dehydrogenase 15-(NAD) -1.674 LTA4H Leukotriene A4 hydrolase -1.276 C5AR2 Complement component 5a receptor 2 -1.605 CD59 CD59 molecule, complement regulatory protein -1.458 CCL19 Chemokine (C-C motif) ligand 19 -4.237 CCL20 Chemokine (C-C motif) ligand 20 -2.965 CXCL12 Chemokine (C-X-C motif) ligand 12 -2.189 CXCL17 Chemokine (C-X-C motif) ligand 17 -2.691 CXCL2 Chemokine (C-X-C motif) ligand 2 -2.957 CXCR4 Chemokine (C-X-C motif) receptor 4 -1.646 LYZ Lysozyme -3.144 SGSH N-sulfoglucosamine sulfohydrolase -1.261 CYBA Cytochrome b-245, alpha polypeptide -1.418 1 Fold changes are as follows: (i) dietary restriction: up or down in restricted fed animals compared with ad libitum control animals during dietary restriction at the end of Period 1; (ii) compensatory growth: up or down in restricted Period 2 animals compared to restricted Period 1 animals during compensatory growth. 10.1371/journal.pone.0177852.g001 Fig 1 Classification of differentially expressed genes according to molecular and cellular function, most significantly affected by restricted feeding in rumen papillae at the end of Period 1. The bars indicate the likelihood [-log(P value)] that the specific molecular and cellular function was affected by restricted feeding compared with other functions represented in the list of differentially expressed genes. 10.1371/journal.pone.0177852.g002 Fig 2 Classification of differentially expressed genes according to molecular and cellular function, most significantly affected by re-alimentation and compensatory growth in rumen papillae. The bars indicate the likelihood [-log(P value)] that the specific molecular and cellular function was affected by re-alimentation induced CG compared with other functions represented in the list of differentially expressed genes. Using IPA software, a total of five networks were identified for DEGs at the end of Period 1 ( S3 Table ), with 25 networks identified in rumen papillae of animals undergoing CG (RES Period 2 v RES Period 1; S4 Table ). Network 6 was of particular interest in rumen papillae undergoing CG. This network consisted of genes involved in carbohydrate metabolism, small molecule biochemistry and cellular assembly and organisation and details are presented in Fig 3 . 10.1371/journal.pone.0177852.g003 Fig 3 Metabolism and cellular assembly/organisation network in rumen papillae following re-alimentation induced compensatory growth (Network 6: Carbohydrate metabolism, small molecule biochemistry and cellular assembly and organisation). The network is displayed graphically as nodes (genes). The node colour intensity indicates the expression of genes; with red representing up-regulation and green, down-regulation in animals following a period of re-alimentation induced compensatory growth relative to following a period of dietary restriction. Discussion The CG phenomenon, typically expressed upon re-alimentation following a prior period of dietary restriction has been associated with improved feed efficiency in a number of cattle studies [ 4 , 8 , 32 ] including for the animals employed in the current study [ 11 ]. Moreover, Sainz et al. [ 14 ] and Keogh et al. [ 11 ] both reported greater feed intake in compensating animals following re-alimentation. Greater feed intake combined with greater animal ADG and the rapid CG of the rumen suggest changes in the activity of the rumen or potentially differences in digestibility capacity or epithelial morphology within the rumen during CG may play a role in accelerated growth. Results from our own study are consistent with previous studies, showing compensating animals consumed a greater amount of feed per unit of body weight when compared to their ad libitum fed counterparts during Period 2 [ 11 ]. Furthermore, increases in digestibility have also been reported in cattle undergoing CG [ 33 ]. Indeed, Sun et al. [ 23 ] observed differences in rumen papillae height, width and surface area in goats following a 48 day period of dietary restriction compared to those that had not been diet restricted. However, following a subsequent period of re-alimentation induced CG, lasting 62 days there were no longer any detectable differences in rumen epithelium morphology between animals that had undergone dietary restriction and subsequent CG compared with their unrestricted counterparts [ 23 ]. Moreover, the rumen has repeatedly been shown to be one of the most responsive organs to both dietary restriction and also subsequent CG, as evidenced by both our own work [ 11 ] as well as by that of others [ 7 , 8 ]. This may be due to the high metabolic rate associated with this organ, with a reduction in rumen size following a period of dietary restriction allowing for a reduction in associated basal metabolic energy requirements of the organ [ 8 , 11 , 34 ]. Indeed, a lowered basal metabolic rate is thought to sustain through to at least the early stages of re-alimentation and contribute to CG by allowing more energy to be partitioned towards growth as opposed to maintenance requirements [ 11 , 34 ]. However, although this tissue is clearly affected by both dietary restriction and CG, knowledge of the underlying biology regulating the expression of CG in rumen epithelial or indeed any tissue of the gastrointestinal tract is still lacking. Therefore, the objective of this study was to quantify and characterize by gene function, the transcriptional changes in rumen papillae of beef cattle in response to both nutrient restriction and subsequent CG and also to determine the contribution of these changes to overall animal CG. This was achieved through an examination of DEGs in rumen papillae following a period of dietary restriction and also a period of subsequent re-alimentation compared to rumen papillae of animals that were fed continuously. Additionally, sequencing data were analysed within treatment group to further assess the effect of CG on transcriptional changes within rumen papillae. The large difference in DEGs between RES and ADLIB groups when analysed within treatment across time (RES: 411 DEGs, ADLIB: 5 DEGs) suggests that the RES within treatment group analysis is reflective of CG and not of normal growth as described in the ADLIB DEG profile. A greater knowledge of molecular changes occurring during CG of highly metabolically important organs such as the rumen may facilitate more accurate identification of animals with improved CG potential and thus the possible incorporation of this economically important information into genomically assisted cattle breeding programs. Gene transcription and protein folding A reduction in feed intake is typically paralleled by a reduction in growth and overall cellular functions. Indeed, this was apparent in the papillae of RES animals following a period of dietary restriction, where genes involved in gene transcription and protein folding tended to be down-regulated compared with ADLIB animals. Specifically these DEGs were manifested as down-regulation of genes coding for proteins involved in chaperone functionality including CRYAB , HSPB8 and HSPH1 . The CRYAB protein displays chaperone-like activity and functions in preventing aggregation of various proteins under a wide range of conditions [ 35 ]. We also found this gene to be down-regulated in the skeletal muscle tissue of these same cattle during dietary restriction [ 25 ]. Both HSPB8 and HSPH1 code for heat shock proteins which also function in the prevention of aggregation of denatured proteins in cells [ 36 ]. HSPH1 was also found to be down-regulated in skeletal muscle of cattle following a period of dietary restriction [ 25 ]. Furthermore, up-regulation of both SATB1 and ZC3H12A was also apparent in the rumen papillae of RES animals at the same time-point. Both of these genes code for proteins involved in repressing transcription, thereby causing down-regulation of gene expression processes [ 37 , 38 ]. SATB1 encodes a matrix protein, which functions to recruit chromatic remodelling factors in order to regulate chromatin structure and gene expression and ultimately functions in transcriptional repression and gene silencing [ 39 ]. Consistent with this ZC3H12A displays RNase activity and functions in selectively degrading specific target mRNA species [ 40 ]. We also found ZC3H12A to be up-regulated in hepatic tissue of these cattle following a period of dietary restriction [ 26 ]. Down-regulation of genes involved in these processes following a period of dietary restriction may be reflective of a reduced requirement for nutrient processing and metabolism, coinciding with lower animal ADG and weight of the rumen complex, at the end of Period 1 [ 11 ]. Moreover, following a 125 day period of dietary restriction, when compared with a reference slaughter group at the start of dietary restriction, the proportional weight of the rumen in feed restricted animals was found to be lower [ 11 ], further evidencing a reduced metabolic requirement of this organ in response to dietary restriction. Lower expression of genes involved in cellular metabolism, following dietary restriction was also apparent in liver tissue of the same animals used in the current study [ 26 ]. While the aforementioned biological processes were down-regulated during dietary restriction, upon re-alimentation up-regulation of these functions was apparent which coincided with greater rumen and overall body growth rates [ 11 ]. Up-regulation of these processes during re-alimentation also coincided with a greater capacity for growth and requirement for cellular metabolism in the rumen tissue as well as in other organs within the body. For example, increased expression of genes associated with metabolism during CG was previously described in the hepatic tissue of the animals used in the current study [ 26 ]. Indeed greater transcription of genes coding for proteins involved in gene expression was evident in papillae of animals displaying CG. Greater expression of these genes may be necessary in order to allow for increased production of proteins to accommodate the increased nutrient availability and metabolic demands of digestion, absorption and ultimately tissue growth. Genes involved in protein folding included those coding for chaperone proteins: AHSA1 ; DNAJB4 ; HSPA8 ; HSPB8 ; HSPD1 ; MDN1 ; as well as those requiring input of ATP: CCT2 ; HSP90AA1 ; HSP90AB1 ; HSPE1 . These genes have also been shown to display greater expression during CG in other tissues and organs, including skeletal muscle ( AHSA1 , DNAJB4 , HSPA8 , HSPB8 and HSPD1 ; [ 25 ]; and liver ( HSPA8 , HSPB8 and HSPD1 ; [ 10 ]). Moreover, HSPA8 , HSPB8 and HSPD1 were also up-regulated in the liver of feed efficient cattle [ 41 ]. These genes may be important in relation to improved feed efficiency consistent with that observed for cattle undergoing CG including those in the current study [ 4 , 11 ]. Greater expression of CCT2 and HSPE1 was also reported in skeletal muscle of our animals while undergoing CG [ 25 ]. Furthermore, up-regulation of HSP90AA1 and HSP90AB1 was also apparent in both hepatic and skeletal muscle tissues of cattle undergoing CG [ 10 , 25 ]. Greater expression of PPID , a protein that functions in accelerating protein folding [ 42 ] as well as STIP1 which regulates both the conformation and ATPase cycles of HSP70 and HSP90 molecular chaperones [ 43 ] was also apparent in rumen papillae of cattle undergoing CG in the current study. Both of these genes were also up-regulated in skeletal muscle tissue of cattle undergoing CG [ 25 ]. Overall these results suggest that increased cellular protein folding activity is required within the rumen papillae as part of the adaption to an increased dietary intake and is consistent with the heightened metabolic state typical of animals undergoing re-alimentation induced CG [ 11 , 12 , 44 , 45 ]. Indeed, this may be a necessary response in order to cope with the typically elevated rate of metabolism associated with greater feed consumption [ 46 ], which appears to be a primary driver of whole animal CG [ 8 , 11 , 14 ]. However, further studies are required to assess the metabolic state of the rumen and indeed other metabolic organs in response to both dietary restriction and CG. In addition to an increase in the expression of genes coding for chaperone and protein folding cellular machinery, up-regulation of genes involved in transcriptional activity was also observed in rumen papillae of animals undergoing CG, again this coincided with greater reticulo-rumen and whole body growth [ 11 ]. Overall, genes coding for proteins involved in transcription ( EMG1 , FOXN1 , FOXP4 , INTS3 ), splicing ( KHDRBS3 , PRCC ) and translation ( EIF4G2 , EIF4G3 , ELL2 ) were up-regulated in rumen papillae of cattle undergoing CG. Transcriptional genes differentially expressed included EMG1 , which encodes a protein involved in ribosome biogenesis [ 47 ], two FoxO proteins which are involved in the regulation of gene transcription [ 48 ] and a subunit of the integrator complex of RNA polymerase II ( INTS3 ; [ 49 ]). Both EIF4G2 and EIF4G3 code for proteins involved in the eukaryotic translation initiation factor and function in the recognition of the mRNA cap, and recruitment of mRNA to the ribosome [ 50 ]. Up-regulation of EIF4G2 was also reported in the data of Connor et al. [ 10 ] in hepatic tissue of cattle undergoing CG. The elongation factor component ELL2 was also up-regulated in skeletal muscle tissue of the same cattle used in the current study [ 25 ]. This gene codes for a complex which is required to increase the catalytic rate of RNA polymerase II transcription [ 51 ]. Genes involved in splicing, and the editing of nascent pre-mRNA [ 52 ] were also detected as differentially expressed in rumen papillae of cattle undergoing CG in the current study. These included KHDRBS3 which functions in the regulation of alternative splicing and influences mRNA splice site selection [ 53 ] and PRCC which functions in pre-mRNA splicing [ 54 ]. Moreover, genes coding for histone proteins ( HIST1H2AC , HIST1H2BD , HIST1H2BN , HISTH2BO , HIST2H4A , KAT2A ) were also up-regulated during the same time. Histones are proteins that package and order DNA into structural nucleosomes, playing a role in gene regulation [ 55 ]. Additionally, KAT2A a histone acetlytransferase that functions primarily as a transcriptional activator was also up-regulated in papillae of animals undergoing CG. Genes coding for histone proteins were also detected as up-regulated in skeletal muscle tissue of cattle expressing CG, these included HIST1H2AC , HIST1H2BD and KAT2A [ 25 ]. Collectively, these results suggest an increase in gene expression and associated translational and protein folding activity in rumen papillae epithelia during CG and associated feed efficiency in cattle. A similar effect was also reported in rumen epithelium of feed efficient cattle (low-residual feed intake) [ 56 ]. This is also apparent in the network presented in Fig 3 , where genes associated with metabolism, biochemistry and cellular assembly and organisation were up-regulated. Up-regulation of these cellular processes during rumen papillae CG may be a consequence of a greater nutrient intake during re-alimentation and be necessary for the replenishment of the associated metabolic machinery required for increased digestion and absorption, which ultimately may be contributing to compensatory tissue growth and development, and as stated earlier was also apparent during CG of hepatic tissue of cattle undergoing CG [ 10 , 26 ]. Cellular interactions and organisation Reduced nutrient intake may be consistent with a down-regulation of cellular processes associated with cellular function and organisation [ 57 ]. This has previously been reported in skeletal muscle of cattle after diet restriction [ 25 ]. Following a period of dietary restriction, genes coding for proteins involved in structural components of ruminal epithelial cells were observed to be down-regulated in RES compared to ADLIB animals. Down-regulation of these genes may be due to a lack of requirement for a large ruminal epithelial surface area as a consequence of a reduction in intake and associated digestive processes in the rumen. Moreover, a reduction in epithelial surface area may allow for a reduction in cellular maintenance requirements in an energetically demanding organ such as the rumen. At the end of Period 1, CDH2 and DSG1 were both down-regulated in RES animals compared to ADLIB animals. CDH2 codes for a cadherin, which are a family of transmembrane proteins involved in cellular adhesion [ 58 ]. The encoded protein CDH2, is a calcium dependent cell-cell adhesion glycoprotein [ 59 ]. The gene DSG1 codes for a desmosome protein, which form junctions between certain cell types including epithelial cells [ 58 ]. DSG1 is a calcium-binding trans-membrane glycoprotein component of desmosomes in vertebrate epithelial cells [ 60 ]. It is involved in maintaining the structural integrity of epithelial cells including rumen epithelium and intermediate filaments mediating cell-cell adhesion [ 61 ]. Structural alterations to rumen papillae in response to differences in dietary intake have previously been reported. For example, Steele et al. [ 62 ] observed structural adaptations in rumen epithelium when cows were fed a diet consisting primarily of grain. Moreover, in that study lower expression of DSG1 was reported in response to a high concentrate diet, with expression subsequently greater upon transition to a high forage diet [ 62 ]. Additionally, Sun et al. [ 23 ] observed reduced rumen epithelial height, width and surface area in goats following a 48 day period of dietary restriction. It is logical to expect that the cumulative surface area of papillae and thus weight of the organ itself reflects the prevailing dietary management of an animal. As a consequence of reduced dietary intake, there may be a decreased necessity for ruminal papillae surface area, which may contribute to the reduction in rumen size, as observed in the current study [ 11 ]. In turn, the reduction in rumen size may allow for rumen metabolic rate to be curtailed which in turn could contribute to reduced animal maintenance requirements during dietary restriction. Conversely though, during re-alimentation induced CG, the corollary was observed in the current study, whereby expression of genes coding for proteins involved in cellular interactions and organisation was greater in papillae of RES animals at the end of Period 2 than at the end of Period 1. Genes involved in cellular adhesion ( ANTXR1 , CEP97 , FAT4 , PCDH12 , PCDH7 ), cellular interactions ( IGCA8 , NRG1 , RELN , SMAGP , THBS4 ) and transport ( SLC1A5 , SLC22A17 , SLC25A15 , SLC25A26 , SLC30A6 , SLC4A7 , SLC6A9 , SLC9A1 , CACNA1G , KCNC4 ) were all up-regulated during the CG of ruminal papillae. A similar effect has also been reported in skeletal muscle for the same cattle population used here [ 25 ]. Of note, up-regulation of the following genes PCDH12 and THBS4 as well as two transporter genes, SLC22A17 and SLC25A15 was consistent between the current study for rumen epithelial tissue and our previous study using muscle tissue [ 25 ]. Up-regulation of these processes during CG in ruminal epithelial may have reflected a necessary adaptive requirement for cells to cope with the increase in cellular metabolic activity as a consequence of increased nutrient availability. Indeed, the observed improved feed efficiency associated with CG may be through potentially increasing the surface area of rumen papaillae. This hypothesis is further fortified following the results of Sun et al. [ 23 ], who showed that rumen papillae height, width and surface area were all lower in goats that had undergone a 6 week period of dietary restriction. However, following a period of CG, full recovery in the morphology of epithelium tissue was observed [ 23 ]. Greater expression of genes involved in cellular adhesion and interaction as well as cellular transport proteins in the current study suggest that the structural state of the rumen papillae may play an important role in governing the expression of entire body CG. Indeed, an increase in rumen papillae structure and consequently surface area during re-alimentation may potentially contribute to an improvement in nutrient absorption during periods of accelerated growth, which is consistent with the increase in appetite and feed intake capacity of animals undergoing CG [ 7 , 9 , 11 , 14 ]. Moreover, Kong et al. [ 56 ] reported up-regulation of genes involved in intracellular adhesion and actin cytoskeleton in the rumen epithelium of feed efficient cattle suggesting that the rumen epithelium may contribute to the enhanced feed efficiency evident during CG. Additionally, restoration of ruminal epithelium may be a necessary requirement in response to re-alimentation in order to cope with the increase in associated metabolic activity concomitant with increased dietary intake which was evident in the animals used in the current study where consumption of feed was greater on a proportional body weight basis in RES compared with ADLIB animals [ 11 ]. However, although DEG profiles suggest alterations to rumen papillae surface area in response to both diet restriction and CG, physical measurements, including papillae height, width, crypt depth are necessary to prove this hypothesis in cattle. Immune function Our global gene expression data suggest that the animal's immune system was also affected by both dietary restriction and subsequent re-alimentation induced CG in rumen epithelial. This was manifested through differential expression of immune related genes namely up-regulation of IL17A , and LBP in animals undergoing dietary restriction. IL17A codes for interleukin 17a, a proinflammatory cytokine [ 63 ], whereas LBP is involved in host defence against gram negative bacteria and plays a role in innate immune response [ 64 ]. Similarly, following a 10-week period of feed restriction, changes in genes regulating immune function and inflammation was apparent in hepatic tissue in the data of Connor et al. [ 10 ]. Moreover, Dhahbi et al. [ 65 ] reported functional groups of genes to be affected by calorie restriction in mice including those involved in the immune response. Periods of moderate dietary restriction have previously been shown to affect the immune system manifested as an up-regulation of immune genes and an overall greater capacity for immune response following a period of dietary restriction [ 66 – 70 ]. Up-regulation of genes governing the immune response during nutrient restriction may represent a potential protective mechanism against pathological disease. Indeed, a study on rodents showed that the immunological status of rodents offered a restricted feed allowance was superior to that of their non-restricted counterparts [ 71 ]. A similar outcome was also apparent in the jejunal epithelial cells of cattle following a period of dietary restriction, whereby CTSW , a gene which functions in T-cell cytolytic activity was also up-regulated in cattle that had undergone a period of dietary restriction compared to their ad libitum counterparts [ 24 ]. Overall, these results suggest that dietary restriction in cattle can elicit a superior immunological status as previously described in other species which may protect against any potential pathological threats to the animal. Alternatively, it has been suggested that the immune response could be involved in nutrient partitioning away from non-essential activities including growth and instead towards activating tissue mobilisation and catabolism [ 72 ]. Nutrient partitioning during diet restriction has been widely reported in cattle [ 11 , 12 , 44 , 73 , 74 ]. When coupled with data from the present study these results indicate that the immune system may be contributing to this observed effect. Indeed in the context of the current study this may be reflective of a change in rumen size and weight in response to a period of dietary restriction. Immune related genes were subsequently down-regulated in ruminal epithelial during re-alimentation compared with previous dietary restriction. Immune genes down-regulated reflected those involved in inflammation ( BDKRB1 , CHI3L1 , HPGD , LTA4H ); the complement system ( C5AR2 , CD59 ); cytokines ( CCL19 ) and chemokines ( CCL20 , CXCL12 , CXCL17 , CXCL2 , CXCR4 ) as well as others ( LYZ , immunoagents; SGSH and CYBA , lysosomal degradation). In the data of Chen et al. [ 75 ], CD59 was also found to be down-regulated in hepatic tissue of feed efficient cattle. Similarly, in the current ruminal papillae study, CD59 was down-regulated in animals undergoing CG consistent with increased feed efficiency [ 4 , 11 ]. Moreover, LTA4H gene was also down-regulated in the skeletal muscle of our cattle when undergoing CG [ 25 ]. Studies in beef cattle divergently selected for feed efficiency have indicated that a large proportion of the variation in efficiency among animals may be attributed to stress or immune related biological pathways [ 76 ]. Moreover, Alexandre et al. [ 77 ] described down-regulation of genes involved in the immune response in feed efficient cattle, which is consistent with the results of the current study, as during CG the animals in the current study displayed a better feed efficiency potential [ 11 ]. Kern et al. [ 78 ] recently suggested that a reduction in an animal's immune response, as described during CG in the current study, may allow for more energy to be directed toward cellular proliferation and growth. As this effect was observed in the current study, it is possible that down-regulation of immune-related genes during re-alimentation may allow for the rapid CG typically observed for the rumen [ 11 ]. Alternatively, the same authors suggested that a reduction in the immune response could benefit both intake and gain through a reduction in papillae swelling, which may allow for improved nutrient absorption [ 78 ]. Gene transcription and protein folding A reduction in feed intake is typically paralleled by a reduction in growth and overall cellular functions. Indeed, this was apparent in the papillae of RES animals following a period of dietary restriction, where genes involved in gene transcription and protein folding tended to be down-regulated compared with ADLIB animals. Specifically these DEGs were manifested as down-regulation of genes coding for proteins involved in chaperone functionality including CRYAB , HSPB8 and HSPH1 . The CRYAB protein displays chaperone-like activity and functions in preventing aggregation of various proteins under a wide range of conditions [ 35 ]. We also found this gene to be down-regulated in the skeletal muscle tissue of these same cattle during dietary restriction [ 25 ]. Both HSPB8 and HSPH1 code for heat shock proteins which also function in the prevention of aggregation of denatured proteins in cells [ 36 ]. HSPH1 was also found to be down-regulated in skeletal muscle of cattle following a period of dietary restriction [ 25 ]. Furthermore, up-regulation of both SATB1 and ZC3H12A was also apparent in the rumen papillae of RES animals at the same time-point. Both of these genes code for proteins involved in repressing transcription, thereby causing down-regulation of gene expression processes [ 37 , 38 ]. SATB1 encodes a matrix protein, which functions to recruit chromatic remodelling factors in order to regulate chromatin structure and gene expression and ultimately functions in transcriptional repression and gene silencing [ 39 ]. Consistent with this ZC3H12A displays RNase activity and functions in selectively degrading specific target mRNA species [ 40 ]. We also found ZC3H12A to be up-regulated in hepatic tissue of these cattle following a period of dietary restriction [ 26 ]. Down-regulation of genes involved in these processes following a period of dietary restriction may be reflective of a reduced requirement for nutrient processing and metabolism, coinciding with lower animal ADG and weight of the rumen complex, at the end of Period 1 [ 11 ]. Moreover, following a 125 day period of dietary restriction, when compared with a reference slaughter group at the start of dietary restriction, the proportional weight of the rumen in feed restricted animals was found to be lower [ 11 ], further evidencing a reduced metabolic requirement of this organ in response to dietary restriction. Lower expression of genes involved in cellular metabolism, following dietary restriction was also apparent in liver tissue of the same animals used in the current study [ 26 ]. While the aforementioned biological processes were down-regulated during dietary restriction, upon re-alimentation up-regulation of these functions was apparent which coincided with greater rumen and overall body growth rates [ 11 ]. Up-regulation of these processes during re-alimentation also coincided with a greater capacity for growth and requirement for cellular metabolism in the rumen tissue as well as in other organs within the body. For example, increased expression of genes associated with metabolism during CG was previously described in the hepatic tissue of the animals used in the current study [ 26 ]. Indeed greater transcription of genes coding for proteins involved in gene expression was evident in papillae of animals displaying CG. Greater expression of these genes may be necessary in order to allow for increased production of proteins to accommodate the increased nutrient availability and metabolic demands of digestion, absorption and ultimately tissue growth. Genes involved in protein folding included those coding for chaperone proteins: AHSA1 ; DNAJB4 ; HSPA8 ; HSPB8 ; HSPD1 ; MDN1 ; as well as those requiring input of ATP: CCT2 ; HSP90AA1 ; HSP90AB1 ; HSPE1 . These genes have also been shown to display greater expression during CG in other tissues and organs, including skeletal muscle ( AHSA1 , DNAJB4 , HSPA8 , HSPB8 and HSPD1 ; [ 25 ]; and liver ( HSPA8 , HSPB8 and HSPD1 ; [ 10 ]). Moreover, HSPA8 , HSPB8 and HSPD1 were also up-regulated in the liver of feed efficient cattle [ 41 ]. These genes may be important in relation to improved feed efficiency consistent with that observed for cattle undergoing CG including those in the current study [ 4 , 11 ]. Greater expression of CCT2 and HSPE1 was also reported in skeletal muscle of our animals while undergoing CG [ 25 ]. Furthermore, up-regulation of HSP90AA1 and HSP90AB1 was also apparent in both hepatic and skeletal muscle tissues of cattle undergoing CG [ 10 , 25 ]. Greater expression of PPID , a protein that functions in accelerating protein folding [ 42 ] as well as STIP1 which regulates both the conformation and ATPase cycles of HSP70 and HSP90 molecular chaperones [ 43 ] was also apparent in rumen papillae of cattle undergoing CG in the current study. Both of these genes were also up-regulated in skeletal muscle tissue of cattle undergoing CG [ 25 ]. Overall these results suggest that increased cellular protein folding activity is required within the rumen papillae as part of the adaption to an increased dietary intake and is consistent with the heightened metabolic state typical of animals undergoing re-alimentation induced CG [ 11 , 12 , 44 , 45 ]. Indeed, this may be a necessary response in order to cope with the typically elevated rate of metabolism associated with greater feed consumption [ 46 ], which appears to be a primary driver of whole animal CG [ 8 , 11 , 14 ]. However, further studies are required to assess the metabolic state of the rumen and indeed other metabolic organs in response to both dietary restriction and CG. In addition to an increase in the expression of genes coding for chaperone and protein folding cellular machinery, up-regulation of genes involved in transcriptional activity was also observed in rumen papillae of animals undergoing CG, again this coincided with greater reticulo-rumen and whole body growth [ 11 ]. Overall, genes coding for proteins involved in transcription ( EMG1 , FOXN1 , FOXP4 , INTS3 ), splicing ( KHDRBS3 , PRCC ) and translation ( EIF4G2 , EIF4G3 , ELL2 ) were up-regulated in rumen papillae of cattle undergoing CG. Transcriptional genes differentially expressed included EMG1 , which encodes a protein involved in ribosome biogenesis [ 47 ], two FoxO proteins which are involved in the regulation of gene transcription [ 48 ] and a subunit of the integrator complex of RNA polymerase II ( INTS3 ; [ 49 ]). Both EIF4G2 and EIF4G3 code for proteins involved in the eukaryotic translation initiation factor and function in the recognition of the mRNA cap, and recruitment of mRNA to the ribosome [ 50 ]. Up-regulation of EIF4G2 was also reported in the data of Connor et al. [ 10 ] in hepatic tissue of cattle undergoing CG. The elongation factor component ELL2 was also up-regulated in skeletal muscle tissue of the same cattle used in the current study [ 25 ]. This gene codes for a complex which is required to increase the catalytic rate of RNA polymerase II transcription [ 51 ]. Genes involved in splicing, and the editing of nascent pre-mRNA [ 52 ] were also detected as differentially expressed in rumen papillae of cattle undergoing CG in the current study. These included KHDRBS3 which functions in the regulation of alternative splicing and influences mRNA splice site selection [ 53 ] and PRCC which functions in pre-mRNA splicing [ 54 ]. Moreover, genes coding for histone proteins ( HIST1H2AC , HIST1H2BD , HIST1H2BN , HISTH2BO , HIST2H4A , KAT2A ) were also up-regulated during the same time. Histones are proteins that package and order DNA into structural nucleosomes, playing a role in gene regulation [ 55 ]. Additionally, KAT2A a histone acetlytransferase that functions primarily as a transcriptional activator was also up-regulated in papillae of animals undergoing CG. Genes coding for histone proteins were also detected as up-regulated in skeletal muscle tissue of cattle expressing CG, these included HIST1H2AC , HIST1H2BD and KAT2A [ 25 ]. Collectively, these results suggest an increase in gene expression and associated translational and protein folding activity in rumen papillae epithelia during CG and associated feed efficiency in cattle. A similar effect was also reported in rumen epithelium of feed efficient cattle (low-residual feed intake) [ 56 ]. This is also apparent in the network presented in Fig 3 , where genes associated with metabolism, biochemistry and cellular assembly and organisation were up-regulated. Up-regulation of these cellular processes during rumen papillae CG may be a consequence of a greater nutrient intake during re-alimentation and be necessary for the replenishment of the associated metabolic machinery required for increased digestion and absorption, which ultimately may be contributing to compensatory tissue growth and development, and as stated earlier was also apparent during CG of hepatic tissue of cattle undergoing CG [ 10 , 26 ]. Cellular interactions and organisation Reduced nutrient intake may be consistent with a down-regulation of cellular processes associated with cellular function and organisation [ 57 ]. This has previously been reported in skeletal muscle of cattle after diet restriction [ 25 ]. Following a period of dietary restriction, genes coding for proteins involved in structural components of ruminal epithelial cells were observed to be down-regulated in RES compared to ADLIB animals. Down-regulation of these genes may be due to a lack of requirement for a large ruminal epithelial surface area as a consequence of a reduction in intake and associated digestive processes in the rumen. Moreover, a reduction in epithelial surface area may allow for a reduction in cellular maintenance requirements in an energetically demanding organ such as the rumen. At the end of Period 1, CDH2 and DSG1 were both down-regulated in RES animals compared to ADLIB animals. CDH2 codes for a cadherin, which are a family of transmembrane proteins involved in cellular adhesion [ 58 ]. The encoded protein CDH2, is a calcium dependent cell-cell adhesion glycoprotein [ 59 ]. The gene DSG1 codes for a desmosome protein, which form junctions between certain cell types including epithelial cells [ 58 ]. DSG1 is a calcium-binding trans-membrane glycoprotein component of desmosomes in vertebrate epithelial cells [ 60 ]. It is involved in maintaining the structural integrity of epithelial cells including rumen epithelium and intermediate filaments mediating cell-cell adhesion [ 61 ]. Structural alterations to rumen papillae in response to differences in dietary intake have previously been reported. For example, Steele et al. [ 62 ] observed structural adaptations in rumen epithelium when cows were fed a diet consisting primarily of grain. Moreover, in that study lower expression of DSG1 was reported in response to a high concentrate diet, with expression subsequently greater upon transition to a high forage diet [ 62 ]. Additionally, Sun et al. [ 23 ] observed reduced rumen epithelial height, width and surface area in goats following a 48 day period of dietary restriction. It is logical to expect that the cumulative surface area of papillae and thus weight of the organ itself reflects the prevailing dietary management of an animal. As a consequence of reduced dietary intake, there may be a decreased necessity for ruminal papillae surface area, which may contribute to the reduction in rumen size, as observed in the current study [ 11 ]. In turn, the reduction in rumen size may allow for rumen metabolic rate to be curtailed which in turn could contribute to reduced animal maintenance requirements during dietary restriction. Conversely though, during re-alimentation induced CG, the corollary was observed in the current study, whereby expression of genes coding for proteins involved in cellular interactions and organisation was greater in papillae of RES animals at the end of Period 2 than at the end of Period 1. Genes involved in cellular adhesion ( ANTXR1 , CEP97 , FAT4 , PCDH12 , PCDH7 ), cellular interactions ( IGCA8 , NRG1 , RELN , SMAGP , THBS4 ) and transport ( SLC1A5 , SLC22A17 , SLC25A15 , SLC25A26 , SLC30A6 , SLC4A7 , SLC6A9 , SLC9A1 , CACNA1G , KCNC4 ) were all up-regulated during the CG of ruminal papillae. A similar effect has also been reported in skeletal muscle for the same cattle population used here [ 25 ]. Of note, up-regulation of the following genes PCDH12 and THBS4 as well as two transporter genes, SLC22A17 and SLC25A15 was consistent between the current study for rumen epithelial tissue and our previous study using muscle tissue [ 25 ]. Up-regulation of these processes during CG in ruminal epithelial may have reflected a necessary adaptive requirement for cells to cope with the increase in cellular metabolic activity as a consequence of increased nutrient availability. Indeed, the observed improved feed efficiency associated with CG may be through potentially increasing the surface area of rumen papaillae. This hypothesis is further fortified following the results of Sun et al. [ 23 ], who showed that rumen papillae height, width and surface area were all lower in goats that had undergone a 6 week period of dietary restriction. However, following a period of CG, full recovery in the morphology of epithelium tissue was observed [ 23 ]. Greater expression of genes involved in cellular adhesion and interaction as well as cellular transport proteins in the current study suggest that the structural state of the rumen papillae may play an important role in governing the expression of entire body CG. Indeed, an increase in rumen papillae structure and consequently surface area during re-alimentation may potentially contribute to an improvement in nutrient absorption during periods of accelerated growth, which is consistent with the increase in appetite and feed intake capacity of animals undergoing CG [ 7 , 9 , 11 , 14 ]. Moreover, Kong et al. [ 56 ] reported up-regulation of genes involved in intracellular adhesion and actin cytoskeleton in the rumen epithelium of feed efficient cattle suggesting that the rumen epithelium may contribute to the enhanced feed efficiency evident during CG. Additionally, restoration of ruminal epithelium may be a necessary requirement in response to re-alimentation in order to cope with the increase in associated metabolic activity concomitant with increased dietary intake which was evident in the animals used in the current study where consumption of feed was greater on a proportional body weight basis in RES compared with ADLIB animals [ 11 ]. However, although DEG profiles suggest alterations to rumen papillae surface area in response to both diet restriction and CG, physical measurements, including papillae height, width, crypt depth are necessary to prove this hypothesis in cattle. Immune function Our global gene expression data suggest that the animal's immune system was also affected by both dietary restriction and subsequent re-alimentation induced CG in rumen epithelial. This was manifested through differential expression of immune related genes namely up-regulation of IL17A , and LBP in animals undergoing dietary restriction. IL17A codes for interleukin 17a, a proinflammatory cytokine [ 63 ], whereas LBP is involved in host defence against gram negative bacteria and plays a role in innate immune response [ 64 ]. Similarly, following a 10-week period of feed restriction, changes in genes regulating immune function and inflammation was apparent in hepatic tissue in the data of Connor et al. [ 10 ]. Moreover, Dhahbi et al. [ 65 ] reported functional groups of genes to be affected by calorie restriction in mice including those involved in the immune response. Periods of moderate dietary restriction have previously been shown to affect the immune system manifested as an up-regulation of immune genes and an overall greater capacity for immune response following a period of dietary restriction [ 66 – 70 ]. Up-regulation of genes governing the immune response during nutrient restriction may represent a potential protective mechanism against pathological disease. Indeed, a study on rodents showed that the immunological status of rodents offered a restricted feed allowance was superior to that of their non-restricted counterparts [ 71 ]. A similar outcome was also apparent in the jejunal epithelial cells of cattle following a period of dietary restriction, whereby CTSW , a gene which functions in T-cell cytolytic activity was also up-regulated in cattle that had undergone a period of dietary restriction compared to their ad libitum counterparts [ 24 ]. Overall, these results suggest that dietary restriction in cattle can elicit a superior immunological status as previously described in other species which may protect against any potential pathological threats to the animal. Alternatively, it has been suggested that the immune response could be involved in nutrient partitioning away from non-essential activities including growth and instead towards activating tissue mobilisation and catabolism [ 72 ]. Nutrient partitioning during diet restriction has been widely reported in cattle [ 11 , 12 , 44 , 73 , 74 ]. When coupled with data from the present study these results indicate that the immune system may be contributing to this observed effect. Indeed in the context of the current study this may be reflective of a change in rumen size and weight in response to a period of dietary restriction. Immune related genes were subsequently down-regulated in ruminal epithelial during re-alimentation compared with previous dietary restriction. Immune genes down-regulated reflected those involved in inflammation ( BDKRB1 , CHI3L1 , HPGD , LTA4H ); the complement system ( C5AR2 , CD59 ); cytokines ( CCL19 ) and chemokines ( CCL20 , CXCL12 , CXCL17 , CXCL2 , CXCR4 ) as well as others ( LYZ , immunoagents; SGSH and CYBA , lysosomal degradation). In the data of Chen et al. [ 75 ], CD59 was also found to be down-regulated in hepatic tissue of feed efficient cattle. Similarly, in the current ruminal papillae study, CD59 was down-regulated in animals undergoing CG consistent with increased feed efficiency [ 4 , 11 ]. Moreover, LTA4H gene was also down-regulated in the skeletal muscle of our cattle when undergoing CG [ 25 ]. Studies in beef cattle divergently selected for feed efficiency have indicated that a large proportion of the variation in efficiency among animals may be attributed to stress or immune related biological pathways [ 76 ]. Moreover, Alexandre et al. [ 77 ] described down-regulation of genes involved in the immune response in feed efficient cattle, which is consistent with the results of the current study, as during CG the animals in the current study displayed a better feed efficiency potential [ 11 ]. Kern et al. [ 78 ] recently suggested that a reduction in an animal's immune response, as described during CG in the current study, may allow for more energy to be directed toward cellular proliferation and growth. As this effect was observed in the current study, it is possible that down-regulation of immune-related genes during re-alimentation may allow for the rapid CG typically observed for the rumen [ 11 ]. Alternatively, the same authors suggested that a reduction in the immune response could benefit both intake and gain through a reduction in papillae swelling, which may allow for improved nutrient absorption [ 78 ]. Conclusions Following a period of dietary restriction, we described evidence for reduced gene expression and cellular interactions in rumen papillae tissue of Holstein Friesian bulls. This was in conjunction with an apparent enhanced immune response potential. During subsequent re-alimentation induced CG, our data suggest that greater nutrient intake is consistent with an up-regulation in transcriptional activity of ruminal epithelial tissue, which may in turn lead to greater nutrient uptake through an increase in papillae surface area and ultimately contribute to increased feed efficiency typical of CG, thus supporting the accelerated growth phenomenon of both the rumen as well as the animal. In contrast to that observed for diet restricted cattle a period of improved feed efficiency was consistent with a reduction in the abundance of transcripts for genes involved in immune response, potentially allowing more energy to be channelled towards growth within the rumen papillae. Our results also suggest that the structural state of the gastrointestinal tract may play an important role in governing feed efficiency, with an increase in rumen papillae surface area during re-alimentation potentially contributing to improvements in nutrient absorption during periods of accelerated growth. The new knowledge generated in this study offers further insights into some of the many molecular processes underlying nutrient restricted and CG states in cattle. However functional studies are now warranted to validate the hypotheses put forward in the current study. Furthermore, our DEG patterns provide baseline data which may be further interrogated and used to identify animals with superior genetic potential for CG and associated feed efficiency. Supporting information S1 Table Genes differentially expressed in rumen epithelium of Holstein Friesian bulls (n = 10) following a 125-day period of restricted feeding at the end of Period 1 relative to ad libitum -fed controls (n = 10). (DOCX) Click here for additional data file. S2 Table Genes differentially expressed in rumen epithelium of Holstein Friesian bulls (n = 10) following a 55-day period of re-alimentation and compensatory growth in Period 2 relative to animals fed a restricted diet for 125 days at the end of Period 1(n = 10). (DOCX) Click here for additional data file. S3 Table Networks generated from gene expression data of restricted versus ad libitum fed bulls by IPA. (DOCX) Click here for additional data file. S4 Table Networks generated from gene expression data of compensating versus restricted fed bulls by IPA. (DOCX) Click here for additional data file.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10882300/

Post-traumatic growth among emergency nurses after COVID-19 in Shanghai, China: a qualitative study

Objectives This study aims to investigate the post-traumatic growth of emergency nurses (ENs) in Shanghai, China, in 2022 following the COVID-19 pandemic. Design A phenomenological qualitative research approach was employed, with 17 ENs being interviewed between July and August 2022. Data collection was conducted through semistructured, in-depth interviews, and data analysis was carried out using the Colaizzi's seven-step analysis method. Setting A third-level hospital in Shanghai. Participants A total of 17 ENs were interviewed through face-to-face, semistructured, in-depth interviews. Results Three main themes and eight subthemes were extracted from the data: (a) stress, (b) restructuring and (c) growth. Conclusion Significant stress was imposed on ENs by the Shanghai COVID-19 pandemic, but cognitive restructuring was successfully undergone by them, leading to the experience of growth. It is recommended that post-traumatic growth levels be enhanced through professional psychological counselling and tailored support measures for different stages. Objectives This study aims to investigate the post-traumatic growth of emergency nurses (ENs) in Shanghai, China, in 2022 following the COVID-19 pandemic. Design A phenomenological qualitative research approach was employed, with 17 ENs being interviewed between July and August 2022. Data collection was conducted through semistructured, in-depth interviews, and data analysis was carried out using the Colaizzi's seven-step analysis method. Setting A third-level hospital in Shanghai. Participants A total of 17 ENs were interviewed through face-to-face, semistructured, in-depth interviews. Results Three main themes and eight subthemes were extracted from the data: (a) stress, (b) restructuring and (c) growth. Conclusion Significant stress was imposed on ENs by the Shanghai COVID-19 pandemic, but cognitive restructuring was successfully undergone by them, leading to the experience of growth. It is recommended that post-traumatic growth levels be enhanced through professional psychological counselling and tailored support measures for different stages. Introduction Traumatic events affect a significant portion of the population, with approximately 70% of individuals experiencing at least one such event in their lifetime. 1 These events can lead to trauma-related psychological responses not only among survivors but also among those who respond to emergencies, such as police officers, firefighters and nurses. Nurses, in particular, are at heightened risk due to their roles in healthcare, often operating in high-stress, high-risk medical environments where they frequently encounter traumatic situations like illnesses and deaths. An often overlooked group within the nursing profession in China is emergency nurses (ENs), who routinely face traumatic incidents during their professional duties. Emergency departments (EDs) are known for being one of the busiest areas in hospitals across China, where ENs work tirelessly in demanding conditions. Their responsibilities include managing critical medical cases, addressing frequent medical disputes and complaints, and even dealing with incidents of violence. They also contend with the emotional toll of witnessing deaths, navigate overcrowded spaces and cope with disrupted sleep patterns due to shift work. In late February 2022, Shanghai, a city with a population of 25 million, faced a rapid surge in COVID-19 cases driven by a variant of the Omicron strain. This pandemic storm placed immense pressure on the city's healthcare system, significantly impacting ENs. They were confronted with increased workloads, a severe shortage of medical resources, and the added responsibility of screening and caring for infected patients. 2 3 This burden was exacerbated by the fact that more hospitals were converted into designated treatment centres, and a quarter of ENs were reassigned to designated hospitals, mobile cabin hospitals and the Shanghai Public Health Clinical Center. Furthermore, the majority of patients were treated within EDs rather than in inpatient units, straining medical resources. Delays in transferring patients from EDs, partly due to the overwhelming workload at the Shanghai Medical Emergency Center ('Shanghai 120'), further intensified the strain on medical resources, ultimately heightening the risk of infection for ENs. 3 Amid the healthcare turmoil and strained medical resources mentioned above, ENs found themselves facing a heightened risk of encountering traumatic events. These traumatic events can lead to various psychological reactions, which can be either negative, positive or a combination of both. 4 For instance, during past traumatic events like the Wenchuan earthquake, Hurricane Floyd and anthrax incidents, nurses played a crucial role in caring for patients while simultaneously dealing with injuries or even the loss of their own family members. 5–7 Among these reactions, one positive response stands out—post-traumatic growth (PTG). 8 PTG involves the positive transformation of an individual's cognitive outlook following a traumatic experience. This phenomenon is often referred to as 'finding benefits', 'stress-related growth' or 'adversary growth'. 2 The concept of traumatic events leading to positive change has spanned cultures for centuries, from the Egyptian myth of the phoenix's rebirth to Friedrich Nietzsche's famous quote, 'what doesn't kill you makes you stronger', and Chinese proverbs like 'good fortune follows upon disaster' and 'a fall into a pit, a gain in your wit'. Numerous studies have highlighted various factors influencing the development of PTG, including age, gender, education level, self-assessment, post-traumatic stress symptoms, self-expression, social support, thoughtfulness, wisdom and self-compassion. 9 In recent years, there has been a growing interest in PTG among medical personnel. Early investigations primarily focused on PTG among nurses following incidents of hospital violence, 10 patient deaths 11 and other traumatic events. Additionally, studies explored PTG after natural disasters like earthquakes and tsunamis. 12 With the onset of the COVID-19 pandemic, an increasing number of studies have delved into the factors influencing nurses' PTG. Some of these studies suggest that the trauma experienced by nurses in patient care enhances their competence and compassion. 13 The physical and mental challenges posed by the pandemic can also provide valuable lessons for nurses, helping them develop coping strategies for difficult situations and improve resource management skills. 14 However, limited research has focused on the PTG experienced by ENs in the aftermath of the COVID-19 epidemic. Given the dynamic and diverse nature of PTG across different populations, 15–18 this study aims to qualitatively explore PTG among ENs and analyse the factors shaping it. The findings will establish a foundation for psychological interventions and support measures in response to future traumatic events experienced by ENs. Aim and research question The objective of this study is to investigate the PTG experienced by ENs following the COVID-19 pandemic in Shanghai, China, in 2022. The research inquiries for this study are as follows: How do ENs undergo personal growth in the aftermath of public health emergencies? and What factors contribute to their PTG? Methods Study design This study employs a phenomenological approach, using semistructured, in-depth interviews to delve into intricate phenomena, such as human emotions and experiences. The aim of phenomenological research is to enhance comprehension of human behaviour, elucidate the essence of actions, and provide a thorough and comprehensive portrayal and interpretation of human life experiences. 19 The adoption of the phenomenological research method enhances the understanding and acknowledgement of the PTG experiences of ENs. We adhered to the Consolidated Criteria for Reporting Qualitative Research guidelines to present our methods and findings in a well-organised manner. 20 Participants and settings To assemble participants for this study, we employed purposive sampling from a third-level hospital located in Shanghai. Inclusion criteria were as follows: (a) enrolment in an ED for a minimum of 1 year; (b) active involvement in frontline emergency medicine for at least 2 months during the Shanghai COVID-19 pandemic; (c) participants capable of articulating their emotional experiences and (d) voluntary consent to participate in the study. The characteristics of the ENs who partook in the survey are summarised in table 1 . In this study, we reached a saturation point where no new information emerged; therefore, 17 subjects were included. No participants withdrew from the interview process. Table 1 Characteristics of participants (n=17) Characteristics N (%) or mean (SD) Sex Male 6 (35.29) Female 11 (64.71) Age 29.88 (2.91) Length of employment 7.71 (2.80) Education level Diploma 2 (11.76) Baccalaureate degree 14 (82.36) Master's degree 1 (5.88) Marital status Married 6 (35.29) Single 11 (64.71) Data collection The interview questions outlined in table 2 were adapted from previous research. 2 3 A preliminary interview was conducted with two ENs to evaluate the suitability of the interview questions; however, the results from these interviews were not included in the data analysis. Between July and August 2022, semistructured interviews were carried out and recorded in a tranquil workplace setting, with each interview lasting approximately 30–40 min. The first author, who possessed substantial experience in conducting interviews and a profound grasp of qualitative research methodologies, conducted the interviews without any prior relationship with the research participants. After conducting 17 interviews, data saturation and validation were achieved. As a token of appreciation, each participant received a small gift (a 200 mL bottle of shampoo), and no participants withdrew from the survey. Table 2 Interview questions No Questions 1 What were your experiences and changes while you fought the pandemic? 2 How did you deal with the difficulties and pressures of this time? 3 What positive changes have you experienced since then? 4 What motivated these positive changes? 5 What are your plans for the future? Data analysis Verbatim transcriptions of the interview data were promptly conducted within 24 hours of the interviews. Two seasoned researchers independently extracted themes and subthemes from the data using Colaizzi's seven-step phenomenological method, 21 the detailed procedures are presented in table 3 . This process involved a meticulous and iterative examination of the collected data to attain a comprehensive understanding of all information conveyed by the participants. Key statements were isolated and recorded in written form, recurring concepts were coded and synopses of similar meanings were crafted to establish thematic prototypes. Throughout this process, every effort was made to prevent subjective judgement. Each identified theme was subsequently meticulously elucidated, incorporating participant quotations to facilitate detailed analysis. Following this, themes and descriptions underwent repeated comparisons to uncover shared perspectives, construct overarching themes and ensure result accuracy, which was further validated through participant validation. In cases where differences of opinion arose, resolution was achieved through discussion with a third researcher. Table 3 Colaizzi's seven-step method 21 No Method 1 Familiarisation 2 Identifying meaningful statements 3 Constructing meanings 4 Topic clustering 5 Describing in detail 6 Producing the basic structure 7 Verifying basic structure Study rigour To enhance the dependability of the study, a qualitative research expert who was not part of the research team re-evaluated the data. The decision-making process was rigorously examined by two seasoned researchers, thereby ensuring the study's confirmability. Techniques employed to ensure credibility encompassed researcher triangulation, data collection triangulation and participant validation. Lastly, transferability was upheld by providing a comprehensive description of the study's background, sampling methods, data collection procedures and analytical processes. 22 Patient and public involvement No patients were involved in this study. Study design This study employs a phenomenological approach, using semistructured, in-depth interviews to delve into intricate phenomena, such as human emotions and experiences. The aim of phenomenological research is to enhance comprehension of human behaviour, elucidate the essence of actions, and provide a thorough and comprehensive portrayal and interpretation of human life experiences. 19 The adoption of the phenomenological research method enhances the understanding and acknowledgement of the PTG experiences of ENs. We adhered to the Consolidated Criteria for Reporting Qualitative Research guidelines to present our methods and findings in a well-organised manner. 20 Participants and settings To assemble participants for this study, we employed purposive sampling from a third-level hospital located in Shanghai. Inclusion criteria were as follows: (a) enrolment in an ED for a minimum of 1 year; (b) active involvement in frontline emergency medicine for at least 2 months during the Shanghai COVID-19 pandemic; (c) participants capable of articulating their emotional experiences and (d) voluntary consent to participate in the study. The characteristics of the ENs who partook in the survey are summarised in table 1 . In this study, we reached a saturation point where no new information emerged; therefore, 17 subjects were included. No participants withdrew from the interview process. Table 1 Characteristics of participants (n=17) Characteristics N (%) or mean (SD) Sex Male 6 (35.29) Female 11 (64.71) Age 29.88 (2.91) Length of employment 7.71 (2.80) Education level Diploma 2 (11.76) Baccalaureate degree 14 (82.36) Master's degree 1 (5.88) Marital status Married 6 (35.29) Single 11 (64.71) Data collection The interview questions outlined in table 2 were adapted from previous research. 2 3 A preliminary interview was conducted with two ENs to evaluate the suitability of the interview questions; however, the results from these interviews were not included in the data analysis. Between July and August 2022, semistructured interviews were carried out and recorded in a tranquil workplace setting, with each interview lasting approximately 30–40 min. The first author, who possessed substantial experience in conducting interviews and a profound grasp of qualitative research methodologies, conducted the interviews without any prior relationship with the research participants. After conducting 17 interviews, data saturation and validation were achieved. As a token of appreciation, each participant received a small gift (a 200 mL bottle of shampoo), and no participants withdrew from the survey. Table 2 Interview questions No Questions 1 What were your experiences and changes while you fought the pandemic? 2 How did you deal with the difficulties and pressures of this time? 3 What positive changes have you experienced since then? 4 What motivated these positive changes? 5 What are your plans for the future? Data analysis Verbatim transcriptions of the interview data were promptly conducted within 24 hours of the interviews. Two seasoned researchers independently extracted themes and subthemes from the data using Colaizzi's seven-step phenomenological method, 21 the detailed procedures are presented in table 3 . This process involved a meticulous and iterative examination of the collected data to attain a comprehensive understanding of all information conveyed by the participants. Key statements were isolated and recorded in written form, recurring concepts were coded and synopses of similar meanings were crafted to establish thematic prototypes. Throughout this process, every effort was made to prevent subjective judgement. Each identified theme was subsequently meticulously elucidated, incorporating participant quotations to facilitate detailed analysis. Following this, themes and descriptions underwent repeated comparisons to uncover shared perspectives, construct overarching themes and ensure result accuracy, which was further validated through participant validation. In cases where differences of opinion arose, resolution was achieved through discussion with a third researcher. Table 3 Colaizzi's seven-step method 21 No Method 1 Familiarisation 2 Identifying meaningful statements 3 Constructing meanings 4 Topic clustering 5 Describing in detail 6 Producing the basic structure 7 Verifying basic structure Study rigour To enhance the dependability of the study, a qualitative research expert who was not part of the research team re-evaluated the data. The decision-making process was rigorously examined by two seasoned researchers, thereby ensuring the study's confirmability. Techniques employed to ensure credibility encompassed researcher triangulation, data collection triangulation and participant validation. Lastly, transferability was upheld by providing a comprehensive description of the study's background, sampling methods, data collection procedures and analytical processes. 22 Patient and public involvement No patients were involved in this study. Results From the data analysis, three main themes and eight subthemes emerged: (a) stress, (b) reconstruction and (c) growth. These themes and subthemes are presented in figure 1 . Figure 1 Themes and subthemes structure. Stress Suffering All interviewees reported enduring physical and psychological hardships due to the use of personal protective equipment (PPE), an unexpected surge in workload and the demanding rescue procedures. One participant expressed, "Wearing the PPE made resuscitation a challenge. Securing a catheter for a ventilated patient was much harder due to the obstructed view caused by the face shield. To save time, I went without food or water for 8 straight hours. The clothing under the PPE was completely soaked after just one shift" (N9). Another nurse shared, "Despite numerous patients being stranded in the ED, we kept receiving a continuous flow of new cases. In addition to emergency cases, we had to identify and treat COVID-19 patients. It was common for multiple cases of respiratory failure requiring intubation to occur simultaneously, keeping me extremely busy. The sound of an ambulance siren felt like being caught in an overwhelming whirlwind" (N12). One nurse remarked, "Some specialists had to come from their respective wards and spend time donning PPE. Patients waited for extended periods and argued with me, which was quite frustrating" (N4). In response to the Shanghai COVID-19 pandemic, ENs faced prolonged separation from their families. The conflict between their familial responsibilities and work commitments proved to be a source of trauma and distress. One EN shared, "My wife was on maternity leave, and my children and parents contracted the virus. I couldn't be there to take care of them due to work, and it left me feeling profoundly saddened" (N15). Fear ENs grappled with various fears during the Shanghai COVID-19 pandemic, primarily stemming from the uncertainties surrounding the virus, the fear of contracting it themselves and the potential for transmitting the infection to other patients due to errors in their work. One EN stated, "I constantly worried about being infected with the virus, and I checked my nucleic acid test results every afternoon. If the result didn't show up, I would become anxious" (N3). Another shared, "I worked in the pre-screening triage office, and I was constantly anxious about the possibility of inadvertently allowing infected patients into an area where uninfected patients were being treated. The pressure was immense" (N6). While hospital management made efforts to deploy additional nursing resources and brought in transdisciplinary nurses (TNs) to support the EDs, the ENs still experienced anxiety. This stemmed from concerns that the TNs, who had not previously worked in EDs, might make mistakes due to their unfamiliarity with the workflow. One EN expressed, "When I was paired with TNs, I had this constant fear that they might make errors. Working on the frontline of emergency care is no easy task" (N14). The aftermath of the Shanghai COVID-19 pandemic left several ENs with traumatic flashbacks, avoidance of trauma-related triggers and sleep disturbances. One EN revealed, "I dreaded receiving a message in our WeChat workgroup indicating that one of our patients had tested positive. This triggered a series of tasks, including gathering patient information, identifying close contacts (including family members and neighboring patients), temporarily relocating patients, disinfecting the environment, reporting to the Department of Hospital Infection Management and arranging patient transfers to designated hospitals… It was overwhelming" (N1). Another nurse mentioned, "My sleep was consistently disrupted by nightmares that frequently replayed scenes of busyness" (N17). Worry The majority of interviewees expressed concerns regarding various challenges arising from centralised isolation and strained medical resources. One EN noted, "I wasn't afraid of the virus itself, but I was apprehensive about being placed in centralised quarantine. If I were to become infected, my family and colleagues would also have to undergo quarantine due to their close contact with me" (N10). With the emergence of the epidemic and the depletion of medical resources, some nurses voiced their concerns about the adequacy of rescue and treatment efforts. One EN mentioned, "Every day when I arrived at work, my first task was to check the emergency equipment inventory to identify any shortages. It was a constant source of worry" (N1). Another nurse shared, "Due to concerns about the epidemic, the medical pneumatic logistic transmission systems were halted. Specimen transportation relied on workers, but the heavy workload raised concerns about potential delays in patient treatment" (N13). As the duration of the Shanghai COVID-19 outbreak persisted, a few participants expressed concerns about their own health and the well-being of their families. One EN revealed, "I was extremely fatigued at night but just couldn't sleep. Despite frontline staff receiving injections of thymosin alpha-1 (a drug that enhances immunity), I worried that my health might be compromised" (N8). Another nurse conveyed, "Because of the lockdown measures, I had been away from home for 40 days. I had concerns about my child, even though we had regular WeChat video calls. I worried about whether he could adapt to online classes and whether prolonged computer use might harm his eyes" (N7). Reconstruction Resilience The majority of interviewees described actively engaging in psychological resilience and cultivating robust psychological defence mechanisms, including fostering positive self-perception, self-motivation and a focus on the future. Self-motivation played a pivotal role in bolstering their resilience, diminishing self-criticism and boosting self-esteem. One EN shared, "Allowing anxiety to fester can harm one's health and quality of life. I constantly remind myself that I am a competent EN, and the pandemic will eventually come to an end!" (N5). Individuals with high levels of resilience exhibited a distinctive perspective and a positive outlook. They drew strength from their well-balanced inner qualities and directed their attention towards the future. Another nurse expressed, "Excessive darkness and sadness can be detrimental to one's well-being, and wisdom lies in walking in the light. Beyond work, my life is filled with delightful food, a loving family and cherished children…" (N8). Another EN remarked, "Instead of dwelling on regrets, we should extract lessons from our experiences and move forward, as the song 'Tomorrow Will Be Better' suggests. 'The best is yet to come,' as the saying goes. (Hahaha)" (N13). Active coping All interviewees actively embraced coping strategies, including socialising with positive individuals, engaging in mindfulness meditation, pursuing physical fitness and exercise, and practising downward comparison. One EN illustrated this by saying, "There's a Chinese saying, 'one who stays near vermilion gets stained red, and one who stays near ink gets stained black.' My roommate is incredibly optimistic, and I feel upbeat when I'm around her (laughs)" (N7). Mindfulness meditation served as a valuable technique for mental self-regulation. 9 Another nurse explained, "I allocate time for meditation to divert my attention from my demanding work" (N11). Engaging in fitness and exercise not only promotes cardiovascular health but also helps maintain healthy blood sugar and cholesterol levels, leading to increased energy and stress relief. One EN shared, "I shifted my focus to activities I enjoy rather than dwelling on negative thoughts. After the quarantine ended, I continued running, working up a good sweat and feeling happy" (N17). Practising downward comparison and maintaining an open-minded outlook proved effective in overcoming challenging situations at work. One EN reflected, "I was heartbroken when I witnessed a 2-year-old boy choking on a piece of beef and his parents crying until they fainted in the ED… No matter what happens, I have a stable job, a wonderful family and a precious child" (N10). Social support Individual healing can be a formidable challenge, and all respondents underscored the vital importance of social support, which encompassed organisational support, family support and peer support. The Chinese culture and system place significant emphasis on teamwork and collective efforts to achieve common goals, a principle fully reflected in the fight against the Shanghai COVID-19 pandemic. One EN recounted, "The hospital management deployed additional staff and volunteers to the ED and increased compensation for high-risk positions. The hospital's labor union even provided us with rice dumplings for the Dragon Boat Festival (Hahaha)" (N11). Chinese families are known for their strong bonds, and family support played a pivotal role in personal reconstruction. One nurse shared, "On Children's Day, 1 June, my daughter's school honored online the anti-epidemic heroes. When she saw my photo, she exclaimed, "Mom is a superwoman." My daughter is my biggest source of motivation to keep moving forward!" (N15). Another EN mentioned, "My husband created a Nurses' Day video for me. Even though we couldn't be together, it brought me immense joy!" (N7). Peer support also proved to be a valuable form of social support. One EN expressed, "My friend who specialized in traditional Chinese medicine helped me with acupressure using beans. It slightly improved my sleep quality (smiled)" (N4). Growth Spiritual dimensions All interviewees reported growth in spiritual dimensions, including cultivating positive thinking, self-identity and gratitude. Positive psychology posits that positive thinking leads to a more wholesome perspective on life, a positive attitude and increased resilience. One EN affirmed, "You can't see the rainbow without enduring the storm. I believe that every day in the future will be filled with hope" (N16). Psychological research has shown that these positive shifts in spiritual dimensions, such as self-identity, often occur following traumatic events and are instrumental in promoting mental well-being. 8 One EN shared, "When I enter my work number into Enterprise WeChat, it generates an electronic anti-epidemic memorial certificate. I've completed 55 days of anti-epidemic work, and looking back, it has all been so challenging! I feel fantastic!" (N8). A spirit of gratitude is regarded as essential for leading a meaningful life, and all interviewees expressed their gratitude and adopted a new perspective on life. One EN remarked, "Many individuals assisted us during the epidemic, and I often reflect on these moments of gratitude" (N1). Behavioural dimensions All interviewees reported experiencing growth in behavioural dimensions, which encompassed individual empowerment, enhanced interpersonal closeness and newfound planning. PTG often leads to individual empowerment, signifying an increased ability to confront crises, which includes heightened strength and the acquisition of new skills. One EN shared, "Following the Shanghai COVID-19 pandemic, I enrolled in a 1-month training course at a reputable third-level hospital. This expanded my horizons and equipped me with new skills, such as extracorporeal membrane oxygenation" (N9). The majority of interviewees noted that they had become closer to their families and developed greater empathy for others. One EN humorously remarked, "My wife loves me more now (smiles) and thinks I'm a hardworking person" (N15). Empathy is known to have a bidirectional effect, 23 and many interviewees found that empathising with patients helped establish emotional connections and offered professional benefits. Another EN explained, "I used to be solely focused on emergency care, but now I take the time to listen to patients. A patient's family often comments, "This young person with glasses is really kind." I truly sense the professional rewards of empathy" (N12). Several interviewees had new plans for their futures. One EN expressed, "There's a saying, 'To forge iron, one must be strong,' and I hope to further refine some of my skills, like intraosseous infusion and continuous renal replacement therapy, in the future" (N3). Many individuals who have experienced a crisis are more inclined to help others and explore new opportunities. One EN mentioned, "I joined a volunteer group to contribute to society" (N6). Another EN shared, "In the future, I aspire to enhance the efficiency of organ procurement organizations to assist more people in need" (N5). Stress Suffering All interviewees reported enduring physical and psychological hardships due to the use of personal protective equipment (PPE), an unexpected surge in workload and the demanding rescue procedures. One participant expressed, "Wearing the PPE made resuscitation a challenge. Securing a catheter for a ventilated patient was much harder due to the obstructed view caused by the face shield. To save time, I went without food or water for 8 straight hours. The clothing under the PPE was completely soaked after just one shift" (N9). Another nurse shared, "Despite numerous patients being stranded in the ED, we kept receiving a continuous flow of new cases. In addition to emergency cases, we had to identify and treat COVID-19 patients. It was common for multiple cases of respiratory failure requiring intubation to occur simultaneously, keeping me extremely busy. The sound of an ambulance siren felt like being caught in an overwhelming whirlwind" (N12). One nurse remarked, "Some specialists had to come from their respective wards and spend time donning PPE. Patients waited for extended periods and argued with me, which was quite frustrating" (N4). In response to the Shanghai COVID-19 pandemic, ENs faced prolonged separation from their families. The conflict between their familial responsibilities and work commitments proved to be a source of trauma and distress. One EN shared, "My wife was on maternity leave, and my children and parents contracted the virus. I couldn't be there to take care of them due to work, and it left me feeling profoundly saddened" (N15). Fear ENs grappled with various fears during the Shanghai COVID-19 pandemic, primarily stemming from the uncertainties surrounding the virus, the fear of contracting it themselves and the potential for transmitting the infection to other patients due to errors in their work. One EN stated, "I constantly worried about being infected with the virus, and I checked my nucleic acid test results every afternoon. If the result didn't show up, I would become anxious" (N3). Another shared, "I worked in the pre-screening triage office, and I was constantly anxious about the possibility of inadvertently allowing infected patients into an area where uninfected patients were being treated. The pressure was immense" (N6). While hospital management made efforts to deploy additional nursing resources and brought in transdisciplinary nurses (TNs) to support the EDs, the ENs still experienced anxiety. This stemmed from concerns that the TNs, who had not previously worked in EDs, might make mistakes due to their unfamiliarity with the workflow. One EN expressed, "When I was paired with TNs, I had this constant fear that they might make errors. Working on the frontline of emergency care is no easy task" (N14). The aftermath of the Shanghai COVID-19 pandemic left several ENs with traumatic flashbacks, avoidance of trauma-related triggers and sleep disturbances. One EN revealed, "I dreaded receiving a message in our WeChat workgroup indicating that one of our patients had tested positive. This triggered a series of tasks, including gathering patient information, identifying close contacts (including family members and neighboring patients), temporarily relocating patients, disinfecting the environment, reporting to the Department of Hospital Infection Management and arranging patient transfers to designated hospitals… It was overwhelming" (N1). Another nurse mentioned, "My sleep was consistently disrupted by nightmares that frequently replayed scenes of busyness" (N17). Worry The majority of interviewees expressed concerns regarding various challenges arising from centralised isolation and strained medical resources. One EN noted, "I wasn't afraid of the virus itself, but I was apprehensive about being placed in centralised quarantine. If I were to become infected, my family and colleagues would also have to undergo quarantine due to their close contact with me" (N10). With the emergence of the epidemic and the depletion of medical resources, some nurses voiced their concerns about the adequacy of rescue and treatment efforts. One EN mentioned, "Every day when I arrived at work, my first task was to check the emergency equipment inventory to identify any shortages. It was a constant source of worry" (N1). Another nurse shared, "Due to concerns about the epidemic, the medical pneumatic logistic transmission systems were halted. Specimen transportation relied on workers, but the heavy workload raised concerns about potential delays in patient treatment" (N13). As the duration of the Shanghai COVID-19 outbreak persisted, a few participants expressed concerns about their own health and the well-being of their families. One EN revealed, "I was extremely fatigued at night but just couldn't sleep. Despite frontline staff receiving injections of thymosin alpha-1 (a drug that enhances immunity), I worried that my health might be compromised" (N8). Another nurse conveyed, "Because of the lockdown measures, I had been away from home for 40 days. I had concerns about my child, even though we had regular WeChat video calls. I worried about whether he could adapt to online classes and whether prolonged computer use might harm his eyes" (N7). Suffering All interviewees reported enduring physical and psychological hardships due to the use of personal protective equipment (PPE), an unexpected surge in workload and the demanding rescue procedures. One participant expressed, "Wearing the PPE made resuscitation a challenge. Securing a catheter for a ventilated patient was much harder due to the obstructed view caused by the face shield. To save time, I went without food or water for 8 straight hours. The clothing under the PPE was completely soaked after just one shift" (N9). Another nurse shared, "Despite numerous patients being stranded in the ED, we kept receiving a continuous flow of new cases. In addition to emergency cases, we had to identify and treat COVID-19 patients. It was common for multiple cases of respiratory failure requiring intubation to occur simultaneously, keeping me extremely busy. The sound of an ambulance siren felt like being caught in an overwhelming whirlwind" (N12). One nurse remarked, "Some specialists had to come from their respective wards and spend time donning PPE. Patients waited for extended periods and argued with me, which was quite frustrating" (N4). In response to the Shanghai COVID-19 pandemic, ENs faced prolonged separation from their families. The conflict between their familial responsibilities and work commitments proved to be a source of trauma and distress. One EN shared, "My wife was on maternity leave, and my children and parents contracted the virus. I couldn't be there to take care of them due to work, and it left me feeling profoundly saddened" (N15). Fear ENs grappled with various fears during the Shanghai COVID-19 pandemic, primarily stemming from the uncertainties surrounding the virus, the fear of contracting it themselves and the potential for transmitting the infection to other patients due to errors in their work. One EN stated, "I constantly worried about being infected with the virus, and I checked my nucleic acid test results every afternoon. If the result didn't show up, I would become anxious" (N3). Another shared, "I worked in the pre-screening triage office, and I was constantly anxious about the possibility of inadvertently allowing infected patients into an area where uninfected patients were being treated. The pressure was immense" (N6). While hospital management made efforts to deploy additional nursing resources and brought in transdisciplinary nurses (TNs) to support the EDs, the ENs still experienced anxiety. This stemmed from concerns that the TNs, who had not previously worked in EDs, might make mistakes due to their unfamiliarity with the workflow. One EN expressed, "When I was paired with TNs, I had this constant fear that they might make errors. Working on the frontline of emergency care is no easy task" (N14). The aftermath of the Shanghai COVID-19 pandemic left several ENs with traumatic flashbacks, avoidance of trauma-related triggers and sleep disturbances. One EN revealed, "I dreaded receiving a message in our WeChat workgroup indicating that one of our patients had tested positive. This triggered a series of tasks, including gathering patient information, identifying close contacts (including family members and neighboring patients), temporarily relocating patients, disinfecting the environment, reporting to the Department of Hospital Infection Management and arranging patient transfers to designated hospitals… It was overwhelming" (N1). Another nurse mentioned, "My sleep was consistently disrupted by nightmares that frequently replayed scenes of busyness" (N17). Worry The majority of interviewees expressed concerns regarding various challenges arising from centralised isolation and strained medical resources. One EN noted, "I wasn't afraid of the virus itself, but I was apprehensive about being placed in centralised quarantine. If I were to become infected, my family and colleagues would also have to undergo quarantine due to their close contact with me" (N10). With the emergence of the epidemic and the depletion of medical resources, some nurses voiced their concerns about the adequacy of rescue and treatment efforts. One EN mentioned, "Every day when I arrived at work, my first task was to check the emergency equipment inventory to identify any shortages. It was a constant source of worry" (N1). Another nurse shared, "Due to concerns about the epidemic, the medical pneumatic logistic transmission systems were halted. Specimen transportation relied on workers, but the heavy workload raised concerns about potential delays in patient treatment" (N13). As the duration of the Shanghai COVID-19 outbreak persisted, a few participants expressed concerns about their own health and the well-being of their families. One EN revealed, "I was extremely fatigued at night but just couldn't sleep. Despite frontline staff receiving injections of thymosin alpha-1 (a drug that enhances immunity), I worried that my health might be compromised" (N8). Another nurse conveyed, "Because of the lockdown measures, I had been away from home for 40 days. I had concerns about my child, even though we had regular WeChat video calls. I worried about whether he could adapt to online classes and whether prolonged computer use might harm his eyes" (N7). Reconstruction Resilience The majority of interviewees described actively engaging in psychological resilience and cultivating robust psychological defence mechanisms, including fostering positive self-perception, self-motivation and a focus on the future. Self-motivation played a pivotal role in bolstering their resilience, diminishing self-criticism and boosting self-esteem. One EN shared, "Allowing anxiety to fester can harm one's health and quality of life. I constantly remind myself that I am a competent EN, and the pandemic will eventually come to an end!" (N5). Individuals with high levels of resilience exhibited a distinctive perspective and a positive outlook. They drew strength from their well-balanced inner qualities and directed their attention towards the future. Another nurse expressed, "Excessive darkness and sadness can be detrimental to one's well-being, and wisdom lies in walking in the light. Beyond work, my life is filled with delightful food, a loving family and cherished children…" (N8). Another EN remarked, "Instead of dwelling on regrets, we should extract lessons from our experiences and move forward, as the song 'Tomorrow Will Be Better' suggests. 'The best is yet to come,' as the saying goes. (Hahaha)" (N13). Active coping All interviewees actively embraced coping strategies, including socialising with positive individuals, engaging in mindfulness meditation, pursuing physical fitness and exercise, and practising downward comparison. One EN illustrated this by saying, "There's a Chinese saying, 'one who stays near vermilion gets stained red, and one who stays near ink gets stained black.' My roommate is incredibly optimistic, and I feel upbeat when I'm around her (laughs)" (N7). Mindfulness meditation served as a valuable technique for mental self-regulation. 9 Another nurse explained, "I allocate time for meditation to divert my attention from my demanding work" (N11). Engaging in fitness and exercise not only promotes cardiovascular health but also helps maintain healthy blood sugar and cholesterol levels, leading to increased energy and stress relief. One EN shared, "I shifted my focus to activities I enjoy rather than dwelling on negative thoughts. After the quarantine ended, I continued running, working up a good sweat and feeling happy" (N17). Practising downward comparison and maintaining an open-minded outlook proved effective in overcoming challenging situations at work. One EN reflected, "I was heartbroken when I witnessed a 2-year-old boy choking on a piece of beef and his parents crying until they fainted in the ED… No matter what happens, I have a stable job, a wonderful family and a precious child" (N10). Social support Individual healing can be a formidable challenge, and all respondents underscored the vital importance of social support, which encompassed organisational support, family support and peer support. The Chinese culture and system place significant emphasis on teamwork and collective efforts to achieve common goals, a principle fully reflected in the fight against the Shanghai COVID-19 pandemic. One EN recounted, "The hospital management deployed additional staff and volunteers to the ED and increased compensation for high-risk positions. The hospital's labor union even provided us with rice dumplings for the Dragon Boat Festival (Hahaha)" (N11). Chinese families are known for their strong bonds, and family support played a pivotal role in personal reconstruction. One nurse shared, "On Children's Day, 1 June, my daughter's school honored online the anti-epidemic heroes. When she saw my photo, she exclaimed, "Mom is a superwoman." My daughter is my biggest source of motivation to keep moving forward!" (N15). Another EN mentioned, "My husband created a Nurses' Day video for me. Even though we couldn't be together, it brought me immense joy!" (N7). Peer support also proved to be a valuable form of social support. One EN expressed, "My friend who specialized in traditional Chinese medicine helped me with acupressure using beans. It slightly improved my sleep quality (smiled)" (N4). Resilience The majority of interviewees described actively engaging in psychological resilience and cultivating robust psychological defence mechanisms, including fostering positive self-perception, self-motivation and a focus on the future. Self-motivation played a pivotal role in bolstering their resilience, diminishing self-criticism and boosting self-esteem. One EN shared, "Allowing anxiety to fester can harm one's health and quality of life. I constantly remind myself that I am a competent EN, and the pandemic will eventually come to an end!" (N5). Individuals with high levels of resilience exhibited a distinctive perspective and a positive outlook. They drew strength from their well-balanced inner qualities and directed their attention towards the future. Another nurse expressed, "Excessive darkness and sadness can be detrimental to one's well-being, and wisdom lies in walking in the light. Beyond work, my life is filled with delightful food, a loving family and cherished children…" (N8). Another EN remarked, "Instead of dwelling on regrets, we should extract lessons from our experiences and move forward, as the song 'Tomorrow Will Be Better' suggests. 'The best is yet to come,' as the saying goes. (Hahaha)" (N13). Active coping All interviewees actively embraced coping strategies, including socialising with positive individuals, engaging in mindfulness meditation, pursuing physical fitness and exercise, and practising downward comparison. One EN illustrated this by saying, "There's a Chinese saying, 'one who stays near vermilion gets stained red, and one who stays near ink gets stained black.' My roommate is incredibly optimistic, and I feel upbeat when I'm around her (laughs)" (N7). Mindfulness meditation served as a valuable technique for mental self-regulation. 9 Another nurse explained, "I allocate time for meditation to divert my attention from my demanding work" (N11). Engaging in fitness and exercise not only promotes cardiovascular health but also helps maintain healthy blood sugar and cholesterol levels, leading to increased energy and stress relief. One EN shared, "I shifted my focus to activities I enjoy rather than dwelling on negative thoughts. After the quarantine ended, I continued running, working up a good sweat and feeling happy" (N17). Practising downward comparison and maintaining an open-minded outlook proved effective in overcoming challenging situations at work. One EN reflected, "I was heartbroken when I witnessed a 2-year-old boy choking on a piece of beef and his parents crying until they fainted in the ED… No matter what happens, I have a stable job, a wonderful family and a precious child" (N10). Social support Individual healing can be a formidable challenge, and all respondents underscored the vital importance of social support, which encompassed organisational support, family support and peer support. The Chinese culture and system place significant emphasis on teamwork and collective efforts to achieve common goals, a principle fully reflected in the fight against the Shanghai COVID-19 pandemic. One EN recounted, "The hospital management deployed additional staff and volunteers to the ED and increased compensation for high-risk positions. The hospital's labor union even provided us with rice dumplings for the Dragon Boat Festival (Hahaha)" (N11). Chinese families are known for their strong bonds, and family support played a pivotal role in personal reconstruction. One nurse shared, "On Children's Day, 1 June, my daughter's school honored online the anti-epidemic heroes. When she saw my photo, she exclaimed, "Mom is a superwoman." My daughter is my biggest source of motivation to keep moving forward!" (N15). Another EN mentioned, "My husband created a Nurses' Day video for me. Even though we couldn't be together, it brought me immense joy!" (N7). Peer support also proved to be a valuable form of social support. One EN expressed, "My friend who specialized in traditional Chinese medicine helped me with acupressure using beans. It slightly improved my sleep quality (smiled)" (N4). Growth Spiritual dimensions All interviewees reported growth in spiritual dimensions, including cultivating positive thinking, self-identity and gratitude. Positive psychology posits that positive thinking leads to a more wholesome perspective on life, a positive attitude and increased resilience. One EN affirmed, "You can't see the rainbow without enduring the storm. I believe that every day in the future will be filled with hope" (N16). Psychological research has shown that these positive shifts in spiritual dimensions, such as self-identity, often occur following traumatic events and are instrumental in promoting mental well-being. 8 One EN shared, "When I enter my work number into Enterprise WeChat, it generates an electronic anti-epidemic memorial certificate. I've completed 55 days of anti-epidemic work, and looking back, it has all been so challenging! I feel fantastic!" (N8). A spirit of gratitude is regarded as essential for leading a meaningful life, and all interviewees expressed their gratitude and adopted a new perspective on life. One EN remarked, "Many individuals assisted us during the epidemic, and I often reflect on these moments of gratitude" (N1). Behavioural dimensions All interviewees reported experiencing growth in behavioural dimensions, which encompassed individual empowerment, enhanced interpersonal closeness and newfound planning. PTG often leads to individual empowerment, signifying an increased ability to confront crises, which includes heightened strength and the acquisition of new skills. One EN shared, "Following the Shanghai COVID-19 pandemic, I enrolled in a 1-month training course at a reputable third-level hospital. This expanded my horizons and equipped me with new skills, such as extracorporeal membrane oxygenation" (N9). The majority of interviewees noted that they had become closer to their families and developed greater empathy for others. One EN humorously remarked, "My wife loves me more now (smiles) and thinks I'm a hardworking person" (N15). Empathy is known to have a bidirectional effect, 23 and many interviewees found that empathising with patients helped establish emotional connections and offered professional benefits. Another EN explained, "I used to be solely focused on emergency care, but now I take the time to listen to patients. A patient's family often comments, "This young person with glasses is really kind." I truly sense the professional rewards of empathy" (N12). Several interviewees had new plans for their futures. One EN expressed, "There's a saying, 'To forge iron, one must be strong,' and I hope to further refine some of my skills, like intraosseous infusion and continuous renal replacement therapy, in the future" (N3). Many individuals who have experienced a crisis are more inclined to help others and explore new opportunities. One EN mentioned, "I joined a volunteer group to contribute to society" (N6). Another EN shared, "In the future, I aspire to enhance the efficiency of organ procurement organizations to assist more people in need" (N5). Spiritual dimensions All interviewees reported growth in spiritual dimensions, including cultivating positive thinking, self-identity and gratitude. Positive psychology posits that positive thinking leads to a more wholesome perspective on life, a positive attitude and increased resilience. One EN affirmed, "You can't see the rainbow without enduring the storm. I believe that every day in the future will be filled with hope" (N16). Psychological research has shown that these positive shifts in spiritual dimensions, such as self-identity, often occur following traumatic events and are instrumental in promoting mental well-being. 8 One EN shared, "When I enter my work number into Enterprise WeChat, it generates an electronic anti-epidemic memorial certificate. I've completed 55 days of anti-epidemic work, and looking back, it has all been so challenging! I feel fantastic!" (N8). A spirit of gratitude is regarded as essential for leading a meaningful life, and all interviewees expressed their gratitude and adopted a new perspective on life. One EN remarked, "Many individuals assisted us during the epidemic, and I often reflect on these moments of gratitude" (N1). Behavioural dimensions All interviewees reported experiencing growth in behavioural dimensions, which encompassed individual empowerment, enhanced interpersonal closeness and newfound planning. PTG often leads to individual empowerment, signifying an increased ability to confront crises, which includes heightened strength and the acquisition of new skills. One EN shared, "Following the Shanghai COVID-19 pandemic, I enrolled in a 1-month training course at a reputable third-level hospital. This expanded my horizons and equipped me with new skills, such as extracorporeal membrane oxygenation" (N9). The majority of interviewees noted that they had become closer to their families and developed greater empathy for others. One EN humorously remarked, "My wife loves me more now (smiles) and thinks I'm a hardworking person" (N15). Empathy is known to have a bidirectional effect, 23 and many interviewees found that empathising with patients helped establish emotional connections and offered professional benefits. Another EN explained, "I used to be solely focused on emergency care, but now I take the time to listen to patients. A patient's family often comments, "This young person with glasses is really kind." I truly sense the professional rewards of empathy" (N12). Several interviewees had new plans for their futures. One EN expressed, "There's a saying, 'To forge iron, one must be strong,' and I hope to further refine some of my skills, like intraosseous infusion and continuous renal replacement therapy, in the future" (N3). Many individuals who have experienced a crisis are more inclined to help others and explore new opportunities. One EN mentioned, "I joined a volunteer group to contribute to society" (N6). Another EN shared, "In the future, I aspire to enhance the efficiency of organ procurement organizations to assist more people in need" (N5). Discussion ENs faced traumatic events amid the chaotic healthcare environment and strained medical resources. This study delineated the PTG experienced by ENs during the Shanghai COVID-19 pandemic into three stages: stress, restructuring and growth. This observation underscores that PTG is a dynamic process, aligning with the research findings of Tedeschi and Calhoun, 8 which suggest that PTG unfolds gradually in the ongoing struggle with stress. The trajectory of PTG identified in this study also parallels the conclusions drawn from previous research. 24 25 As a hospital manager, it is imperative to offer comprehensive support to ENs at each juncture of their post-traumatic journey and implement targeted psychological interventions tailored to the specific stages of PTG. ENs who have endured the pandemic have faced stressors, including suffering, fear and worry. Trauma can shatter an individual's world view and pose a significant challenge to coping with stress. A study examining the psychological status of 41 frontline nurses combating COVID-19 at Wuhan Union Hospital (Wuhan, China) over a 2-week period found that 92.68% of nurses experienced psychological issues. These problems included anxiety in 32 nurses, somatisation disorder in 11 nurses, obsession in 8 nurses and depression in 2 nurses. 26 Therefore, effective stress management following early trauma is of paramount importance. Managers can draw upon authoritative literature on psychological support and stress management strategies to expedite ENs' recovery from stress. For instance, the SAFER-R model, 27 developed by Dr Everly, encompasses five key areas: stabilisation, acknowledging the crisis, facilitating understanding, encouraging effective coping and referring for recovery. It is considered an effective trauma-informed response strategy. Additionally, critical incident stress management is a comprehensive peer support programme designed to address psychological crises following traumatic events. Furthermore, common early interventions for psychological crises include psychological first aid, critical incident stress debriefing, trauma-focused cognitive–behavioural therapy, eye movement desensitisation and reprocessing, pharmacotherapy and Balint group. 28 It is important to note that in this study, the stressors experienced by ENs during the COVID-19 pandemic in Shanghai primarily revolved around concerns related to infection, a shortage of medical resources and excessive workloads. Therefore, in addition to the aforementioned stress management strategies, hospital management should optimise the allocation of medical resources and adopt an ambulatory management approach to ensure a rational distribution of resources. 29 Simultaneously, increasing investment in medical resource development and promoting their information-driven and intelligent utilisation is crucial. Providing adequate care and support to ENs is imperative. Comprehensive emergency training programmes and opportunities can enhance their response capabilities and self-protection skills while reducing anxiety levels among nurses, which aligns with previous research findings. 30 By implementing these measures comprehensively, it is possible to alleviate trauma and mitigate the pressures faced by nurses during challenging periods. PTG theory posits that stressors, such as suffering, fear and worry, triggered by traumatic events can serve as indicators of cognitive restructuring in response to these events. Resilience may play a pivotal role in influencing the development of cognitive restructuring. Therefore, enhancing the resilience of ENs during the restructuring phase can facilitate the achievement of PTG. 31 Resilience, defined as the ability to bounce back when individuals face adversity, trauma, misfortune or change, encompasses positive feedback, self-motivation and a focus on the future. Resilient individuals tend to maintain emotional composure and are more likely to successfully complete cognitive restructuring. Improving public emergency response capabilities, establishing resilient teams and enhancing salary incentives are effective methods to promote resilience. 3 Among these, the establishment of a resilient team is closely tied to resilient leadership and fostering a conducive team atmosphere. 32 In addition to employing various training methods to enhance the resilience of ENs, managers should also prioritise the development of leaders' resilience and foster a positive team culture to strengthen internal cohesion within the team. Furthermore, our study unveiled that many nurses adopted positive coping strategies during the restructuring process, such as engaging in fitness and meditation, as well as 'rationalising' their current circumstances and using downward comparisons to identify their strengths. This finding aligns with the conclusions of numerous previous studies. 33 34 Programmes like Stress Management and Resiliency Training, mindfulness courses and educational seminars on compassion fatigue have proven to be effective. Additionally, traditional Chinese culture should be considered as a novel dimension of the strategies offered. Traditional Chinese philosophies have deeply explored the relationship between humans and nature, exemplified by concepts like the five elements and eight trigrams. Integrating traditional Chinese culture may influence the thinking patterns of ENs and assist them in better coping with traumatic events. Strategies like the five-tone therapy (Gong, Shang, Jue, Zhi, Yu), Baduanjin 3 and Yikong skills can all serve as effective tools in this regard. Social support plays a crucial role in the process of post-traumatic restructuring for ENs. 35 Tedeschi and Calhoun categorised social factors into distal and proximal categories as key contributors to PTG. 8 Among proximal factors, organisations, family and peers are the factors with which individuals have the most direct interaction. Distal factors refer to the prevailing cultural perspective on PTG. Traditional Chinese culture has a deep understanding of the concept that suffering and trauma can foster personal growth. Buddhism, for instance, teaches that 'Everyone has hardships to bear, and individuals must overcome them on their own', providing a solid foundation for PTG. In response to the outbreak, the Chinese government implemented a policy that embodied the idea of 'When trouble occurs at one spot, help comes from all quarters'. On 15 May 2022, 25 000 healthcare providers from 12 provinces were deployed to assist with the Shanghai COVID-19 pandemic. 36 A strong sense of national pride and a psychological sense of belonging to the country propelled ENs towards growth amid the continuous challenges of trauma and cognitive restructuring. In addition to the support of their families and loved ones, managers can enhance the PTG of ENs by providing support from hospitals, organisations and society. This entails shaping public opinion, emphasising the remarkable contributions of nurses, elevating their societal status, attentively addressing their intrinsic needs, establishing a psychological counselling platform to promptly address ENs' negative emotions, strategically distributing work responsibilities, improving compensation packages and facilitating ample opportunities for professional development through external learning and training programmes. The effectiveness of each of these support strategies has been discussed in previous literature. 30 Growth, encompassing both spiritual and behavioural dimensions, serves as an indicator of successful coping with trauma. ENs who have weathered the pandemic underwent cognitive restructuring, revealing latent potentials within themselves, which can be interpreted as a form of growth. They cultivated a positive mindset towards life, embraced self-identification and gratitude, and harnessed personal resilience, acquiring new skills in the process. ENs also deepened their empathy and altruism, which facilitated their problem-solving abilities, and they harboured greater aspirations for their professional careers through new planning. Conclusions In conclusion, this qualitative study uncovered three distinct stages of PTG among ENs who confronted the challenges of the Shanghai COVID-19 pandemic: stress, restructuring and growth. Nurse managers should recognise that ENs at different phases require tailored interventions to enhance their mental well-being and achieve higher levels of PTG ( figure 2 ). Figure 2 Interventional strategies for emergency nurses' post-traumatic growth. Study limitations This research did not consider specific factors that may influence PTG among ENs, such as religious affiliation, gender and individual experiences of traumatic events. Additionally, the findings cannot be generalised to all healthcare professionals due to the inherent characteristics of qualitative research. Future replications of this study in various clinical settings, with larger and more diverse samples from various cultural backgrounds, are needed to yield a broader range of insights. Furthermore, involving patients and the public in future studies could contribute to more diverse and enriched conclusions. Supplementary Material Reviewer comments Author's manuscript Data availability statement Data are available upon reasonable request. All data can be obtained by contacting the corresponding author. Ethics statements Patient consent for publication Not applicable. Ethics approval This study involves human participants and received approval from the Institutional Review Board of Shanghai Tenth People's Hospital (Shanghai, China) under approval no. 22KN08. Prior to the formal interviews, each participant received a clear explanation of the study's objectives and provided informed consent. To protect participant confidentiality, their real identities were replaced with numerical identifiers (eg, N1). Patient consent for publication Not applicable. Ethics approval This study involves human participants and received approval from the Institutional Review Board of Shanghai Tenth People's Hospital (Shanghai, China) under approval no. 22KN08. Prior to the formal interviews, each participant received a clear explanation of the study's objectives and provided informed consent. To protect participant confidentiality, their real identities were replaced with numerical identifiers (eg, N1).

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3042556/

Activation of the Edema Factor of Bacillus anthracis by Calmodulin: Evidence of an Interplay between the EF-Calmodulin Interaction and Calcium Binding

Calmodulin (CaM) is a remarkably flexible protein which can bind multiple targets in response to changes in intracellular calcium concentration. It contains four calcium-binding sites, arranged in two globular domains. The calcium affinity of CaM N-terminal domain (N-CaM) is dramatically reduced when the complex with the edema factor (EF) of Bacillus anthracis is formed. Here, an atomic explanation for this reduced affinity is proposed through molecular dynamics simulations and free energy perturbation calculations of the EF-CaM complex starting from different crystallographic models. The simulations show that electrostatic interactions between CaM and EF disfavor the opening of N-CaM domains usually induced by calcium binding. Relative calcium affinities of the N-CaM binding sites are probed by free energy perturbation, and dissociation probabilities are evaluated with locally enhanced sampling simulations. We show that EF impairs calcium binding on N-CaM through a direct conformational restraint on Site 1, by an indirect destabilization of Site 2, and by reducing the cooperativity between the two sites.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10272415/

Dengue virus neutralizing antibody: a review of targets, cross-reactivity, and antibody-dependent enhancement

Dengue is the most common viral infection spread by mosquitoes, prevalent in tropical countries. The acute dengue virus (DENV) infection is a benign and primarily febrile illness. However, secondary infection with alternative serotypes can worsen the condition, leading to severe and potentially fatal dengue. The antibody raised by the vaccine or the primary infections are frequently cross-reactive; however, weakly neutralizing, and during subsequent infection, they may increase the odds of antibody-dependent enhancement (ADE). Despite that, many neutralizing antibodies have been identified against the DENV, which are thought to be useful in reducing dengue severity. Indeed, an antibody must be free from ADE for therapeutic application, as it is pretty common in dengue infection and escalates disease severity. Therefore, this review has described the critical characteristics of DENV and the potential immune targets in general. The primary emphasis is given to the envelope protein of DENV, where potential epitopes targeted for generating serotype-specific and cross-reactive antibodies have critically been described. In addition, a novel class of highly neutralizing antibodies targeted to the quaternary structure, similar to viral particles, has also been described. Lastly, we have discussed different aspects of the pathogenesis and ADE, which would provide significant insights into developing safe and effective antibody therapeutics and equivalent protein subunit vaccines. 1 Introduction Dengue epidemiology has grown dramatically with dengue fever, now endemic in 128 tropical and subtropical nations, where more than half of the world's population lives. The data analysis from Global Burden of Diseases, Injuries, and Risk Factors Study (GBD) 2017 revealed 107.6% increase in disability-adjusted life years (DALYs) since 1990. Dengue incident cases and age-standardized incidence rates (ASR) increased from 30.7 million and 557.15 per 100,000 people in 1990 to 56.9 million and 740.4 per 100,000 in 2019, respectively, as per the GBD 2019 data study. The three largest endemic regions with the greatest dengue incidence in 2019 are Oceania, South Asia, and Southeast Asia ( 1 ). Dengue has spread throughout the Pacific and the Americas as a result of international trade and tourism ( 2 ). Due to worldwide travel and warming, the prevalence of dengue is rapidly rising ( 3 ). Dengue virus (DENV) is one of the most significant members of the Flaviviridae family, which has four different serotypes (DENV1-4) that only differ by 30–35% of the amino acid sequence ( 4 ), and a large number of genotypes within each serotype that varries only by 6–8% nucleotide and 3% amino acid sequence ( 5 ). All the variants are infectious and can result in a variety of clinical conditions, from mild dengue fever to severe dengue ( 6 ). Every serotype's primary infection confers lifelong immunity against it, and usually recovers after a self-limiting condition ( 7 ). According to the updated dengue case classification, recurrent infection with a heterologous serotype only confers partial or transitory protection, often leading to dengue with or without warning signs (DWS+/-) and severe dengue (SD) ( 8 , 9 ). Serotype-cross-reactive antibodies may also significantly contribute to the spread of DENV by facilitating viral entry into FcγR-bearing monocytes, a mechanism known as antibody-dependent enhancement (ADE) ( 10 ). The specific antibody response generated after primary infection maintains an immunological memory, and is able to bind and to neutralize a homologous dengue serotype providing lifelong specific-immunity. While in subsequent infection with heterologous serotype, the pre-existing IgG antibodies are able to bind to the heterologous dengue serotype, however, instead of neutralizing the virus, it enhances the infection ( 11 ). This impairment in development of robust immune response against heterologous serotype is due to the phenomenon of original antigenic sin ( 12 , 13 ). The titre of antibodies, targeted towards the virus serotype, are responsible for the primary infection tend to significantly rise and can often remain higher than the levels of antibodies induced towards currently infecting serotype ( 14 ). During recurrent infection, the antibody-mediated amplification of the disease severity sets a high bar to develop a vaccine against DENV. There is currently no such vaccination technique that can completely protect against all four serotypes in naive people ( 15 ). The most leading vaccine, a tetravalent live attenuated candidate (CYD-TDV), has received approval in certain dengue-prevalent nations; however, it has the unintended consequence of raising the risk of SD in individuals who have never been infected by DENV ( 16 ). As a result, the administration of vaccines, and their development have been limited by the concern that vaccine delivery may put naive people at a high risk of alarming dengue ( 17 ). Nonetheless, several alternative attempts have been made towards the development of safe vaccine with efficient protective immunity and to exempt all risky issues ( 18 ). The Laboratory of Infectious Diseases at the National Institutes of Health evaluated a number of monovalent and tetravalent dengue candidate vaccines to determine which had the greatest safety, infectivity, and immunogenicity profiles. TV003/TV005, a vaccine that combines four live attenuated recombinant dengue virus vaccines (rDEN1D30, rDEN2/4D30, rDEN3D30/31, and rDEN4D30), has cleared Phase III clinical trials, and been licenced to a number of manufacturers, including Butantan, VaBiotech, and Merk ( 19 ). Furthermore, various additional recombinant tetravalent vaccines (e.g., DEN-80E, TVDV) have passed phase I clinical trials, expressing the prM and E genes of each of the four DENV serotypes from plasmid DNA ( 18 , 20 ). DENVax candidates, Takeda's live tetravalent dengue vaccine TDV, and a chimeric dengue-2 PDK-53-based tetravalent vaccination are also undergoing preclinical and clinical trials to stimulate humoral and cellular protective immune responses ( 21 ). Most recently a nucleotide-modified mRNA vaccine encoding the membrane and envelope structural proteins from DENV serotype 1 encapsulated in lipid nanoparticles (prM/E mRNA-LNP) was found to induce neutralizing antibody and cellular immune responses in immunocompetent mice. In comparison to a live DENV1 viral infection, this vaccine design displayed serotype-specific protection with little serum cross-reactivity and decreased ADE ( 22 ). Since the pre-existing antibodies either raised by the vaccine or the primary infection may increase the disease severity, none of the vaccine candidates are still allowed for human application. Due to the continued absence of a safe vaccine and suitable drugs, other antiviral measures, especially antibody therapeutics, are becoming an attractive option to reduce SD. Many infectious and non-infectious diseases are now being treated with a variety of therapeutic antibodies ( 23 ). However, antibody therapeutics against DENV is still challenging due to the ADE effect and higher production cost. Several expression systems and protein engineering options are currently being utilized for the economical production of antibodies and to overcome the biosafety concern of ADE ( 24 , 25 ). For the last ten years, our research group has been working on designing and production of monoclonal antibody (mAb) fragments and short peptide vaccines against DENV. The challenges and difficulties observed in developing a safe and effective therapeutic mAb and vaccine motivated us to write this review. Therefore, we explored the NCBI database to find published articles on this topic in the last two decades. In this review, we aim to describe the disease severity, general features, and characteristics of the DENV, along with its surface protein structures, following the different immune targets. Then, we focus on the envelope protein as a potential target in the third part of the review. Here, we deliberate the serotype-specific and cross-reactive epitopes and the therapeutic potential of the neutralizing and cross-reactivity antibodies targeted to these epitopes. After that, the mAbs developed against virus particle-like quaternary structure is described. Lastly, the insights on severe pathogenesis and ADE would provide extended knowledge to design future subunit vaccines and therapeutic mAbs. 2 Dengue virus and immune targets The flavivirus genus includes a group of more than seventy single-stranded, positive-polarity RNA viruses that are mostly spread by arthropod vectors. Most of which are spread by mosquitoes, including the Zika virus (ZIKV), West Nile virus (WNV), Japanese encephalitis virus (JEV), (DENV), and Yellow fever virus (YFV), and some are spread by ticks, which mainly cause Alkhurma diseases, Tick-Borne Encephalitis (TBE), Omsk hemorrhagic fever, and Kyasanur forest disease ( 26 ). These viruses can cause a wide variety of illnesses in vertebrates including asymptomatic to moderate fever, flu-like symptoms to fatal encephalitis, arthralgia to severe hemorrhagic fever. Despite genetic and anatomical similarities within the genus or species, infection can have a wide range of clinical outcomes ( 27 ). Infection with WNV, JEV, TBEV, Powassan virus (POWV), and ZIKV can result in neuroinvasive infection, whereas DENV, YFV, and ZIKV cause visceral illness ( 28 ). Low blood pressure can result from the NS1 protein's capacity to change vascular permeability. Depending on the flavivirus, lesions may form in the brainstem, cerebral cortex, hippocampus, thalamus, cerebellum, or spinal cord ( 27 ). Acute flaccid paralysis (AFP), which is a distinct manifestation of WNV infection, and neuromuscular weakness is seen in 50% cases. Seizures and dystonia are common in JEV patients, and 50% of survivors continue to experience long-term psychological effects ( 28 ). Thus, fatality may linger in patients for a very long time and possesses a risk of long-term morbidity. However, there is no antiviral treatment for flavivirus infection, with the exception of certain vaccines against YFV, JEV, and TBEV, and their outbreaks persist until the weather eliminates their vectors ( 29 ). The whole flavivirus genome is approximately an 11-kb RNA with a single ORF that is bordered by the 3' and 5' non-coding sequences. This ORF translates 7 non-structural proteins and 3 structural proteins (capsid, membrane and envelope) which are necessary for viral generation. The RNA genome is stored in a structural core of the underlying capsid (C) protein, which is covered on the surface of the mature virion by envelope (E) glycoprotein and membrane (M) protein ( 30 ). The envelope protein has to interact with the proper receptor for the flavivirus genome to enter the host cell. The interaction between the envelope (E) protein and the glycosaminoglycans on the surface of the host cell gradually raises the viral concentration over the surface, ensuring tight contact with the receptor ( 31 ). Three ecto-domains of the envelope protein, termed as domain I, II, and III, engage with receptors and attachment factors to initiate endocytosis. A conserved fusion loop of domain II, located at the end of envelope proteins, is revealed to be particularly important for starting receptor-mediated endocytosis. In addition, the domain III has also been mapped by the vast majority of strong, neutralising antibodies, underscoring its clinical significance. Evidently, the anti-domain I and anti-domain II antibodies are less effective, nonetheless have greater cross-reactivity, and comprise a large fraction of anti-E IgG antibodies that are now being investigated for therapeutic purposes ( 32 ). The virus that causes dengue fever is carried mostly by female mosquitoes Aedes aegypti (Linnaeus, 1762), and to a minor degree by Aedes albopictus (Skuse, 1895). The mature form of the virus has a spherical shape with a diameter of 40 to 50 nm, and a single-stranded, positive-polar RNA genome is encased in a bilayer membrane ( 33 ). DENV has four distinct serotypes in the Flaviviridae family, DENV1 through DENV4, and a large number of genotypes within each serotype. Infection by any of those can cause a variety of well-defined medical conditions, from mild fever to alarming dengue, which is sometimes associated with a number of fatal complications like plasma leakage, respiratory distress, fluid accumulation, severe bleeding, and organ impairment ( 34 ). The serotype and genotype variation may affect the clinical condition differently. For instance, compared to the American DENV-2 genotypes, Asian genotypes tend to produce more serious infections in people ( 5 ). According to a research work conducted in Thailand, DENV-2 was most likely to cause DHF (44%) than other serotypes and DENV-4 was less likely (31%) ( 35 ). Additionally, the clinical severity varies depending on the interactions between the virus and the host's immune system. Children who have already acquired immunity to the DENV-1 serotype are susceptible to SD from DENV-2NI-1, whereas DENV-2NI-2B is more virulent in children who have already infected with the DENV-3 serotype ( 36 ). Various phylogenetic studies have reported the presence of recombinant genotype within particular DENV serotype ( 37 , 38 ). This intra-serotype recombination along with high mutation rate of this RNA virus is responsible for antigenic variation ( 39 ). However, the effect of recombination is weaker in creating DENV diversity, and is not driven by natural selection to enhance the fitness of the virus ( 40 ). Intra-serotype recombination is a chance phenomenon and cannot solely drive the evolution of a new genotype; however, it may aid the evolution process of DENV towards better viral fitness. Nonetheless, nucleotide variation analysis showed that the occurrence of synonymous mutations was more than non-synonymous mutations in DENV1-4, and mainly found in the non-structural genes which are responsible for viral replication ( 41 ). The evolutionary studies provided insight into the processes of positive selection and antigenic diversity in structural and non-structural genes. Therefore, most of the antiviral efforts are concentrating on the conserved areas to avoid diversity. The DENV is generally spherical in shape, and is made up mostly of a 10.7 kb RNA, capsid proteins, and envelope proteins. It enters into the host cells by interacting with cell surface receptors followed by clathrin-mediated endocytosis and trafficking into the late endosomes. The low pH condition of the late endosome triggers structural rearrangements of the DENV E protein and promotes fusion of viral and host membranes ( 33 ). Once viral genome is released into host cell cytoplasm, viral RNA is translated into a single polyprotein that is subsequently cleaved into three structural and seven non-structural proteins (NS1, NS2A, NS2B, NS3, NS4A, NS4B, and NS5) by viral (NS3) and host proteases ( Figure 1 ). The capsid (C), membrane (M), and envelope (E) proteins give the structural shape, and the non-structural proteins are engaged in virus multiplication ( 23 ). These non-structural proteins show different levels of sequence variability among different serotypes. NS2 proteins show the highest level of variability among all the non-structural proteins ( Figure 2 ). These non-structural proteins direct RNA replication and packaging. Negative-strand viral RNA is synthesized first that directs positive-strand RNA synthesis. Progeny RNA further associates with C proteins to form a nucleocapsid which buds into the ER to acquire lipid membrane containing heterodimers of E and prM proteins. The immature assembled virus is transported through Golgi apparatus where prM protein is cleaved by furin. After release into extracellular space, the cleaved pr peptide is dissociated and virions become fully infectious ( 42 , 43 ). Figure 1 Structural organization of dengue virus envelope protein. The RNA genome of dengue virus is composed of a single ORF, flanking by 5' and 3' untranslated region. The ORF includes coding sequence of three structural proteins, and seven non-structural proteins. The envelope (E) is the most significant structural protein of dengue virus surface. The monomer of each of the envelope protein is arranged into head-to-tail homo-dimer and 90 of such dimers are arrange in icosahedral symmetry to constitute the complete surface of dengue virus. Each of the monomeric envelope protein consists of three ecto-domains: EDI, EDII and EDIII. The domain I, II and III are denoting by red, golden and violet color respectively. Figure 2 Multiple sequence alignment of non-structural (NS) proteins of DENV serotypes 1 to 4. The amino acid sequences of (A) NS1 (B) NS2A and NS2B (C) NS3 (D) NS4A, 2K and NS4B (E) NS5 are taken from polyprotein having accession no. QTH36735.1, NP_056776.2, YP_001621843.1, NP_073286.1 for DENV serotypes 1 to 4 respectively. Fully conserved amino acids are indicated with black dots while variable amino acids are displayed in the Clustal Omega format using Jal view. As per the Global Similarity (BLOSUM62) analysis, the sequence similarity index for NS1, NS2, NS3, NS4 and NS5 proteins are 0.77 ± 0.03, 0.51 ± 0.04, 0.82 ± 0.03, 0.75 ± 0.02, and 0.79 ± 0.02, respectively. The complete RNA genome is enclosed in a nucleocapsid core, which is surrounded by a lipid bilayer membrane. The outer surface of the membrane also remains covered with the E protein. There are 180 copies of E proteins which are arranged into 90 head-to-tail homo-dimers, and three of such homo-dimers are arranged in a raft with two asymmetric units. These 30 units are finally organized in an icosahedral symmetry ( Figure 1 ) ( 44 ). There are three domains in the envelope protein, such as domain I, II, and III. Although domain I belongs at the N-terminal end, in the three-dimensional structure it is located in the middle, hanging domains I and II on their opposite sides. In addition, domain II has two extended loops, including a conserved fusion loop at one end of the E protein structure and a neighbouring bc loop. Whereas domain III, an immunoglobulin-like structure at the other end of the molecule, is thought to connect with the host cell surface receptor to induce endocytosis ( 45 ). As the virus enter in a host cell endosome, the acidic condition causes envelope proteins to undergo conformational changes, causing fusion loops to produce trimeric spikes that penetrate the endosomal membrane. This rearrangement of the trimeric envelope protein enables the release of the viral RNA into the host cell cytoplasm by merging the viral membrane with the endosomal membrane ( 46 ). 3 Antibodies targeting the DENV envelope protein The mAbs have certain unique advantages in dengue treatment as compared to the polyclonal antibodies. Neutralization escape mutations might be prevented by targeting numerous epitopes at a time to neutralize DENV ( 9 ). Nonetheless, weak or sub-neutralizing antibodies may induce ADE, a phenomenon most often associated with DENV infection ( 47 ). The major fraction of antibodies against DENV are not neutralizing, and the use of polyclonal antibody preparation carries a significant risk of ADE. Therefore, highly specific neutralizing mAbs are thought to be produced in vitro in a larger quantity, and the passive immunization with these antibodies can significantly reduce the DENV associated severity ( 9 ). The outer surface of DENV is made up of 180 copies of a monomeric envelope protein that is arranged with icosahedral symmetry ( 44 , 45 , 48 ) and is widely recognized as a target of neutralizing antibodies ( 49 ). Since the amino acid sequence of E proteins being 72% to 80% similar across all four serotypes, DENV neutralizing antibodies may either be type-specific or cross-reactive ( Figure 3 ) ( 50 ). Structural studies reveal that each E subunit consists of three distinct β-barrel domains termed as domain I, II, and III. Two well-characterized hydrophobic loops (fusion and bc) are present at domain II and a receptor-binding site at domain III ( 51 ). The primary repertoire of antibodies in human serum targets the fusion and bc loop of domain II, which are identified mostly as cross-reactive but poorly neutralizing ( 52 , 53 ). A minor repertoire of antibodies is found to target domain III, and they are highly neutralizing to some specific serotypes but not equally effective against all the four serotypes ( 51 ). Therefore, we have categorized the antibodies mainly into two groups: serotype-specific and cross-reactive; and also, have shown their target site specific critical response along with the ADE effects ( Table 1 ). Besides, a small number of mAbs have been identified for interacting with different structural residues of domains I, II, and III ( 79 , 85 ), some of which are found serotype-specific, and some are cross-reactive; therefore, they are further categorized as quaternary epitope-specific antibodies. Figure 3 Sequence alignment of Envelope proteins ecto-domains (60 to 400 residues) of four dengue serotypes. Fully conserved amino acids are indicated with black stars (*) at the bottom of each row. Domains I, II, and III are schematically indicated with red, yellow, and blue color bars. Secondary structures are shown above the amino acid cluster (arrows indicate β strands, and the lines indicate helices). Each of the β strands of domains I, II, and III have been denoted with (A 0 to H 0 ), (a to l), and (A to G), respectively. The interacting residues of serotype-specific and cross-neutralizing antibodies are marked with blue and yellow shaded boxes, respectively, and their corresponding antibodies are mentioned above the bars with the same color-shed. The antibodies which were identified without any ADE effects are shown with blue letter code name and mentioned above the sequence alignment. Table 1 DENV specific antibodies with their corresponding target epitopes, key interacting residues, neutralizing effects and ADE effects. Antibody Target Virus Epitope Key interacting residues Neutralizing effects ADE effects References Antibodies targeting serotype specific epitope DV1-E1 & E2 DENV1 DIII lateral ridge LR(Q347, N348 & D360) Neutralizes only DENV1 NA ( 54 ) E106 DENV1 DIII lateral ridge A(K307,K310);B(K325,Y326); BC(E327,T329,D330);K361,362 Neutralizes 5 DENV-1 genotypes NA ( 55 ) E111 DENV1 DIII & CC' loop N-terminal (300,301); C strand, CC' loop (E345), C' strand (334–351); EF loop (372);FG(382-384) Neutralizes DENV-1 genotype 2 NA ( 56 ) E105/E106 DENV1 DIII Lateral ridge BC(G328, T329 & D330), DE (K361 & E362), & FG (K385) Neutralizes 5 DENV-1 genotypes Not tested ( 57 ) 1A1D-2 DENV2 DIII Lateral ridge LR(K310) E383, P384 Strongly neutralizes DENV2 NA ( 58 ) 9D12 DENV2 Lateral ridge LR(K310) E383, P384 Moderately neutralizes DENV2 Not tested ( 58 ) 3F9 DENV2 DI centre DI-centre Neutralizes DENV2 serotypes NA ( 59 ) DV2-44, 76, 104 DENV2 DI & DII, DIII DI & II(K88, Q233, H244); DV2-76_DIII(V309); DV2-104 _DIII(P334, M340,H346) Neutralizes diverse DENV2 strain Prevent ADE ( 51 ) 5H2 DENV4 DI (F0, G0 and H0 and G0H0) DI(E172, K174, P176, D177, E180, R293) Neutralizes DENV4 NA ( 60 ) 1G6 DENV4 EDIII EDIII (387-ALTLH-390) T388, H390 Neutralize DENV4 Not mentioned ( 61 ) 3H2 DENV-2 EDIII DIII-DI hinge residues (298–303) BC loop (328–332), K361,V282, E383, P384 & C strand, CD loop Neutralizes DENV2 Marginal ADE ( 62 ) 2C8 DENV-2 EDIII DIII-DI hinge residues (298-303) BC loop (328-332), K361,V282, E383, P384 Neutralizes DENV2 Induce ADE ( 63 ) DENV-290 DENV-2 EDII EDII Neutralizes DENV3 Not mentioned ( 62 ) Antibodies targeting cross-reactive epitope 7F4 DENV-1,2,4 EDII DII (E67, E69 and E118) Cross-reactive but neutralizes DENV1 Undetected ( 46 ) E53 WNV, DENV1-4 EDII fusion loop WNV_FL(G104, C105, G106, L107, G109,K110); bc(C74, T76, M77, G78, E79) Cross-reactive Prevailing ADE ( 64 ) m366.6 DENV1-4 DIII A strand DIII_A(K310) Cross-neutralizing Undetectable level ( 65 ) 3E31 DENV1-4 DIII AB (314–370), E(365-370), Q316, H317, E368, E370 Cross-neutralizing Does not promote ( 66 ) Ab513 (4E5A) DENV1-4 DIII DE(358-365),Y360, S363 Broadly neutralizing Undetectable level ( 67 ) VIS513 (4E11) DENV1-4 DIII A strand DIII_A(K310, E311) Cross-neutralizing Minor level ( 68 ) 2H12 DENV1-4 DIII A strand A(314-317); FL(K110, E114, T115, Q116, H117) Cross-neutralizing except DENV2 Minor or negative ( 63 ) 9F12 WNV, DENV1-4 DIII A strand A(K305,K307, K310); B(327-331) Cross-neutralizing Not mentioned ( 69 ) N297Q-B3B9 DENV1-4 DII fusion loop Fusion loop Cross-neutralizing Lack of ADE ( 70 ) SigN-3C DENV1-4 DII fusion loop & DIII FL(G100, W101); EDIII(K310) Cross-neutralizing Abrogates ADE ( 71 ) 2A10G6 DENV1-4 DII fusion loop ZIKV, DENV1-4 FL(98D,99R,100X,101W) Cross-neutralizing Undetected ( 72 , 73 ) 1C19 DENV1-4 DII-bc loop DII-bc(R73,G78,E79) Cross-neutralizing Undetected ( 74 ) d448 DENV1-4 DII and M DII (D215, P219, M237, Q256, and G266) Cross-neutralizing Undetected ( 75 ) DM25-3 DENV1-4 VLPs DII-FL(W101) Cross-neutralizing Undetected ( 76 ) Z004 ZIKV DENV1 DIII lateral ridge LR(P384, G385) Neutralizes ZIKV and DENV1 Not tested ( 77 ) MZ4 ZIKV DENV2 DIII & DI linker ZIKV(residues 299-306) Neutralizes ZIKV and DENV2 Lesser extend ( 78 ) Antibodies targeting quaternary epitope HM14c10 DENV1 Hinge region between DI- DII & DIII DI(I161, T165, Q167,E172); DII(S273, G274,T275,T276); DIII(E311,E309,E327) Potentially neutralizes DENV1 Not mentioned ( 79 ) 5J7 DENV3 DI–DII HR & DIII, FL loop DI & DII (E123, E126, HR-Q269, N270) EDIII(K307, K308) Neutralizes DENV3 Not mentioned ( 80 ) 1F4 DENV1 DI & DI-DII HR of one ED1 and DIII of ED2 Doa; Eo, EoFo loop, Fo (155–165), Go strand, GoHo loop (170–177), and kl loop (272–276) Strongly neutralizes DENV1 Not mentioned ( 81 ) 2D22 DENV2 Dimer of DIII; and DII DIII(R323) & EDII (FL and bc loop) Neutralizes DENV2 Prevent ADE ( 82 ) 1L12 DENV2 DIII DIII(R323) Neutralizes DENV2 Not mentioned ( 59 ) A11, B7, C8, C10 DENV2 Dimer of EDI & EDII DI(148-159, N153 glycan); DII(67-74, N67 glycan), FL loop (97-106), ij loop (246-249) Neutralizes DENV1-4 Not mentioned ( 83 ) J8, J9 DENV1-4 DI DI (K47,H149, V151, F279) Cross-neutralizing No ADE ( 84 ) NA, Not Available. 3.1 Antibodies targeting serotype-specific epitopes Several research groups are currently focusing on the therapeutic potential of type-specific antibodies, because they are examined as highly neutralizing, and provide persistent immunity against homologous DENV serotypes ( 9 , 58 ). Since DENV1 is the most dominant type of the four serotypes (DENV1-4) to cause infections, DENV1-specific antibodies have been prioritized for therapeutic studies. Two of such DENV1 specific mAbs (E1 and E2) have been reported which exhibit strong neutralizing potency, but have no cross-reactivity to other serotypes. Further, structure-guided epitope mapping reveals that these strong inhibitory antibodies bind with a unique epitope between Thr346 and Asp360 on the lateral ridge of DENV1's domain III ( Figure 4 ). Mutational study and sequence analysis identify that Gln347 and Glu360 residues are critical for inhibition of DENV1 infection, and these residues are highly conserved among most of the DENV1 genotypes, whereas diverse from DENV2-4 ( 54 ). Figure 4 Monoclonal antibodies targeting DENV E protein. Domains I, II, and III are schematically indicated with red, yellow, and blue colors, respectively. Antibodies in the light blue shed are serotype-specific that target mainly domain I, hinge between domain I and II, and lateral ridge of domain III. In contrast, the antibodies in the light-yellow color are cross-reactive and target mainly FL and bc loop of domain II and A strand of domain III. The antibodies which were identified without any ADE effects are shown with blue letter code name. Similarly, E105 and E106 antibodies bind potentially with all five DENV1 genotypes and exhibit strong neutralizing activity ( 57 ), however, their binding sites are distinct from E1 and E2 binding sites in domain III. Epitope mapping reveals that both of these antibodies lose their binding activity after the alteration of residues Gly328, Thr329, and Asp330 in the BC region, Lys385 in the FG loop, and the substitution of Lys361Glu and Glu362Lys in the DE region of domain III's lateral ridge area ( 57 ). Protein's crystal study divulges that the contact residues of the E106 binding site are mostly conserved among the five genotypes of DENV1 but vary with other serotypes. Therefore, mAbs, E105 and E106, neutralize all the DENV1 genotypes, nonetheless, it shows poorly detectable neutralizing potential against other serotypes ( 55 , 57 ). Several studies have shown that the domain III lateral ridge specific mAbs are not influenced by temperature or time-dependent structural changes, whereas the neutralization efficiency of domain III specific mAbs, E105 and E106, are found to be enhanced with the increment of temperature and incubation period ( 56 ). Another mAb has been reported which distinguishes between distinct DENV1 strains by neutralizing a cryptic epitope in domain III. The domain III of DENV1 has a non-conserved CC' loop that mAb E111 binds to and is anticipated to be inaccessible on both immature and mature virions. Nonetheless, it is capable of neutralizing some of the DENV1 genotypes ( 56 ). Therefore, it is assumed that the epitope recognition and virus neutralization by the E111 antibody may happen independently to the temperature, incubation time, and the epitope residues. In other words, the structural alteration of the virus surface influences the neutralization by binding antibodies with a more or less exposed epitope in a strain-dependent manner ( 86 ). Therefore, it is suggested that the sequence variation within CC' loop and their surrounding binding sites are responsible for such differential behavior of these antibodies with various DENV1 genotypes ( 56 ). The second most serotype-specific neutralizing mAbs have been identified against DENV2, which binds with complex structural epitopes of domain III that overlap between A strand and neighbouring lateral ridge residues. Two of such sub-complexes targeting antibodies are 9D12 and 1A1D-2; 1A1D-2 is a firmly, and 9D12 is a moderately neutralizing antibody ( 58 ). Their tertiary binding sites are highly conserved, and share residues of both the cross-neutralizing A strand epitope, and serotype-specific lateral ridge epitope. Further epitope mapping reveals that this kind of specificity is determined by the centre residues Glu383 and Pro384 in the FG loop, that are conserved among various DENV2 strains, but diverse in other serotypes ( 58 ). Whereas, the residues Lys305, Lys307, and Lys310 of A strand interact with the sub-complex specific mAbs, which share binding residues of cross-neutralizing mAb 4E11 ( 87 ). Among these three residues, the Lys310 is totally conserved in all the serotypes, and Lys305 and Lys307 are partially conserved, only between DENV4 and DENV1 strains ( Figure 3 ). Therefore, it suggests that their distinct binding specificities and neutralizing efficiency to DENV2 strains are due to the precise interactions with the specific lysine residues on the A strand and two centre residues (Glu383 and Pro384) in the FG loop. Furthermore, mAb 5H2 was identified from chimpanzee, can potentially bind to the DI of the DENV4 envelope protein, and can potentially neutralize it. Crystal-structure studies reveal that the residues of 5H2 Fab fragment form a hydrogen bond with the side chains of epitope residues, Glu172 and Lys174 of G 0 strand, Asp177 of G 0 H 0 loop, Glu180 of H 0 strand, and Arg293. Except for the Glu172, all the epitope residues that engage in hydrogen bonding are variable across all the DENV serotypes ( Figure 3 ). However, the 5H2's key binding residues are conserved among all of the sequences of the DENV4 strains available in the database ( 60 ). Further investigation reveals that the 5H2 Fab antibody binds with the linker residue Arg293 (between DI and DIII), thereby preventing the conformational change of fusogenic event. Locking these two subunits in the pre-hairpin trimer ultimately blocks the membrane fusion into the endosome ( 60 , 88 ). Another DENV4 envelope domain III specific neutralizing mAb, 1G6, plays role at one step before the fusogenic event. Pre- and post- adsorption inhibition experiment reveals that the 1G6 antibody prevents viral adsorption at the primary stage of life cycle ( 61 ). Since domain III engages in binding with many cell receptors ( 89 ), the binding of 1G6 antibody with domain III may block the host cell receptor binding as well as potential endocytosis. Serotype-specific mAbs have also been identified against other regions of DENV envelope protein such as the epitopes located especially at the centre of domain I, hinge region between domain I and II, dimer interface of domain I and II, fusion and bc loop of domain II, lateral ridge, CC' loop, and the A strand of domain III ( 57 , 58 ) ( Figure 4 ). For instance, mAb DV2-48 and DV2-51 bind primarily with the key residues Gly177 and Glu184, respectively of the domain I epitope, in the DENV2 envelope. These binding sites are similar to the binding site of DENV4 specific protective antibody 5H2, which targets key amino acid residue Lys174 in DI ( 61 ). There are three more antibodies, DV2-44, DV2-76, and DV2-104, which are also capable of neutralizing DENV2 strains without inducing ADE ( 51 ). Epitope mapping shows that DV2-44 interact with the residues Lys88, Asn233, or His244 in the dimer interface of domain I and II, whereas DV-76 loses binding due to the mutation of conserved residues Lys305, Lys307, and Val309 in the A strand of domain III; which are similarly targeted by the sub-complex specific antibodies 9D12 and1A1D-2 ( 51 , 58 ). Unlike the domain III lateral ridge-specific mAb targeting FG loop (especially the Glu383 and Pro384 residues), DV2-104 interacts with a conserved residue Pro336 located on the bridge between BC loop and C strand, and also two additional DENV2 specific residues Met340 and His346, resided respectively in the C strand and CC' loop ( 44 , 51 ). Due to the differential binding specificity, these antibodies so far provide protection exclusively against the DENV2 serotype. Generally, most serotype-specific DENV antibodies induce ADE at a sub-neutralizing concentration ( 90 ). Nevertheless, recently two unusual DENV2 type-specific mouse mAbs, 2C8 and 3H5, were characterized to domain III of the DENV2 envelope protein, which are potentially neutralizing; however, they significantly differ in their ADE properties. Analysis of their crystal structures revealed that 3H5 binds with some extra amino acids in the C strand and CD loop compare to 2C8. The crucial interacting residues Lys344 and Arg345 are situated in close proximity to the viral membrane and remain hidden in the inter-dimeric envelope structure. Due to this differential binding, 3H5 exerts strong neutralizing potency with low or negligible ADE, whereas 2C8 possesses a similar neutralizing capacity but promotes a higher level of ADE ( 62 ). The target epitope and the functional properties of 3H5 are symmetrical to the previously described antibody DV2-104 that also strengthen the explanation about their specificity and the reasons for their strong neutralizing potency without ADE. Lastly, 15 more human mAbs have been reported, some of which shows efficient neutralizing activity against DENV3, although their neutralization potency varies from 10 to 15-fold across different genotypes. These antibodies were categorized into three major subsets viz. group 1 mAbs, targeting domain I epitopes (DENV-437, -286, -298, -354, -404, -406, -443); group 2 mAbs, targeting domain II epitopes (DENV-115, -290 and-419); and group 3 mAbs, targeting domain III epitopes (DENV-66, -144). Group 1 mAbs poses both modest and robust neutralizing activity in vivo , as was the case with DENV-443 and -298 respectively. Whereas DII specific group 2 mAbs mostly reduce in vivo viral load, and among these, DENV-290 is the best neutralizing candidate ( 62 ). Therefore, it suggests that in addition to viral genotype, selection of an appropriate target site is also crucial for protective immunity. 3.2 Antibodies targeting cross-reactive epitopes Most of the DENV envelope protein-specific mAbs are serotype-specific, and targets mainly the domain III epitopes ( 57 , 58 ). However, a number of cross-reactive mAbs have been identified, which exclusively target the fusion and bc loop of domain II; and a small fraction of them are also found to target some of the conserved epitopes in domain III ( Figure 4 ). In general, domain II specific anti-E antibodies constitute the majority of cross-reactive anti-E antibodies identified from dengue patients. These antibodies have low neutralizing efficacy but are extremely cross-reactive with many serotypes, even comparatively more cross-reactive than domain III specific mAbs ( 91 , 92 ). A significant fraction of them is concentrated on the evolutionary conserved fusion (FL) and bc loop of domain II, and capable of neutralizing all the four DENV serotypes. Sometimes the fusion loop specific antibodies are bifunctional, having both neutralizing and enhancing activities. For instance, a fusion loop specific human mAb, D23-1B3B9 (B3B9) is strongly neutralizing against all of the DENV serotypes ( 70 , 93 ). However, it shows viral enhancing activity in FcγR bearing cells at a sub-neutralizing concentration that further limits its therapeutic trial. Despite the fact, this antibody's chimeric form (N297Q-B3B9 rIgG) has in vivo neutralizing effectiveness against all four serotypes without viral boosting activity ( 70 , 94 ). Another fusion loop specific mAb, 2A10G6, is cross-reactive with DENV1 to 4, ZIKV, JEV, TIEV, YFV, and WNV; and potentially neutralizes all DENV strains, ZIKV, YFV and WNV. Structural and functional studies reveal that 2A10G6 bind with a conserved motif 98DRXW101 of the domain II fusion loop and confers protection against the lethal doses with DENV1-4, WNV, and ZIKV viruses in the mouse model ( 72 , 73 ). Further functional tests show that this antibody prevents infection at a stage of the DENV life cycle after attachment. Recently, a potent cross-neutralizing antibody SIgN-3C has been identified, which targets a novel dimeric epitope consisting fusion loop of domain II and III. Alanine scanning and structural study reveal that SIgN-3C binds with a conserved dimeric cluster of DENV envelope protein comprising residues domain II (G100, W101) in one monomer and domain III (K310, R323) in another monomer. Importantly, the SIgN-3C and its LALA-variant (substitution mutations from leucine to alanine at positions 234 and 235) are cross-reactive and equally effective with in vitro and in vivo protection against all of the DENV serogroups, therefore it seems to escape the ADE danger ( 71 ). Although VIS513 and 1A1D-2 bind with a similar epitope in domain III, they are not equally potent in specificity and in vivo neutralization ( 68 , 71 ). Consequently, it implies that invariant residues of the fusion loop and domain III are crucial for maintaining the structural integrity of the virus and responsible for inducing cross-neutralizing antibodies. On the other hand, an anti-fusion loop antibody (E53), which predominantly concentrates on the immature form of the dengue and West Nile viruses. Compare to other anti-fusion loop antibodies, E53 neutralizes partly mature viruses but not completely mature viruses ( 95 ). X-ray crystallographic studies reveal that E53 interacts with twelve residues, comprising the fusion loop (Gly104, Cys105, Gly106, Leu107, Gly109, and Lys110) and the neighboring bc loop (Cys74, Pro75, Thr76, Met77, Gly78, and Glu79) residues of domain II. Further functional studies demonstrate that E53 inhibits furin-catalyzed prM processing in mild acidic pH and prevents the maturation of the immature virus particle ( 64 ). Despite preserving cross-reactivity among all DENV serotypes and WNV, E53 can still induce ADE and increase disease severity by rendering the immature or partially mature form of DENV particle. Therefore, to exempt ADE, an in silico designed scFv mutant variants were developed using Z-docking and molecular dynamic simulation ( 96 ). In vitro binding study reveals that some of these scFv variants possesses nano-molar (nM) range affinity towards the recombinant Fu-bc subunit protein ( 97 , 98 ). Another fusion loop specific mAb 'DM25-3' reacts more potentially with mature virus-like particles (mD2VLP) than immature virus-like particles (imD2VLP). Further, epitope mapping reveals that conserved Trp101 residue in the fusion loop is critical for its proper binding and neutralization. Hence, it anticipates that some residues, like Trp101 of the fusion loop, remain exposed at the mature stage, which facilitates its binding with mature form of virus particle. In contrast, E53 can only attach while virion particles go through a structural shift brought on by low pH at the "breathing" stage and the fusion loop gets fully exposed ( 76 , 95 ). Moreover, a human mAb, 1C19, recognizes the invariant region of bc loop, located at close-proximity of the domain II fusion loop, and exhibits ultra-high neutralizing potency for all the four DENV serotypes ( 74 ). Since some of the anti-fusion loop antibodies, such as E53 and 2J21, also bind with the adjacent bc loop residues, probably due to this reason, these antibodies may block the 1C19 binding site and prevent more potent bc loop specific immune response ( 58 , 74 ). Therefore, a high dose of 1C19 application might compete for binding with fusion loop specific antibodies and thereby suppressing fusion loop-mediated virus transmission, resulting in a higher level of neutralization potency. Alternatively, an scFv variant of IC19 antibody has been developed and synthetically engineered in the Aedes aegypti genome. The resulting homozygous mosquitoes are found to complete refractoriness to DENV infection, and thus it appears as a novel anti-dengue strategy ( 99 ). Despite targeting fusion and bc loop of DII, some of the cross-reactive antibodies target other conserved residues in DII. Among these, antibody d448 interacts with five crucial residues (Asp215, Pro219, Met237, Gln256, and Gly266) that are located in the hidden junction between the membrane and the ecto-domain of envelope proteins. As these residues occupy a crucial structural location, it is possible that binding of the D448 antibody to the important structural component of the DENV coat protein may interfere with the membrane-envelope interaction and prevent "breathing" during viral maturation ( 75 ). The exact role of antibodies in the prevention of recurrent DENV infections and increase the disease severity is still unknown. Most of the neutralizing mAbs against DENV1 to 4 show enhancing activities at sub-neutralizing doses ( 90 ). This bilateral phenomenon further discloses that each of the DENV antibodies has two distinct roles; one is neutralizing, and another is enhancing depending on its available concentration. Paradoxically, 3H12 only displays promoting activities even at higher concentrations whereas 7F4 displays neutralizing abilities at high IgG levels but no increasing activities even at lower values. Additionally, a IgG3 subclass of 7F4 antibody recognizes a novel epitope near the glycan moiety of 67 asparagine on the domain II of the envelope protein ( 100 ), which has a neutralizing potency of 10- to 1,000-fold stronger than any other previously reported humanized mAbs targeting either DI or DII. However, the IgG1 subclass of 3H12 exhibits only enhancing activities, which targets epitope other than DI and DII ( 76 ). Besides, mAb 15C12 primarily recognizes the A strand residue Glu309 in domain III, has both neutralizing and boosting capabilities ( 46 ). Therefore, it suggests that the neutralization potency of an anti-DENV mAb not only depends on its available serum concentration but also depends on the target epitopes and the antibody sub-types. A small number of domain III specific antibodies are also found to cross-reactive and capable of neutralizing all the four DENV serotypes with minor or no antibody-mediated virus enhancement. One of these cross-neutralizing antibodies, 3E31, identifies a thermo-sensitive, invariant epitope in domain III ( 101 ). X-ray crystallographic analysis reveals that it binds residues in the AB loop from positions 314 to 319, the β strand from positions 365 to 370, as well as residues at positions 321, 323, and 352. Moreover, its Fab format forms H-H bonds with the residues Gln368, His317, and Glu370, which are remarkably conserved from serotype DENV1 to 4 ( Figure 4 ). Reportedly, mAb 3E31 neutralizes all the dengue serotypes by inhibiting envelope-mediated membrane fusion, which sterically hinders the trimer formation during receptor-mediated endocytosis. Unlike the anti-fusion loop antibodies, 3E31 does not induce viral amplification even at the minimum-neutralizing level ( 66 ). Similarly, a mouse mAb, 2H12, can cross-react with all of the DENV serotypes and also neutralizes all except DENV2 with or without a minimal ADE. Co-crystallization study of 2H12 Fab antibody with recombinant domain III reveals that it shares the highly conserved epitope (residues 314-317) in AB loop with 3E31 binding site that usually resides buried in the mature form of virion; however, at the post-antibody binding state, it undergoes rearrangement of overall surface conformation. Furthermore, this binding is characterized as temperature-sensitive, and the neutralizing potency is significantly inadequate toward DENV2 at lower temperatures compared to the other serotypes ( 63 ). Moreover, another mouse antibody, 9F12, cross-reacts with all DENV serotypes even with WNV; and neutralizes all four serotypes in plaque reduction assays. Further structural and functional studies reveal that this antibody binds with nano-molar affinity to the conserved epitope of domain III that comprises residues of Lys305, Lys307, Lys310, and Gly330 present in the AB loop. Adsorption and fusion assays divulge that 9F12 neutralizes all DENV serotypes by inhibiting the early fusion steps; however, the post-exposure ADE responses are still unknown ( 69 ). Despite sharing a common antigenic part in the AB loop with 2H12 and 9F12, 3E31 is still unique as it does not promote ADE, and so far, it is due to its additional binding residues. On the other hand, two murine antibodies, 1A1D-2 and 4E11, bind with an overlapping epitope including A and G strands residues of domain III ( 102 ). Both of the antibodies bind and strongly neutralize all three DENV serotypes except serotype 4, although, 4E11 provides intermediate level protection against it. Structural study reveals that the difference in the neutralization scheme of these two mAbs is due to a shift in the epitope of domain III; the epitope shift is towards the A strand in the case of 1A1D-2, whereas the epitope is located at the center of A/G strands in the case of 4E11. Therefore, 4E11 antibody has been modified to increase its binding potency toward the domain III of DENV4 ( 103 ). One such engineered scFv variant is Ab513, which has increased affinity and neutralization potency at least 75-fold for the DENV4 compare to 4E5A, an early version of the engineered 4E11 antibody ( 104 ). In addition to DENV4, the resulting variant can also neutralize all three serotypes (DENV1-3) with EC50 values less than 200 ng/ml ( 67 ). Another 4E11-based engineered antibody, VIS513, targets the exposed region of domain III, overcoming antibody-enhanced infection ( 70 ). Structural study indicates that mutation at the conserved Glu311 with Asp311 may decrease VIS513 binding by disruption of hydrogen bond and van der Waals interactions. Preclinical research reveals that pre- and post-peak viremia treatment with VIS513 led to neutralize all the DENV infections, although RNAemia remains detectable at post-treatment in non-human primates (NHPs) ( 105 ). Recently, another mAb, m366.6, targeting domain III, has been isolated from a human naïve antibody library and is capable of neutralizing all four DENV serotypes in vitro , and protecting a DENV-infected mouse model without causing any detectable ADE effect. Computational docking study reveals that unlike serotype-specific antibody 1A1D-2, m366.6 interacts with key epitope residue K310 with CDR-L1 rather than CDR-H1, which explains why m366.6 is cross-reactive despite sharing a similar epitope with serotype-specific mAb 1A1D-2 ( Figure 4 ) ( 65 ). In addition to inter-serotype cross-neutralizing mAbs, a novel group of DENV/ZIKV cross-neutralizing antibodies has been identified. One such example is Z004, which targets the lateral ridge of domain III and potentially neutralizes both ZIKV and DENV but no other DENV serotypes ( 77 ). Similarly, MZ4 is also a potent ZIKV/DENV cross-neutralizing antibody which targets a novel epitope positioned at the linker between DIII and DI. It was found exceptionally protective against both of the deadly ZIKV and DENV2, for which IC 50 values are lesser than previously identified any other ZIKV and DENV2 specific mAbs ( 78 ). Since MZ4 shares the binding site (domain III and I linker) with previously described chimpanzee 5H2, it suggests that MZ4 may prevent the structural rearrangement during the fusogenic event and blocks membrane fusion as like 5H2 ( 88 , 102 ). 3.3 Antibodies targeting quaternary epitopes In addition to cross-reactive and serotype-specific antibodies, a novel class of highly neutralizing mAbs has been reported that recognize the viral particle rather than a particular epitope of the envelope protein ( 106 ). These antibodies are referred to as quaternary structure-specific because they specifically recognize a certain structural component of viral surfaces ( Figure 5 ) ( 85 , 107 ). For instance, 5J7 is a strongly neutralizing DENV3-specific antibody that detects a cryptic quaternary structure located at an inter-molecular junction of the dimeric E protein. This structural element includes the hinge region residues Gln269 and Asn270 of one envelope molecule A, the domain III lysine residues of positions 310 and 317 of another envelope molecule B, and the domain II tip residues Glu121 and Glu126 of adjacent dimeric molecule B' ( 80 , 108 ). Likewise, two of these Fab variants from 5J7 bind tightly to two dimeric unit of envelope proteins made up molecules A, B, A' and B' to impede infection by preventing membrane fusion at the post-attachment stage. Figure 5 Neutralizing antibodies targeting quaternary epitope of dengue E protein. Domains I, II, and III are schematically indicated with red, yellow, and blue colors, respectively. Both the top and the side view are displayed one 'head to tail' dimer of dengue E protein. Seven of the neutralizing antibodies are shown here schematically that target quaternary epitope in both top and side view of an E dimer. The antibodies which were identified without any ADE effects are shown with red letter code name. Several other reported antibodies, specially 1F4 and 14c10, also bind to the quaternary structure epitope of DENV1, determined by their binding ability only with the whole virus particle rather than with recombinant E protein. Both of these antibodies mainly bind with the domain I and the inter-domain hinge region between domain I and II ( Table 1 ). Although 1F4 exclusively binds to a single monomeric envelope protein, it is distinctive among quaternary structure-specific antibodies, while antibody 14c10 binds to two separate E proteins. Protein crystallization studies reveal that both of the antibodies interact with the β strand residues of F 0 and G 0 and the loop F 0 G 0 of domain I ( Figure 5 ). Since DENV4 specific 5H2 antibody shares binding site mostly with DENV1 specific mAbs 1F4 and 14c10, except the hinge region between domain I and II; hence it might be a critical site for serotype specificity ( 80 ). Further research demonstrates that 1F4 is very sensitive to the domain I and II hinge angle and only attaches to the viral particle, not the isolated envelope protein ( 81 ). Moreover, several human mAbs have been identified, which bind with some anti-parallel dimeric epitope or a highly ordered DENV surface quaternary structure. For example, 2D22, a highly neutralizing DENV2 specific mAb, interacts with eight adjacent residues, including R323 in the domain III of one monomer, and fusion and bc loop in the domain II of the adjacent anti-parallel monomers. Although the contact residues of 2D22 in domain III are not conserved among all DENV serotypes, but the residues in domain II are highly conserved between DENV2 and DENV4 ( 107 ). These bilateral characteristics of the quaternary epitopes so far determine both the serotype specificity and the neutralization potency ( 82 ). Similarly, another DENV2 specific neutralizing mAb 1L12 binds with highly ordered quaternary structure epitopes on the DENV surface E homodimers. They may detect adjacent or overlapped epitopes on the viral surface because 2D22 interfere 1L12 from binding to DENV2 ( 59 ). In contrast, four E-dimer specific mAbs (A11, B7, C10, and C8) have been identified, which are not serotype-specific, however, can potentially neutralize all the four DENV serotypes ( Figure 5 ). The contact sites of all these antibodies are aligned in a valley bordered by β strand on domain II and β strand E 0 F 0 of domain I. In domain II, both of the E-dimer engages in binding with the same residues of the β strand (67-74 with Asn67 glycan), fusion loop, immediate upstream residues from position 97 to 106), and ij loop residues from 246 to 249 position, Figure 3 ). In domain I, the target residues within two E dimers are different; EDE2 interacts with residue N150 in the loop spanning from 148 to159 with Asn150 glycan chain, whereas EDE1 engages with domain I and III to displace Asn150 loop that facilitate antibody binding. Notably, the Asn67 and Asn153 glycans of the E protein, which are conserved throughout the four DENV serotypes, interact extensively with these antibodies. Probably, due to this conserved interaction, the E-dimer-specific mAbs are cross-neutralizing ( 83 ). In addition, two broadly neutralizing antibodies, J8 and J9, are also recently reported to neutralize all the four serotypes in the pico-molar range. These antibodies majorly interact with some novel residues in domain I that are distinct from previously described epitopes. Since both of these antibodies predominantly attach to viral particles but not to soluble E protein, the specific location of the proper binding site remains unknown. Therefore, it suggests that the epitope might be quaternary in structure but it needs to be centered in the E monomer ( 84 ). 3.1 Antibodies targeting serotype-specific epitopes Several research groups are currently focusing on the therapeutic potential of type-specific antibodies, because they are examined as highly neutralizing, and provide persistent immunity against homologous DENV serotypes ( 9 , 58 ). Since DENV1 is the most dominant type of the four serotypes (DENV1-4) to cause infections, DENV1-specific antibodies have been prioritized for therapeutic studies. Two of such DENV1 specific mAbs (E1 and E2) have been reported which exhibit strong neutralizing potency, but have no cross-reactivity to other serotypes. Further, structure-guided epitope mapping reveals that these strong inhibitory antibodies bind with a unique epitope between Thr346 and Asp360 on the lateral ridge of DENV1's domain III ( Figure 4 ). Mutational study and sequence analysis identify that Gln347 and Glu360 residues are critical for inhibition of DENV1 infection, and these residues are highly conserved among most of the DENV1 genotypes, whereas diverse from DENV2-4 ( 54 ). Figure 4 Monoclonal antibodies targeting DENV E protein. Domains I, II, and III are schematically indicated with red, yellow, and blue colors, respectively. Antibodies in the light blue shed are serotype-specific that target mainly domain I, hinge between domain I and II, and lateral ridge of domain III. In contrast, the antibodies in the light-yellow color are cross-reactive and target mainly FL and bc loop of domain II and A strand of domain III. The antibodies which were identified without any ADE effects are shown with blue letter code name. Similarly, E105 and E106 antibodies bind potentially with all five DENV1 genotypes and exhibit strong neutralizing activity ( 57 ), however, their binding sites are distinct from E1 and E2 binding sites in domain III. Epitope mapping reveals that both of these antibodies lose their binding activity after the alteration of residues Gly328, Thr329, and Asp330 in the BC region, Lys385 in the FG loop, and the substitution of Lys361Glu and Glu362Lys in the DE region of domain III's lateral ridge area ( 57 ). Protein's crystal study divulges that the contact residues of the E106 binding site are mostly conserved among the five genotypes of DENV1 but vary with other serotypes. Therefore, mAbs, E105 and E106, neutralize all the DENV1 genotypes, nonetheless, it shows poorly detectable neutralizing potential against other serotypes ( 55 , 57 ). Several studies have shown that the domain III lateral ridge specific mAbs are not influenced by temperature or time-dependent structural changes, whereas the neutralization efficiency of domain III specific mAbs, E105 and E106, are found to be enhanced with the increment of temperature and incubation period ( 56 ). Another mAb has been reported which distinguishes between distinct DENV1 strains by neutralizing a cryptic epitope in domain III. The domain III of DENV1 has a non-conserved CC' loop that mAb E111 binds to and is anticipated to be inaccessible on both immature and mature virions. Nonetheless, it is capable of neutralizing some of the DENV1 genotypes ( 56 ). Therefore, it is assumed that the epitope recognition and virus neutralization by the E111 antibody may happen independently to the temperature, incubation time, and the epitope residues. In other words, the structural alteration of the virus surface influences the neutralization by binding antibodies with a more or less exposed epitope in a strain-dependent manner ( 86 ). Therefore, it is suggested that the sequence variation within CC' loop and their surrounding binding sites are responsible for such differential behavior of these antibodies with various DENV1 genotypes ( 56 ). The second most serotype-specific neutralizing mAbs have been identified against DENV2, which binds with complex structural epitopes of domain III that overlap between A strand and neighbouring lateral ridge residues. Two of such sub-complexes targeting antibodies are 9D12 and 1A1D-2; 1A1D-2 is a firmly, and 9D12 is a moderately neutralizing antibody ( 58 ). Their tertiary binding sites are highly conserved, and share residues of both the cross-neutralizing A strand epitope, and serotype-specific lateral ridge epitope. Further epitope mapping reveals that this kind of specificity is determined by the centre residues Glu383 and Pro384 in the FG loop, that are conserved among various DENV2 strains, but diverse in other serotypes ( 58 ). Whereas, the residues Lys305, Lys307, and Lys310 of A strand interact with the sub-complex specific mAbs, which share binding residues of cross-neutralizing mAb 4E11 ( 87 ). Among these three residues, the Lys310 is totally conserved in all the serotypes, and Lys305 and Lys307 are partially conserved, only between DENV4 and DENV1 strains ( Figure 3 ). Therefore, it suggests that their distinct binding specificities and neutralizing efficiency to DENV2 strains are due to the precise interactions with the specific lysine residues on the A strand and two centre residues (Glu383 and Pro384) in the FG loop. Furthermore, mAb 5H2 was identified from chimpanzee, can potentially bind to the DI of the DENV4 envelope protein, and can potentially neutralize it. Crystal-structure studies reveal that the residues of 5H2 Fab fragment form a hydrogen bond with the side chains of epitope residues, Glu172 and Lys174 of G 0 strand, Asp177 of G 0 H 0 loop, Glu180 of H 0 strand, and Arg293. Except for the Glu172, all the epitope residues that engage in hydrogen bonding are variable across all the DENV serotypes ( Figure 3 ). However, the 5H2's key binding residues are conserved among all of the sequences of the DENV4 strains available in the database ( 60 ). Further investigation reveals that the 5H2 Fab antibody binds with the linker residue Arg293 (between DI and DIII), thereby preventing the conformational change of fusogenic event. Locking these two subunits in the pre-hairpin trimer ultimately blocks the membrane fusion into the endosome ( 60 , 88 ). Another DENV4 envelope domain III specific neutralizing mAb, 1G6, plays role at one step before the fusogenic event. Pre- and post- adsorption inhibition experiment reveals that the 1G6 antibody prevents viral adsorption at the primary stage of life cycle ( 61 ). Since domain III engages in binding with many cell receptors ( 89 ), the binding of 1G6 antibody with domain III may block the host cell receptor binding as well as potential endocytosis. Serotype-specific mAbs have also been identified against other regions of DENV envelope protein such as the epitopes located especially at the centre of domain I, hinge region between domain I and II, dimer interface of domain I and II, fusion and bc loop of domain II, lateral ridge, CC' loop, and the A strand of domain III ( 57 , 58 ) ( Figure 4 ). For instance, mAb DV2-48 and DV2-51 bind primarily with the key residues Gly177 and Glu184, respectively of the domain I epitope, in the DENV2 envelope. These binding sites are similar to the binding site of DENV4 specific protective antibody 5H2, which targets key amino acid residue Lys174 in DI ( 61 ). There are three more antibodies, DV2-44, DV2-76, and DV2-104, which are also capable of neutralizing DENV2 strains without inducing ADE ( 51 ). Epitope mapping shows that DV2-44 interact with the residues Lys88, Asn233, or His244 in the dimer interface of domain I and II, whereas DV-76 loses binding due to the mutation of conserved residues Lys305, Lys307, and Val309 in the A strand of domain III; which are similarly targeted by the sub-complex specific antibodies 9D12 and1A1D-2 ( 51 , 58 ). Unlike the domain III lateral ridge-specific mAb targeting FG loop (especially the Glu383 and Pro384 residues), DV2-104 interacts with a conserved residue Pro336 located on the bridge between BC loop and C strand, and also two additional DENV2 specific residues Met340 and His346, resided respectively in the C strand and CC' loop ( 44 , 51 ). Due to the differential binding specificity, these antibodies so far provide protection exclusively against the DENV2 serotype. Generally, most serotype-specific DENV antibodies induce ADE at a sub-neutralizing concentration ( 90 ). Nevertheless, recently two unusual DENV2 type-specific mouse mAbs, 2C8 and 3H5, were characterized to domain III of the DENV2 envelope protein, which are potentially neutralizing; however, they significantly differ in their ADE properties. Analysis of their crystal structures revealed that 3H5 binds with some extra amino acids in the C strand and CD loop compare to 2C8. The crucial interacting residues Lys344 and Arg345 are situated in close proximity to the viral membrane and remain hidden in the inter-dimeric envelope structure. Due to this differential binding, 3H5 exerts strong neutralizing potency with low or negligible ADE, whereas 2C8 possesses a similar neutralizing capacity but promotes a higher level of ADE ( 62 ). The target epitope and the functional properties of 3H5 are symmetrical to the previously described antibody DV2-104 that also strengthen the explanation about their specificity and the reasons for their strong neutralizing potency without ADE. Lastly, 15 more human mAbs have been reported, some of which shows efficient neutralizing activity against DENV3, although their neutralization potency varies from 10 to 15-fold across different genotypes. These antibodies were categorized into three major subsets viz. group 1 mAbs, targeting domain I epitopes (DENV-437, -286, -298, -354, -404, -406, -443); group 2 mAbs, targeting domain II epitopes (DENV-115, -290 and-419); and group 3 mAbs, targeting domain III epitopes (DENV-66, -144). Group 1 mAbs poses both modest and robust neutralizing activity in vivo , as was the case with DENV-443 and -298 respectively. Whereas DII specific group 2 mAbs mostly reduce in vivo viral load, and among these, DENV-290 is the best neutralizing candidate ( 62 ). Therefore, it suggests that in addition to viral genotype, selection of an appropriate target site is also crucial for protective immunity. 3.2 Antibodies targeting cross-reactive epitopes Most of the DENV envelope protein-specific mAbs are serotype-specific, and targets mainly the domain III epitopes ( 57 , 58 ). However, a number of cross-reactive mAbs have been identified, which exclusively target the fusion and bc loop of domain II; and a small fraction of them are also found to target some of the conserved epitopes in domain III ( Figure 4 ). In general, domain II specific anti-E antibodies constitute the majority of cross-reactive anti-E antibodies identified from dengue patients. These antibodies have low neutralizing efficacy but are extremely cross-reactive with many serotypes, even comparatively more cross-reactive than domain III specific mAbs ( 91 , 92 ). A significant fraction of them is concentrated on the evolutionary conserved fusion (FL) and bc loop of domain II, and capable of neutralizing all the four DENV serotypes. Sometimes the fusion loop specific antibodies are bifunctional, having both neutralizing and enhancing activities. For instance, a fusion loop specific human mAb, D23-1B3B9 (B3B9) is strongly neutralizing against all of the DENV serotypes ( 70 , 93 ). However, it shows viral enhancing activity in FcγR bearing cells at a sub-neutralizing concentration that further limits its therapeutic trial. Despite the fact, this antibody's chimeric form (N297Q-B3B9 rIgG) has in vivo neutralizing effectiveness against all four serotypes without viral boosting activity ( 70 , 94 ). Another fusion loop specific mAb, 2A10G6, is cross-reactive with DENV1 to 4, ZIKV, JEV, TIEV, YFV, and WNV; and potentially neutralizes all DENV strains, ZIKV, YFV and WNV. Structural and functional studies reveal that 2A10G6 bind with a conserved motif 98DRXW101 of the domain II fusion loop and confers protection against the lethal doses with DENV1-4, WNV, and ZIKV viruses in the mouse model ( 72 , 73 ). Further functional tests show that this antibody prevents infection at a stage of the DENV life cycle after attachment. Recently, a potent cross-neutralizing antibody SIgN-3C has been identified, which targets a novel dimeric epitope consisting fusion loop of domain II and III. Alanine scanning and structural study reveal that SIgN-3C binds with a conserved dimeric cluster of DENV envelope protein comprising residues domain II (G100, W101) in one monomer and domain III (K310, R323) in another monomer. Importantly, the SIgN-3C and its LALA-variant (substitution mutations from leucine to alanine at positions 234 and 235) are cross-reactive and equally effective with in vitro and in vivo protection against all of the DENV serogroups, therefore it seems to escape the ADE danger ( 71 ). Although VIS513 and 1A1D-2 bind with a similar epitope in domain III, they are not equally potent in specificity and in vivo neutralization ( 68 , 71 ). Consequently, it implies that invariant residues of the fusion loop and domain III are crucial for maintaining the structural integrity of the virus and responsible for inducing cross-neutralizing antibodies. On the other hand, an anti-fusion loop antibody (E53), which predominantly concentrates on the immature form of the dengue and West Nile viruses. Compare to other anti-fusion loop antibodies, E53 neutralizes partly mature viruses but not completely mature viruses ( 95 ). X-ray crystallographic studies reveal that E53 interacts with twelve residues, comprising the fusion loop (Gly104, Cys105, Gly106, Leu107, Gly109, and Lys110) and the neighboring bc loop (Cys74, Pro75, Thr76, Met77, Gly78, and Glu79) residues of domain II. Further functional studies demonstrate that E53 inhibits furin-catalyzed prM processing in mild acidic pH and prevents the maturation of the immature virus particle ( 64 ). Despite preserving cross-reactivity among all DENV serotypes and WNV, E53 can still induce ADE and increase disease severity by rendering the immature or partially mature form of DENV particle. Therefore, to exempt ADE, an in silico designed scFv mutant variants were developed using Z-docking and molecular dynamic simulation ( 96 ). In vitro binding study reveals that some of these scFv variants possesses nano-molar (nM) range affinity towards the recombinant Fu-bc subunit protein ( 97 , 98 ). Another fusion loop specific mAb 'DM25-3' reacts more potentially with mature virus-like particles (mD2VLP) than immature virus-like particles (imD2VLP). Further, epitope mapping reveals that conserved Trp101 residue in the fusion loop is critical for its proper binding and neutralization. Hence, it anticipates that some residues, like Trp101 of the fusion loop, remain exposed at the mature stage, which facilitates its binding with mature form of virus particle. In contrast, E53 can only attach while virion particles go through a structural shift brought on by low pH at the "breathing" stage and the fusion loop gets fully exposed ( 76 , 95 ). Moreover, a human mAb, 1C19, recognizes the invariant region of bc loop, located at close-proximity of the domain II fusion loop, and exhibits ultra-high neutralizing potency for all the four DENV serotypes ( 74 ). Since some of the anti-fusion loop antibodies, such as E53 and 2J21, also bind with the adjacent bc loop residues, probably due to this reason, these antibodies may block the 1C19 binding site and prevent more potent bc loop specific immune response ( 58 , 74 ). Therefore, a high dose of 1C19 application might compete for binding with fusion loop specific antibodies and thereby suppressing fusion loop-mediated virus transmission, resulting in a higher level of neutralization potency. Alternatively, an scFv variant of IC19 antibody has been developed and synthetically engineered in the Aedes aegypti genome. The resulting homozygous mosquitoes are found to complete refractoriness to DENV infection, and thus it appears as a novel anti-dengue strategy ( 99 ). Despite targeting fusion and bc loop of DII, some of the cross-reactive antibodies target other conserved residues in DII. Among these, antibody d448 interacts with five crucial residues (Asp215, Pro219, Met237, Gln256, and Gly266) that are located in the hidden junction between the membrane and the ecto-domain of envelope proteins. As these residues occupy a crucial structural location, it is possible that binding of the D448 antibody to the important structural component of the DENV coat protein may interfere with the membrane-envelope interaction and prevent "breathing" during viral maturation ( 75 ). The exact role of antibodies in the prevention of recurrent DENV infections and increase the disease severity is still unknown. Most of the neutralizing mAbs against DENV1 to 4 show enhancing activities at sub-neutralizing doses ( 90 ). This bilateral phenomenon further discloses that each of the DENV antibodies has two distinct roles; one is neutralizing, and another is enhancing depending on its available concentration. Paradoxically, 3H12 only displays promoting activities even at higher concentrations whereas 7F4 displays neutralizing abilities at high IgG levels but no increasing activities even at lower values. Additionally, a IgG3 subclass of 7F4 antibody recognizes a novel epitope near the glycan moiety of 67 asparagine on the domain II of the envelope protein ( 100 ), which has a neutralizing potency of 10- to 1,000-fold stronger than any other previously reported humanized mAbs targeting either DI or DII. However, the IgG1 subclass of 3H12 exhibits only enhancing activities, which targets epitope other than DI and DII ( 76 ). Besides, mAb 15C12 primarily recognizes the A strand residue Glu309 in domain III, has both neutralizing and boosting capabilities ( 46 ). Therefore, it suggests that the neutralization potency of an anti-DENV mAb not only depends on its available serum concentration but also depends on the target epitopes and the antibody sub-types. A small number of domain III specific antibodies are also found to cross-reactive and capable of neutralizing all the four DENV serotypes with minor or no antibody-mediated virus enhancement. One of these cross-neutralizing antibodies, 3E31, identifies a thermo-sensitive, invariant epitope in domain III ( 101 ). X-ray crystallographic analysis reveals that it binds residues in the AB loop from positions 314 to 319, the β strand from positions 365 to 370, as well as residues at positions 321, 323, and 352. Moreover, its Fab format forms H-H bonds with the residues Gln368, His317, and Glu370, which are remarkably conserved from serotype DENV1 to 4 ( Figure 4 ). Reportedly, mAb 3E31 neutralizes all the dengue serotypes by inhibiting envelope-mediated membrane fusion, which sterically hinders the trimer formation during receptor-mediated endocytosis. Unlike the anti-fusion loop antibodies, 3E31 does not induce viral amplification even at the minimum-neutralizing level ( 66 ). Similarly, a mouse mAb, 2H12, can cross-react with all of the DENV serotypes and also neutralizes all except DENV2 with or without a minimal ADE. Co-crystallization study of 2H12 Fab antibody with recombinant domain III reveals that it shares the highly conserved epitope (residues 314-317) in AB loop with 3E31 binding site that usually resides buried in the mature form of virion; however, at the post-antibody binding state, it undergoes rearrangement of overall surface conformation. Furthermore, this binding is characterized as temperature-sensitive, and the neutralizing potency is significantly inadequate toward DENV2 at lower temperatures compared to the other serotypes ( 63 ). Moreover, another mouse antibody, 9F12, cross-reacts with all DENV serotypes even with WNV; and neutralizes all four serotypes in plaque reduction assays. Further structural and functional studies reveal that this antibody binds with nano-molar affinity to the conserved epitope of domain III that comprises residues of Lys305, Lys307, Lys310, and Gly330 present in the AB loop. Adsorption and fusion assays divulge that 9F12 neutralizes all DENV serotypes by inhibiting the early fusion steps; however, the post-exposure ADE responses are still unknown ( 69 ). Despite sharing a common antigenic part in the AB loop with 2H12 and 9F12, 3E31 is still unique as it does not promote ADE, and so far, it is due to its additional binding residues. On the other hand, two murine antibodies, 1A1D-2 and 4E11, bind with an overlapping epitope including A and G strands residues of domain III ( 102 ). Both of the antibodies bind and strongly neutralize all three DENV serotypes except serotype 4, although, 4E11 provides intermediate level protection against it. Structural study reveals that the difference in the neutralization scheme of these two mAbs is due to a shift in the epitope of domain III; the epitope shift is towards the A strand in the case of 1A1D-2, whereas the epitope is located at the center of A/G strands in the case of 4E11. Therefore, 4E11 antibody has been modified to increase its binding potency toward the domain III of DENV4 ( 103 ). One such engineered scFv variant is Ab513, which has increased affinity and neutralization potency at least 75-fold for the DENV4 compare to 4E5A, an early version of the engineered 4E11 antibody ( 104 ). In addition to DENV4, the resulting variant can also neutralize all three serotypes (DENV1-3) with EC50 values less than 200 ng/ml ( 67 ). Another 4E11-based engineered antibody, VIS513, targets the exposed region of domain III, overcoming antibody-enhanced infection ( 70 ). Structural study indicates that mutation at the conserved Glu311 with Asp311 may decrease VIS513 binding by disruption of hydrogen bond and van der Waals interactions. Preclinical research reveals that pre- and post-peak viremia treatment with VIS513 led to neutralize all the DENV infections, although RNAemia remains detectable at post-treatment in non-human primates (NHPs) ( 105 ). Recently, another mAb, m366.6, targeting domain III, has been isolated from a human naïve antibody library and is capable of neutralizing all four DENV serotypes in vitro , and protecting a DENV-infected mouse model without causing any detectable ADE effect. Computational docking study reveals that unlike serotype-specific antibody 1A1D-2, m366.6 interacts with key epitope residue K310 with CDR-L1 rather than CDR-H1, which explains why m366.6 is cross-reactive despite sharing a similar epitope with serotype-specific mAb 1A1D-2 ( Figure 4 ) ( 65 ). In addition to inter-serotype cross-neutralizing mAbs, a novel group of DENV/ZIKV cross-neutralizing antibodies has been identified. One such example is Z004, which targets the lateral ridge of domain III and potentially neutralizes both ZIKV and DENV but no other DENV serotypes ( 77 ). Similarly, MZ4 is also a potent ZIKV/DENV cross-neutralizing antibody which targets a novel epitope positioned at the linker between DIII and DI. It was found exceptionally protective against both of the deadly ZIKV and DENV2, for which IC 50 values are lesser than previously identified any other ZIKV and DENV2 specific mAbs ( 78 ). Since MZ4 shares the binding site (domain III and I linker) with previously described chimpanzee 5H2, it suggests that MZ4 may prevent the structural rearrangement during the fusogenic event and blocks membrane fusion as like 5H2 ( 88 , 102 ). 3.3 Antibodies targeting quaternary epitopes In addition to cross-reactive and serotype-specific antibodies, a novel class of highly neutralizing mAbs has been reported that recognize the viral particle rather than a particular epitope of the envelope protein ( 106 ). These antibodies are referred to as quaternary structure-specific because they specifically recognize a certain structural component of viral surfaces ( Figure 5 ) ( 85 , 107 ). For instance, 5J7 is a strongly neutralizing DENV3-specific antibody that detects a cryptic quaternary structure located at an inter-molecular junction of the dimeric E protein. This structural element includes the hinge region residues Gln269 and Asn270 of one envelope molecule A, the domain III lysine residues of positions 310 and 317 of another envelope molecule B, and the domain II tip residues Glu121 and Glu126 of adjacent dimeric molecule B' ( 80 , 108 ). Likewise, two of these Fab variants from 5J7 bind tightly to two dimeric unit of envelope proteins made up molecules A, B, A' and B' to impede infection by preventing membrane fusion at the post-attachment stage. Figure 5 Neutralizing antibodies targeting quaternary epitope of dengue E protein. Domains I, II, and III are schematically indicated with red, yellow, and blue colors, respectively. Both the top and the side view are displayed one 'head to tail' dimer of dengue E protein. Seven of the neutralizing antibodies are shown here schematically that target quaternary epitope in both top and side view of an E dimer. The antibodies which were identified without any ADE effects are shown with red letter code name. Several other reported antibodies, specially 1F4 and 14c10, also bind to the quaternary structure epitope of DENV1, determined by their binding ability only with the whole virus particle rather than with recombinant E protein. Both of these antibodies mainly bind with the domain I and the inter-domain hinge region between domain I and II ( Table 1 ). Although 1F4 exclusively binds to a single monomeric envelope protein, it is distinctive among quaternary structure-specific antibodies, while antibody 14c10 binds to two separate E proteins. Protein crystallization studies reveal that both of the antibodies interact with the β strand residues of F 0 and G 0 and the loop F 0 G 0 of domain I ( Figure 5 ). Since DENV4 specific 5H2 antibody shares binding site mostly with DENV1 specific mAbs 1F4 and 14c10, except the hinge region between domain I and II; hence it might be a critical site for serotype specificity ( 80 ). Further research demonstrates that 1F4 is very sensitive to the domain I and II hinge angle and only attaches to the viral particle, not the isolated envelope protein ( 81 ). Moreover, several human mAbs have been identified, which bind with some anti-parallel dimeric epitope or a highly ordered DENV surface quaternary structure. For example, 2D22, a highly neutralizing DENV2 specific mAb, interacts with eight adjacent residues, including R323 in the domain III of one monomer, and fusion and bc loop in the domain II of the adjacent anti-parallel monomers. Although the contact residues of 2D22 in domain III are not conserved among all DENV serotypes, but the residues in domain II are highly conserved between DENV2 and DENV4 ( 107 ). These bilateral characteristics of the quaternary epitopes so far determine both the serotype specificity and the neutralization potency ( 82 ). Similarly, another DENV2 specific neutralizing mAb 1L12 binds with highly ordered quaternary structure epitopes on the DENV surface E homodimers. They may detect adjacent or overlapped epitopes on the viral surface because 2D22 interfere 1L12 from binding to DENV2 ( 59 ). In contrast, four E-dimer specific mAbs (A11, B7, C10, and C8) have been identified, which are not serotype-specific, however, can potentially neutralize all the four DENV serotypes ( Figure 5 ). The contact sites of all these antibodies are aligned in a valley bordered by β strand on domain II and β strand E 0 F 0 of domain I. In domain II, both of the E-dimer engages in binding with the same residues of the β strand (67-74 with Asn67 glycan), fusion loop, immediate upstream residues from position 97 to 106), and ij loop residues from 246 to 249 position, Figure 3 ). In domain I, the target residues within two E dimers are different; EDE2 interacts with residue N150 in the loop spanning from 148 to159 with Asn150 glycan chain, whereas EDE1 engages with domain I and III to displace Asn150 loop that facilitate antibody binding. Notably, the Asn67 and Asn153 glycans of the E protein, which are conserved throughout the four DENV serotypes, interact extensively with these antibodies. Probably, due to this conserved interaction, the E-dimer-specific mAbs are cross-neutralizing ( 83 ). In addition, two broadly neutralizing antibodies, J8 and J9, are also recently reported to neutralize all the four serotypes in the pico-molar range. These antibodies majorly interact with some novel residues in domain I that are distinct from previously described epitopes. Since both of these antibodies predominantly attach to viral particles but not to soluble E protein, the specific location of the proper binding site remains unknown. Therefore, it suggests that the epitope might be quaternary in structure but it needs to be centered in the E monomer ( 84 ). 4 Pathogenesis and antibody-dependent enhancement of DENV After minor clinical issues, the majority of primary dengue patients normally recover, however, the infection confers lifetime protection against that serotype ( 7 ). Conversely, re-infection with an alternative serotype not only provides low protection but also leads to more severe diseases ( 8 ). Infection with DENV or any other flavivirus, especially ZIKV induces widely cross-reactive but weak or non-neutralizing antibodies, and these antibodies remain detectable for a long period of time ( 109 ). During secondary infection, antibodies produced at the initial flavivirus infection may cross-react with alternative serotypes, although they are not entirely neutralizing. These cross-reactive antibodies can thereby increase the uptake of heterologous viruses into the host immune cells, leading to an over-reactive immune response that results in plasma leakage and potentially fatal hypovolemic shock ( 110 ). The non-neutralizing antibodies produced during the initial infection may cross-react with the virus of subsequent infections. This virus-IgG immune complexes may then be internalized into myeloid cells through FcγR binding, and increased viral replication leading to the ADE ( Figure 6 ) ( 10 ). Figure 6 Pathogenesis and antibody-dependent dengue virus enhancing model. While antibodies of primary dengue virus infection bind to an infectious DENV particle during a subsequent infection with the homologous serotype, protective immunity is conferred. However, secondary dengue virus infection with alternative serotypes unable to completely neutralize the virus particle. Instead, the Ab–virus complex interacts to the Fc receptors (FcγR) on circulating monocytes, making it simpler for the virus to infect them, leads to increased viral replication and a higher risk of dengue with warning signs (DWS+)/severe dengue (SD). The ADE hypothesis states that the antibodies produced during primary infection are insufficient to prevent subsequent infections with alternative serotypes that only have 30–35 percent sequence variation. Instead, this low level of cross-reactive non-neutralizing antibodies can exacerbate the condition by accelerating FcγR-mediated endocytosis into the monocyte cells ( 111 ). The Fc receptor-mediated pathway of ADE was also found to suppress the expression of anti-inflammatory cytokines, including IL-12, IFN-γ, and innate anti-DENV mediators, nitric oxide radicals, allowing unrestricted DENV replication in the monocyte cell. Although the expression of the anti-inflammatory cytokines IL-6 and IL-10 was facilitated by this route, it is not clear yet how these cytokines contribute to DENV replication ( 112 ). A second DENV infection in children increases the risk of SD 10 times more than a first infection ( 113 ). Infants under one-year old who have DENV antibodies from their moms have a greater chance of developing SD than infants born to mothers who have never had the virus ( 114 ). Since ADE is an important cause of SD, it poses a fundamental challenge in the development of vaccines. The C-prM-E protein, which is the major component of the DENV vaccine, and the presence of prM antibodies in the host after immunization may cause ADE upon secondary infection with atypical dengue ( 115 ). In addition, the memory T cells, either type-specific or cross-reactive, are produced at initial infection. Contrarily, in repeated DENV infections, viral antigens are produced on the surfaces of infected cells, activating cross-reactive memory T cells, and causing the release of TNF-α and other pro-inflammatory cytokines that cause plasma leakage in the endothelium ( 116 ). One of the main characteristics of dengue with warning signs (DWS+) is plasma leakage, which is thought to be caused by vascular endothelial damage brought on by pro-inflammatory cytokines ( 117 ). While the precise cause of SD infection is uncertain, it is believed that re-infection with an alternate DENV serotype or even new homologous variants are major alarming factors for SD; this is most likely owing to the ADE ( 118 ). 5 Scopes of antibody engineering against DENV The American Food and Drug Administration (FDA) has licensed more than 79 mAbs for curing a wide variety of illnesses, ranging from cancer to chronic inflammatory syndrome and infection to other neurological disorders ( 119 ). Many mAbs have shown promising results against several infectious diseases as well as cancer in preclinical evaluation ( 20 , 120 ). Currently, more than 570 mAbs are under clinical development ( 121 ). Out of these, 38 mAbs are under active clinical process against several infectious diseases, including HIV, human respiratory syncytial virus (RSV), SARS, coronavirus, rabies virus, WNV, anthrax, influenza, hepatitis and Ebola ( 120 ). Even so, during the Covid-19 pandemic, the FDA has approved convalescent plasma treatment for seriously sick patients, which primarily entails possible antibody collection from recovered individuals ( 122 ). Although a variety of dengue mAbs that are either cross-neutralizing or serotype neutralizing have been proven, decreasing ADE is still a significant difficulty ( 123 ). Recently, protein-engineering technologies have been exploited to eliminate ADE and to improve therapeutic efficacy ( 124 ). Fc modifications have been commonly used to reduce interaction between Fc and FcγR, which is the most leading cause of ADE. Currently, several engineered antibody variants such as scFv, Fab, modified Fc conjugates, and bispecific constructs are under investigation against several infectious diseases. Substitution or deletion mutations in Fc regions or replacement with IgG4 Fc have been shown to reduce ADE ( 109 , 125 ). A single point mutation at position 297 (Asn297Gln) in a human mAb, B3B9, has resulted to cross-neutralize all variants of DENV without ADE enhancing effect ( 70 ). Similar to this, it is discovered that a widely neutralizing human IgG1 mAb, D23-1G7C2-IgG1 with a mutant Fc (Asn297Ala), impairs the binding to FcγR, which consistently has lower ADE than wild-type ( 24 ). Double substitutions from leucine to alanine (LALA) at positions 234 and 235 in the IgG Fc domain have also been demonstrated to abolish contact with FcγR ( 50 ). Current research shows that the MZ4 LALA version of the ZIKV antibody has strong neutralizing power against both DENV-2 and ZIKV without encouraging ADE ( 81 ). In contrast, there are two bispecific antibody platforms (DART and DVD-Ig), which have been utilized to prepare therapeutic antibodies against DENV ( 126 ). DART (dual-affinity retargeting molecule) is made up of two modified antibody fragments that are connected by short peptide linkers, and each of the chains has a cysteine residue at the C terminus to facilitate the creation of inter-Fv bonds ( 127 ). In comparison, DVD-Ig (dual variable domain immunoglobulin) consists of typical IgG with an additional VH/VL domain of another specificity linked by peptide linkers with compatible heavy or light chains. For example, DII-FL specific antibody E60 and A strand of domain III specific antibody 4E11 have been used to develop a bispecific and tetravalent Ig-DART molecule, which retains its in vitro neutralizing activity as well as in vivo therapeutic activity ( 84 ). Along with the resulting construct, an engineered Fc region with the beneficial point mutation (N297Q) could eliminate the ADE ( 70 , 128 ). Moreover, a novel bispecific DVD-Ig molecule 1A1D-2A10 has been created using two additional sets of well-characterized anti-DENV mAbs, 2A10G6 (2A10) and 1A1D-2 (1A1D). The 2A10 antibody binds to domain II to prevent the virus from fusing with the endosomal membrane, whereas the 1A1D antibody attaches to domain III to limit the viral attachment to the host cell surface ( 129 ). Further, an engineered Fc domain with nine residues deletion from the N terminus (position 231-239) abrogates binding with FcγR ( 109 ). The resultant design (DVD-1A1D-2A10) is able to neutralize all four DENV serotypes without causing any ADE since it targets both the attachment and fusion phases ( 129 ). Therefore, it suggests that these bispecific molecules have great potential against DENV specific antiviral designing. Finally, some of the Fc modifications are in charge of increasing the serum half-life of mAbs. Fc mutation at His310 and His415 may change the interaction capability to its salvage receptor, FcRn, and prevent lysosomal degradation by redirecting the antibodies for recycle back into the blood stream ( 130 ). While mutation at position 250 (T250Q) and 428 (M428L) increases binding affinity with FcRn and enhance the serum half-life by at least two-fold in rhesus monkey without affecting ADCC and CDC ( 131 , 132 ). Besides, some of the Fc modifications can also alter the binding of ADE inducing complement component C1q. Mutations at positions 326 (K326W) and 333 (E333S) in the epicenter of the C1q binding site are observed at least five-fold improvement in terms of binding and CDC effects, without influencing ADCC ( 133 ). Therefore, combining all of these beneficial mutations in the Fc region may improve the effector functions of DENV mAbs and simultaneously may reduce the risk of ADE. The process of producing antibodies involves the selection of B cells, their proliferation, and differentiation into plasma cells. The process of affinity maturation, takes place primarily in antigen-selected germinal center B cells through somatic hypermutation (SHM) ( 134 ). A study determined immunoglobulin heavy chain variable frequency usage and SHM levels using high-throughput sequencing of peripheral blood IgG antibody repertoires from DENV patients. The study reported overall low proportion of somatic hypermutated antibody genes during the acute phase plasmablasts, particularly in secondary infections and those cases with more severe disease ( 135 ). So, lower SHM will result in antibodies with suboptimal avidity and affinity towards the epitopes of the secondary-infecting virus thus favouring ADE. The generation of skewed antibody response towards the memory of primary infecting serotype and lower affinity response for secondary infecting serotype due to lower SHM is attributed to original antigenic sin ( 14 ). 6 Conclusion and future directions The antibodies targeting serotype-specific epitopes are mostly neutralizing, however, none of them are completely free from the risk of ADE. A few of the cross-reactive antibodies such as Sign3c, 2H12, and VIS513, either targeting the fusion loop of domain II or A strand of domain III, are highly neutralizing, and characterized with minor or no enhancing activities. Some of the quaternary structure specific antibodies including m366.3 and 3E31 are also found cross-neutralizing without any ADE effects. The engineered version of fusion loop specific B3B9 antibody (N293Q-B3B9 rIgG) is also found to be protective against all of the four serotypes without viral enhancing activities. The epitope-paratope binding analyses of this antibody could provide valuable insights which would be helpful in designing of the antigen for the effective vaccine development. Moreover, this information along with the sequence alignment suggest that the fusion loop of domain II and A strand of domain III are highly conserved and generate cross-reactive antibodies. The antibodies generated against these two epitopes are highly neutralizing and mostly they have potential to escape viral enhancing activities. The detailed description of the most potent and cross-reactive antibodies along with the information of ADE enhancing or reducing epitopes may be beneficial for the discovery of next-generation antibody therapeutics. On the other hand, with a detailed understanding of the binding sites, the potential immune responses, and associated risk analysis may be useful for designing subunit vaccines, possibly in the way of heterologous prime-boosting strategies that can be administrated by following the sequential responses of natural infections. In addition, antibody engineering, especially Fc modifications are currently being utilized to improve therapeutic efficacy either by reducing the risk of ADE or by increasing half-life and effector functions. Although cross-reactive antibodies may increase viremia at sub-neutralizing concentrations and serotype-specific antibodies prevent viral fusion with lower risk of ADE, these dual phenomena can be utilized for therapeutic application by controlling its applications as well as subsequent doses. Finally, antibody cocktail targeting the potential epitopes may also be useful to reduce the chances of neutralizing escape variant, which may be protective against diverse DENV strains regardless of ADE. Author contributions AS and RG contributed to conception and design of the study. AS wrote the first draft of the manuscript. ND contributed by writing a part of the manuscript. RG, AS, and ND revised the manuscript. All authors contributed to the article and approved the submitted version. Conflict of interest The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Publisher's note All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7119210/

Mesenchymal Stem Cell-Derived Exosomes: A Promising Therapeutic Ace Card to Address Autoimmune Diseases

With the development of novel treatments for autoimmune disorders, it has become a popular research focus which mesenchymal stem cells (MSCs) have the capacity to counteract with autoimmune diseases progression. One of the underlying mechanisms behind their activities is the release of extracellular vesicles especially exosomes. MSC-derived exosomes are hypoimmunogenic nanocarriers which contain numerous immunoregulatory factors and similar to other exosomes, are able to pass through boundaries like the blood-brain barrier (BBB). Accumulating evidence provided by animal studies has demonstrated that MSC-derived exosomes, as a novel therapy, can re-induce self-tolerance, without subsequent complications reported for other treatments. Therefore, therapeutic applications of MSC-derived exosomes are contributing to core advances in the field of autoimmune diseases. Here, we briefly describe the biological characteristics of MSC-derived exosomes and review the experimentally verified outcomes for autoimmune disease therapy purposes. Introduction Mesenchymal stem cells (MSCs) are multipotent, self-renewable, and spindle shape cells ( 1 ) which were firstly isolated from guinea-pig bone marrow and spleen by Friedenstein et al. in 1970 ( 2 ). Now, MSCs can be isolated from cord blood, umbilical cord, bone marrow, adipose tissue, and many other tissues ( 3 , 4 ). The single connecting feature of all these tissues is the presence of perivascular space around the blood vessels where pericytes come off and differentiate into MSCs after vessels got inflamed or injured. According to their common origin, pericytes and MSCs both are negative for CD34, CD45, CD56 and positively express CD29, CD44, CD73, CD90, CD105, CD146 ( 5 - 7 ). MSCs are functionally defined based on their capacity to produce molecules that provide a regenerative and immunoregulatory milieu. Moreover, these multipotent cells can differentiate into a closely related family of cells in vitro or following transplantation and subsequently recover the damages. However, recent findings have reported that MSCs do not significantly differentiate into other lineages in vivo under normal condition ( 8 ). Consequently in a new paradigm, Arnold Caplan coined "Medical Signaling Cell" for MSCs to precisely express that mesenchymal stem cells dock at the sites of injury and produce regenerative and immunoregulatory components ( 9 , 10 ). Afterwards, a great number of clinical trials have exploited MSC-based modalities in clinical conditions such as Type 1 diabetes mellitus, Rheumatoid Arthritis (RA), Multiple Sclerosis (MS), Crohn's disease, Lupus, kidney transplant, Graft versus Host Disease (GvHD) and many cancers (clinicaltrials.gov). Unlike other stem cells (embryonic stem cells as well as induced pluripotent stem cells), MSCs have a dynamic immunomodulatory profile which lies on their ability to sense the injury or inflammation and then switch to the required response ( 11 ). For instance, MSCs produce transforming growth factor beta (TGF- β ), indoleamine 2,3-dioxygenase (IDO), nitric oxide (NO), soluble HLA-G, and prostaglandin E2 (PGE2) in response to high concentration of interferon gamma (IFN γ ), tumor necrosis factor alpha (TNF- α ), and toll like receptor 3 (TLR3) agonists ( 11 - 13 ). This reaction activates CD4 + CD25 hi FoxP3 + regulatory T cells (Tregs) while inhibits natural killer cell and T cell responses ( 11 - 13 ). In contrast, low level of inflammatory agents and TLR4 agonists, decrease IDO, NO, PGE2 production and increase expression of pro-inflammatory cytokines (CXCL9, CXCL10, MIP-1 α , MIP-1 β , CCL5), indicating their double-edged sword function within the immune system ( 14 ). Accumulating investigations have demonstrated that these factors are mainly delivered through direct secretions of soluble molecules and release of extracellular vesicles particularly exosomes. Exosomes As shown in Table 1 ( 15 - 20 ), exosomes are one of the nanoscale cell-derived extracellular vesicles (EV) with 30∼100 nm diameter ( 15 ). Exosomes were first thought to be required for discarding unnecessary proteins out of the cells and termed as "garbage cans" ( 21 ). Later, they were introduced as gulps of cytosol delivering bioactive molecules especially proteins and small RNAs (like microRNAs) in intercellular communications. Recent investigation have shown that exosomes transfer bioactive molecules among cells and therefore is able to play an essential role in pathological conditions including Alzheimer's disease ( 22 ), transmissible spongiform encephalopathies ( 23 ), Parkinson's disease ( 24 ), and amyotrophic lateral sclerosis ( 25 ) or could be employed as nanoscale platforms for drug delivery. Constitutive shedding of exosomes is mostly reported for immune cells (i.e. MSCs, lymphocytes, mast cells and dendritic cells) and non-immune cells particularly tumor cells ( 26 , 27 ). Additionally, interaction between T lymphocytes and antigen presenting cells strongly induce the release of the exosomes bearing TCR/CD3/zeta complex ( 28 ) and MHC II-peptide ( 29 ), respectively. It was lately demonstrated that in hypoxic preconditioning and presence of lipopolysaccharide, MSCs significantly enhance the release of exosomes which shuttle anti-inflammatory agents to polarize macrophages into the M2 phenotype ( 30 , 31 ). Generally, extracellular vesicles affect the neighboring cells via two different strategies: 1) fusion dependent 2) fusion independent. The first mechanism takes place when vesicle is taken up by a recipient cell. There are two types of fusions: 1) direct fusion of the extracellular vesicle with the plasma membrane of recipient cell that requires receptor-ligand interactions 2) Back fusion which confers to the endocytosis of vesicle by the recipient cell and subsequently incorporation of vesicle with endosome's membrane ( 32 , 33 ). In fusion independent manner, the vesicle and recipient cells interact with each other, like when exosomes bearing MHC-peptide are being specifically attached to the T cell receptor (TCR) on T cells ( 34 ). MSC-Derived Exosome Therapy: Cons and Pros As shown in Table 2 ( 35 - 46 ), recent findings depicted that MSC-derived exosomes deliver various cytoplasmatic constituents of the MSC secretomes, relevant to the stemness, angiogenesis, and particularly inflammatory factors. They were first used in 2010 for regeneration of tissues in a mouse model of myocardial ischemia/reperfusion injury ( 47 ) and are extensively on investigations at present. As a cell-free therapeutic tools provided a potential opportunity to unlock existing boundaries of cell therapy for autoimmune diseases. Exosomes introduce several advantages including capacity to cross the barriers (blood-brain barrier, capillaries) ( 48 ) and small enough to evade from being cleared by reticuloendothelial system (RES) ( 49 ). In addition to their general characteristics, MSC-derived exosomes are normally hypo-immunogenic due to the lack of MHC-II and low expression of MHC-I similar to their parental cells ( 50 ). It has also been revealed that MSCs can produce higher amount of exosomes compared with human cell lines like human embryonic kidney (HEK) and human acute monocytic leukemia (THP-1); Therefore, MSCs appear to be one of the best sources for high amount of immunoregulatory exosomes ( 51 , 52 ). Despite therapeutic advantages of MSC-derived exosomes, quantification of their protein profile and more importantly durability of cryopreserved exosomes as well as their in vivo tracking are still effortful. Regarding frozen exosomes, Cosenza et al. ( 53 ) analyzed functions of cryopreserved MSC-derived exosomes compared to freshly isolated exosomes and demonstrated that storage destabilize the integrity of membrane, leading to leakage of its content, aggregation of vesicles, and decrease in their immunosuppressive capacity. MSC-derived exosomes obviously express both the common MSC markers, such as CD44, CD90, and CD73 and markers of exosomes, CD9, CD63, and CD81 ( 51 ). These markers are used to track exosomes especially for drug delivery. Alternatively, cell tracing dyes can also help to chase exosomes in vivo . For instance, intravenous (i.v.) administration of Vybrant Cell Tracer DiD- or DiI-labeled exosomes to the mouse model of acute kidney injury, were used to monitor accumulation of MSC-derived exosomes. In this investigation, exosomes were mostly found in kidney, spleen, liver, and lung as filtering organs of the body ( 54 ). Apart from utilization of fluorescent molecules to chase organ localization of EVs, lower sensitivity of fluorophores negatively affect pharmacokinetic analysis of their applications (especially for quantification of exosome accumulated in a target tissue) ( 55 ). Radiolabeling with iodine-125 and bioluminescence emitted from luciferase provide a more practical solution for whole-body imaging of exosome-treated animals with significantly higher sensitivity of detection ( 56 , 57 ). Also, stability of B16-BL6 melanoma cell-derived exosomes in blood circulation were evaluated using exosomes expressing a fusion protein of Gaussia luciferase-Lactadherin, a reporter protein attached to one of the exosome's trophic protein, respectively. The results indicated exosomal half-life of about 2 minutes following i.v. injection. This finding suggests short half-life of exosomes is because of their fast uptake by liver and spleen resident macrophages ( 56 ). MSC-Derived Exosome Therapy for Autoimmune Diseases Utilization of the MSC-derived exosomes for autoimmune diseases therapy is at early stages. Most of studies are focused on immunomodulatory effects of these small vesicles in animal models. Here we will discuss the roles of these exosomes in the treatment of autoimmune diseases. Multiple Sclerosis Multiple Sclerosis (MS) is a chronic autoimmune disease characterized by neuroinflammation and demyelination in the central nervous system (CNS) ( 58 ). Numerous studies have revealed activation and recruitment of autoaggressive CD4 + and CD8 + T cells into the CNS as the main pathogenic players of MS ( 59 , 60 ). Disease-modifying treatments (DMTs) especially IFN- β are commonly prescribed to treat the affected individuals ( 61 , 62 ). Despite their beneficial impacts, DMTs increase the risk of adverse effects including lymphopenia, lymphadenopathy, dyspnea and lipoatrophy ( 63 - 65 ). Hence, identification and evaluation of alternative approaches with lower side effects are necessary. We have already demonstrated that MSCs and their activity have been altered after DMT consumption in relapsing-remitting MS patients ( 66 ). Indeed, patients own stem cells may have more deficiencies which not efficiently prevent for MS initiation and progression ( 67 ). However, MSCs can differentiate into neural cells and release neurotrophic molecules, indicating their strong neuroprotective effects for treating MS ( 68 ). Accumulating evidence has shown that MSCs can regulate autoreactive cells in a paracrine manner through secretion of inhibitory cytokines or molecules as well ( 69 - 71 ). MSC-derived exosomes can also induce peripheral tolerance toward autoreactive T cells via transferring the tolerogenic molecules to the autoreactive immune cells. Since exosomes are lipid-bound nano-vesicles, they can cross the blood-brain barrier. Therefore, incubation of exosomes in presence of neurotrophic and remyelination factors elevate their therapeutic profits. Additionally, expression of receptor(s) with ability to recognize neural markers in the MSCs increase efficiency of the delivery system. In this case, Zhuang et al. ( 72 ) have administered exosomes bearing anti-inflammatory compounds such as curcumin to the EAE mouse and demonstrated that clinical symptoms of the diseases were improved. In 2012, Mokarizadeh et al. ( 73 ) showed that MSC-derived extracellular vesicles exert immunomodulatory effects on splenic mononuclear cells (MNCs) of EAE mouse. These small vesicles not only reduced autoreactive lymphocytes proliferation, but also potentially induced MNCs to secrete significant amount of IL-10. They also demonstrated that MSC-derived vesicles contain PD-L1, galectin-1, and TGF- β , which help them with their immunoregulatory functions. This investigation proposed that MSC-derived vesicles including exosomes are capable of transferring anti-inflammatory molecules into autoreactive cells and re-inducing a self-tolerance. Later in 2018, Laso-García et al. ( 74 ) revealed that i.v. administration of MSC-derived exosomes to a progressive model of MS, Theiler's murine encephalomyelitis virus (TMEV)-induced demyelinating disease, recovered the disabilities observed before treatment. They indicated that administrated MVs prompt remyelination, reduce the brain atrophy, and infiltration of inflammatory CD4 + T cells to the spinal cord of TMEV-infected mice. Recently, Li et al. ( 75 ) demonstrated that MSC-derived exosomes polarize microglia cells mainly into the M2 phenotype and consequently alleviate clinical scores of EAE. In this study, the balance of M1/M2 in EAE rates significantly deviated toward M2 phenotype and their cytokines profile (IL-10 and TGF- β ), while frequency and activity of M1 cells were decreased. Interestingly for the first time, Hosseini Shamili et al. ( 76 ) bioconjugated the MSC-derived exosomes to an aptamer which targets oligodendrocyte markers. The armed exosome improved proliferation of oligodendroglia cell line (OLN93) in vitro and reduced both inflammatory responses and demyelinated lesions in the CNS of EAE mice. These methodologies may provide a new perspective toward indirect applications of MSCs in the treatment of MS. Rheumatoid Arthritis Rheumatoid arthritis (RA) is an inflammatory disease with unknown etiology that causes articular destruction and subsequently functional loss in joints. RA immunopathogenesis is characterized by disordered innate and adaptive immunity that result in immune complex-mediated complement activation, autoreactive lymphocyte response against self-antigens, and also dysregulated cytokine networks (77, 78). As with MS, there are various biological disease-modifying anti-rheumatic drugs (bDMARDs) available to control the symptoms of RA. For example, blockade of proinflammatory cytokines including GM-CSF, TNF- α , and IL-6 with antibodies or soluble decoy receptors have been mainly utilized for patients suffering from RA for many years ( 79 ). Along with beneficial effects of immunosuppressive and immunomodulatory drugs, they present various side effects including nausea, liver problems, lymphopenia, high risk of infection, and etc. ( 79 ). There is indeed a crucial need for new treatment modules with fewer side effects. MSC-derived exosomes have been considered to exert immunosuppressive activities, supporting their application in rheumatic diseases. Although the use of exosomes derived from various origins have been studied extensively to date ( 80 , 81 ), to date, the role of MSC-derived exosomes is still in the beginning. In 2018, Cosenza et al. ( 53 ) showed that MSC-derived exosomes reveal anti-inflammatory impacts on T and B lymphocytes in collagen-induced arthritis (CIA) mice model. In this study, MSCs-derived exosomes effectively inhibit T lymphocyte proliferation in a dose-dependent manner and decreased the percentage of mature T and B cell subsets. Interestingly, exosomes increased regulatory T cell populations much better than MSCs and MSC-derived microvesicles. Further investigations confirmed that microvesicles derived from MSCs are less effective to induce TGF- β and IL-10 production in T and B lymphocytes compared to MSCs alone ( 82 ). In 2018, Chen et al. ( 83 ) demonstrated that modified MSC-derived exosomes which contain miRNA-150, target the matrix metalloproteinase 14 (MMP14) and thereafter, decrease migration and invasion of fibroblast-like synoviocytes (FLS). They also indicated that injection of these engineered exosomes to the CIA mouse can effectively reduce the hind paw thickness and the clinical arthritic scores of animals. Another study evaluated beneficial impacts of human umbilical cord mesenchymal stem cells-derived exosomes (hUCMSC-derived exosomes) on bone destruction in CIA rats. Interestingly, the treatment reduced recruitment of inflammatory cell to the joints and prevent from joint synovial hyperplasia. The authors proposed that the underlying mechanism could be due to regulating of the imbalance of RANKL/OPG. To be clear, they showed that RANKL was decreased in serum and synovial tissues of the CIA rats, while the treatment raised the OPG concentration ( 84 ). In consensus with this study, another group demonstrated that intra-articular injection of hUCMSC-derived exosomes significantly decreased autoreactive infiltering chemokines CCL2 and CXCL12 in serum and synovial fluid ( 85 ). Type 1 Diabetes Mellitus (Insulin-Dependent Diabetes Mellitus) Type 1 diabetes mellitus (T1DM) is an incurable and chronic autoimmune disease which is classically identified by reduced level of insulin due to autoimmune destruction of insulin-producing pancreatic β cell, leading to hyperglycemia ( 86 ). Although underlying mechanism of disease is unknown, disruption of immune tolerance in autoreactive B cells, CD4 + T cells and particularly CD8 + cytotoxic T lymphocytes (CTL) are believed to contribute to the pathogenesis of type 1 diabetes. As a result, diabetic patients need exogenous insulin regiment to control their blood glucose level and avoid from consequences caused by hyperglycemia like fatigue, cataract, and also diabetic coma. Repetitive injection of insulin point to the fact that insulin titer is not stable in affected subjects and patients require a durable cure for diabetes. In order to develop novel therapies for T1DM, researchers have taken different approaches so far. Since first whole pancreas transplantation in 1966, long-term normoglycemia has been achieved with a graft survival rate of 50 to 70 % after 5 years ( 87 , 88 ). But sequential complications including pancreatitis, thrombosis, and pseudocyst appeared in graft are major disadvantages of this surgery ( 89 ). Eventually, transplantation of endocrine part of pancreas (known as islet transplantation) is a well-developed procedure to minimize complications of insulin therapy and pancreas transplantation ( 90 ). last two methods prevent from the disease progression for a limited period of time, but can not correctify autoimmune response against the β cells. The cure must be in the region where autoimmune response is originated. Various strategies have been applied to recover immune tolerance at critical points of regulatory pathways. Of the different approaches, anti-CD3 antibodies nonspecifically induce T cell anergy by co-stimulation blockade. Thereafter, beta cells are preserved in disease established non-obese diabetic (NOD) mice ( 91 ). Traditional α CD3 antibodies like OKT3 were provided from mouse and could not be administered continuously for chronic diseases. This drawback is overwhelmed by humanizing the antibody sequence ( 92 ). Moreover, mitogenic effect of CD3 specific antibodies on T cells, owning to their cross-linking ability to bind to the Fc receptors on natural killer cells and monocytes, brings about a notable cytokine release ( 93 ). This problem can be solved via engineering of C-terminal sequence of antibody to stop its binding possibility to Fc γ receptors. MSC therapy has also shown antidiabetic effects to restore immune response balance in pancreas. Ezquer et al. ( 94 ) injected GFP expressing MSCs to diabetic mouse and reported normalized Th1/Th2 response in diabetic group compared to control. They also traced labeled MSCs after 7 and 65 days to be sure if MSCs differentiate into beta cells and lead to normoglycemia. Surprisingly, they found MSCs in heart and secondary lymphoid organs (like Payer's patch, inguinal lymph node, and pancreatic lymph nodes) not in pancreas. It suggests that the antidiabetic effect of MSC is due to its endocrine activity not differentiation capacity. In 2016, Nakano et al. ( 95 ) have used MSC-derived exosomes to diminish cognitive impairments of diabetic rats by repairing damaged astrocytes and more importantly neurons. Furthermore, administration of MSC-derived exosomes simultaneously with islet transplantation, suppressed proliferation of PBMCs, improved regulatory T cell activities and increased graft survival in syngeneic mouse model of T1DM ( 96 ). Similar to previous studies, a group of scientists also evaluated immunomodulatory effects of MSC-derived exosomes compared with control and reported significant increase of regulatory T cell population among splenic MNCs, accompanied by upregulation of IL-4, IL-10, TGF- β plus downregulation of IL-17 and IFN γ cytokines after treatment. Accordingly treatment stabilized the blood glucose and vanished the signs of diabetes ( 97 ). Uveitis Uveitis is an autoimmune disorder which causes severe visual disability ( 98 ). Current clinical treatment includes corticosteroids, other immunosuppressive drugs, and newly developed biologics. The long-term utilization of each is followed by serious systemic side effects and the local risk of cataracts and glaucoma ( 99 , 100 ). Hence, most investigation intended to find localized and specific therapies. Although, it has been recently revealed that i.v. injection of MSCs improved experimental autoimmune uveitis (EAU) in animal model of uveitis ( 101 - 103 ), the underlying mechanisms of MSCs is not clear yet. In 2017, Bai et al. ( 104 ) showed that, human umbilical cord MSC-derived exosomes can reduce the severity of EAU. They demonstrated that MSC-derived exosomes inhibited the infiltration of inflammatory leukocytes to the eyes and these outcomes could, at least partly, be due to expression of CD73 on the MSCs. As expected, exosomes bearing CD73 interact with CD39 on activated immune cells and induce production of adenosine which result in proliferation inhibition. This study is the only study on the role of exosomes uveitis, indicating the importance of further investigation in this field. Conclusions Until last decade, the contribution of MSC-derived exosomes in the control of inflammatory responses has been less clear. As a functional vesicle derived from MSCs, exosomes have shown same beneficial impacts of the cell of origin, to suppress several autoreactive cells in autoimmune diseases. Moreover, although mesenchymal-derived exosomes expose immunomodulatory functions by self, it seems that engineering these vesicles through adding anti-inflammatory molecules and certain receptors, they can specifically target tissue/organ of interest. This attitude might consider as a potentially novel candidate to treat other autoimmune disorders like spondylitis ankylosing and scleroderma as well. With respect to remaining challenges, more investigations are required to elucidate unclear aspects of cell-free therapy using MSC-derived exosomes in ongoing studies.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4477118/

Wild chimpanzees are infected by Trypanosoma brucei

Although wild chimpanzees and other African great apes live in regions endemic for African sleeping sickness, very little is known about their trypanosome infections, mainly due to major difficulties in obtaining their blood samples. In present work, we established a diagnostic ITS1-based PCR assay that allows detection of the DNA of all four Trypanosoma brucei subspecies ( Trypanosoma brucei brucei , Trypanosoma brucei rhodesiense , Trypanosoma brucei gambiense , and Trypanosoma brucei evansi ) in feces of experimentally infected mice. Next, using this assay we revealed the presence of trypanosomes in the fecal samples of wild chimpanzees and this finding was further supported by results obtained using a set of primate tissue samples. Phylogenetic analysis of the ITS1 region showed that the majority of obtained sequences fell into the robust T. brucei group, providing strong evidence that these infections were caused by T. b. rhodesiense and/or T. b. gambiense . The optimized technique of trypanosome detection in feces will improve our knowledge about the epidemiology of trypanosomes in primates and possibly also other endangered mammals, from which blood and tissue samples cannot be obtained. Finally, we demonstrated that the mandrill serum was able to efficiently lyse T. b. brucei and T. b. rhodesiense , and to some extent T. b. gambiense , while the chimpanzee serum failed to lyse any of these subspecies. 1 Introduction African trypanosomes of the Trypanosoma brucei group are the causative agents of sleeping sickness and nagana ( Simarro et al., 2011 ). T. brucei has been found throughout sub-Saharan Africa and comprises three morphologically identical but genetically different subspecies, specifically Trypanosoma brucei brucei , Trypanosoma brucei rhodesiense , and Trypanosoma brucei gambiense ( Tait et al., 2011 ), and Trypanosoma brucei evansi ( Lai et al., 2008 ). Except the last one, these subspecies are transmitted by tsetse flies of the genus Glossina ( Franco et al., 2014 ), with T. b. gambiense and T. b. rhodesiense , causing the West-African and East-African human sleeping sickness, respectively, both being lethal, yet having distinct clinical syndromes ( Brun et al., 2010 ). T. b. brucei is limited to animals and non-infective to humans due to trypanolytic factors found in human serum, whereas other two subspecies responsible for human trypanosomiasis, T. b. gambiense and T. b. rhodesiense , have developed mechanisms for escaping from lysis mediated by the trypanosome lytic factor (TLF) ( Lugli et al., 2004; Wheeler, 2010 ), which is primarily composed of Apolipoprotein L1 (ApoL1) and a haptoglobin-related protein ( Raper and Friedman, 2013 ). Both trypanosomes able to infect humans exhibit different ways of developing the TLF-mediated lysis. In the case of T . b . gambiense , the resistance is achieved by the interplay among a unique modification of the TLF receptor, expression of a specific TgSGP glycoprotein, and changes in lysosomal physiology ( Uzureau et al., 2013 ). T. b. rhodesiense developed the serum-resistance-associated protein, which is alone sufficient to confer complete resistance to the human TLF ( Xong et al., 1998; Vanhollebeke and Pays, 2010; Stephens et al., 2012 ). The sera of some non-human primates (baboons, sooty mangabeys, mandrills and gorillas) were shown to be capable of ApoL1-mediated killing of the flagellates, while the serum of chimpanzees showed no trypanolytic activity due to secondary loss of the ApoL1 gene ( Lugli et al., 2004; Poelvoorde et al., 2004; Thomson et al., 2009, 2014 ). Moreover, early studies showed that experimental infections with T. b. rhodesiense and T. b. brucei mostly caused the death of untreated chimpanzees, while infections with T. b. gambiense were mild and did not result in apparent clinical symptoms ( Baker, 1962, 1968; Baker and Taylor, 1971; Godfrey and Killick-Kendrick, 1967; Hoare, 1972 ). Within recent years, with the advent of genetic characterization, molecular markers have been developed to study the genetic diversity of trypanosomes ( Tait et al., 2011 ). These methods, such as determination of microsatellite markers for determination of allelic variations, demonstrated great diversity of the T. brucei group in humans due to frequent genetic exchange ( Capewell et al., 2013a; Duffy et al., 2014 ). Recently, Echodu et al. (2015) showed the emergence of the human infective strains from the non-infective T. b. brucei strains of different genetic backgrounds and highlighted the importance of cattle as possible reservoir of sleeping sickness. Moreover, Capewell et al. (2013a) revealed the existence of hybrids of T. b. gambiense and T. b. brucei with novel resistance mechanism, which are able to infect humans. These facts showed the importance of research on the trypanosomatids in other mammalian host, especially free-ranging non-human primates. Obtaining blood samples from wild African apes is complicated and in most cases ethically unacceptable ( Leendertz et al., 2006 ). As a result, the blood and tissue samples have been collected mostly from chimpanzees and other primates that died of anthrax or respiratory diseases in the Taï National Park (NP), Ivory Coast ( Leendertz et al., 2004; Kondgen et al., 2010 ) have been screened for the presence of Trypanosoma spp. Being aware of the breakthroughs in our understanding of the evolution of human Plasmodium species precipitated by their detection in feces of African apes ( Kaiser et al., 2010 ), we designed an assay capable of amplifying trypanosome DNA from this material. 2 Methods 2.1 Ethics statement General permission for sample collection from deceased wild primates was obtained from the authorities in charge of Côte d'Ivoire, Guinea, Uganda, and Democratic Republic of Congo. Deceased animals were found during the course of a long term project focused on the behavior and infectious diseases in wild primates ( Leendertz et al., 2006 ). No animal was anaesthetized or handled for the sole purpose of sample collection. Fecal samples were collected without disturbing the animals within 2–3 min after defecation. All samples from African sanctuary-living wild-born great apes were collected during routine health checks by the sanctuary on-site veterinarians. Since no animal was sampled specifically for this study, approval from the relevant institutional committee was not needed. All samples were collected according to the guidelines - Chimpanzee sanctuaries: guidelines and management workshop report (available at http://pages.ucsd.edu/∼jmoore/courses/methprimconsweb08/chimpsanct.pdf ; date of access to website in 2004). All sera from animals living in Czech zoological gardens were collected during preventive veterinary health checks. No animal was sampled specifically for this study. All samples were collected in accordance with legal requirements of the Czech Republic (Act no. 161/1992) and with the rules of the respective zoos. Samples collected during necropsies of primates that died from various causes in zoos and primate facilities were also included. Imports of samples from free-ranging chimpanzees proceeded according to German veterinary regulations for import of organic materials. Tissue and blood samples were exported with the appropriate CITES permissions from the respective country and Germany. No permit was needed for transport between laboratories in Germany and Czech Republic as both are members of European Union. The experimental infection of mice was approved by the Ethical Committees of the Czech Ministry of Education and Biology Centre (no. 90/2013). 2.2 Detection of trypanosomes in feces of experimentally infected mice Four laboratory mice were infected with T. b. brucei (STIB 920), T. b. rhodesiense (Etat 1.2 R variant), T. b. gambiense (LiTat 1.3) and T. b. evansi (STIB805) by intraperitoneal inoculation of 10 5 bloodstream stages (trypomastigotes), and the course of infection was controlled by searching for the flagellates in blood smears obtained as a drop from the tail. First search for the presence of parasites was performed 72 h post-infection and continued until trypomastigotes appeared in the blood every 8 h. At that time, fecal and blood samples were collected and used for isolation of total DNA using protocols described below. Blood samples were placed on a filter paper for ethanol evaporation and total DNA was isolated by incubation in a 200 μL volume with Chelex 100 (Sigma–Aldrich, St. Louis, MO, USA) (final concentration 5%) at 56 °C for 1 h, followed by boiling for 10 min. Before use, samples were centrifuged for 1 min at 15,000 rpm, and 2.5 μL of supernatant was used for PCR. DNA from fecal samples was isolated by a commercial stool kit (Qiagen, Venlo, Limburg, Netherlands). PCR and sequencing were performed for each sample at least in triplicate and, to avoid contamination, in a laboratory that does not work with trypanosomes, always with positive and negative controls included. A nested approach was used to amplify a region of the Internal Transcribed Spacer 1 (ITS-1) of Trypanosoma spp. (∼180-640 bp). Primers Tryp_3 (5′- TGCAATTATTGGTCGCGC -3′) and Tryp_4 (5′- CTTTGCTGCGTTCTT -3′) were used for the first round of PCR, while internal primers Tryp_1 (5′- AAGCCAAGTCATCCATCG -3′) and Tryp_2 (5′- TAGAGGAAGCAAAAG -3′) were used for the second round of PCR ( Adams et al., 2006 ). PCR was performed with Taq-polymerase (TopBio, Prague, Czech Republic) using the following program: 1 min at 95 °C, 35 cycles of 1 min at 94 °C, 1 min at 54 °C, 30 s at 72 °C, and 5 min at 72 °C, and PCR products were resolved in ethidium-bromide stained agarose gels. 2.3 Collection of primate samples Tissue samples were obtained from carcasses found mainly in Taï NP, Côte d'Ivoire. Furthermore, necropsy samples were collected from 93 wild primates from the Democratic Republic of Congo and Uganda. Additional 16 tissue samples were obtained from sanctuaries from Gambia, Guinea, Sierra Leone, Cameroon, Republic of Congo, Uganda, and two German zoos ( Table 1 ). Finally, 13 fecal samples were collected without disturbing the animals within 2–3 min after defecation and fixed in 96% ethanol. DNA extractions and PCR were performed in physically isolated facilities. DNA from tissue samples and feces was extracted using the DNAeasy tissue kits (Qiagen, Venlo, Limburg, Netherlands) and the EURx Gene Matrix Stool Kit (Roboklon), respectively. All amplicons were gel-purified using the Gel Extraction Kit (Qiagen, Venlo, Limburg, Netherlands) and cloned in TOPO TA Cloning (Invitrogen, Carlsbad, CA, USA) as specified by the manufacturer. Samples positive for ITS1 in nested PCR reactions were further analyzed for the presence of the ∼900 bp-long T. b. gambiense glycoprotein (TgSGP) gene, again via nested PCR, using outer primers 65-F (5′-GTGGCAATTACTAGCAATAGCG-3′) and 66-R (5′-GCCATCGTGCTTGCCGCTC-3′), and inner primers 61-F (5′-TCACGGCCATCAGACGGAGA-3′) and 62-R (5′-GGGCTCCTGCCTCAATTGCTGCA-3). Conditions were the same for both rounds: 2 min at 95 °C, 24 cycles of 1 min at 95 °C, 1 min at 55 °C, 2 min at 72 °C, and 2 min at 72 °C. As a positive control, DNA from blood of mouse experimentally infected with T. b. gambiense (LiTat 1.3) was used. 2.4 Phylogenetic analysis Highly polymorphic ITS1 sequences obtained by nested PCR from primate fecal and tissue samples were aligned with Clustal-X (ver. 2.0; gap opening penalty 12; gap extension penalty 5) and neighbor joining clustering with K2P distances was performed on the unmodified alignment using PAUP (4.0, beta version). The final alignment included 452 characters. The accession numbers of sequences retrieved from the GenBank ( Trypanosoma sp. ex Wildebeest JN673403 , for Trypanosoma theileri JX178185 , HQ664848 , and HQ664849 ) and used in phylogenetic reconstructions and ITS1 alignment are available from the authors upon request. 2.5 Trypanosome lysis in primate sera We used fresh sera from five non-human primate species (barbary macaque [ Macaca sylvanus ], mandrill [ Mandrillus sphinx ], spider monkey [ Ateles geoffroyi ], chimpanzee [ Pan troglodytes ], and Sumatran orangutan [ Pongo abelii ]) from the Plzeň, Liberec and Prague zoos, Czech Republic. At final concentration of 10% the sera were added to trypanosomatids cultivated in either HMI-9 or M199 medium with 10% fetal bovine serum (FBS), when concentration reached 5 × 10 5 cells/ml. The following isolates were used: T. b. brucei (STIB 920), T. b. rhodesiense (Etat 1.2 R variant), and T. b. gambiense (LiTat 1.3). Fresh normal human serum (NHS) obtained from a Central European individual and FBS were used as positive and negative controls, respectively. Trypanosomatids were cultivated in triplicate with the various primate sera for 24 h at 37 °C. After this period, resazurin (Sigma–Aldrich, St. Louis, MO, USA) was added to 1 ml of each culture (final concentration 58 pM) and cell viability was measured after 24 h (i.e. 48 hrs after the cells were mixed with primate sera) in triplicate using Tecan Infinite M200PRO under emission at 590 nm and excitation at 560 nm, and the results were analyzed using Anova. The relative survival was calculated as a ratio in a given serum compared to FBS. 2.1 Ethics statement General permission for sample collection from deceased wild primates was obtained from the authorities in charge of Côte d'Ivoire, Guinea, Uganda, and Democratic Republic of Congo. Deceased animals were found during the course of a long term project focused on the behavior and infectious diseases in wild primates ( Leendertz et al., 2006 ). No animal was anaesthetized or handled for the sole purpose of sample collection. Fecal samples were collected without disturbing the animals within 2–3 min after defecation. All samples from African sanctuary-living wild-born great apes were collected during routine health checks by the sanctuary on-site veterinarians. Since no animal was sampled specifically for this study, approval from the relevant institutional committee was not needed. All samples were collected according to the guidelines - Chimpanzee sanctuaries: guidelines and management workshop report (available at http://pages.ucsd.edu/∼jmoore/courses/methprimconsweb08/chimpsanct.pdf ; date of access to website in 2004). All sera from animals living in Czech zoological gardens were collected during preventive veterinary health checks. No animal was sampled specifically for this study. All samples were collected in accordance with legal requirements of the Czech Republic (Act no. 161/1992) and with the rules of the respective zoos. Samples collected during necropsies of primates that died from various causes in zoos and primate facilities were also included. Imports of samples from free-ranging chimpanzees proceeded according to German veterinary regulations for import of organic materials. Tissue and blood samples were exported with the appropriate CITES permissions from the respective country and Germany. No permit was needed for transport between laboratories in Germany and Czech Republic as both are members of European Union. The experimental infection of mice was approved by the Ethical Committees of the Czech Ministry of Education and Biology Centre (no. 90/2013). 2.2 Detection of trypanosomes in feces of experimentally infected mice Four laboratory mice were infected with T. b. brucei (STIB 920), T. b. rhodesiense (Etat 1.2 R variant), T. b. gambiense (LiTat 1.3) and T. b. evansi (STIB805) by intraperitoneal inoculation of 10 5 bloodstream stages (trypomastigotes), and the course of infection was controlled by searching for the flagellates in blood smears obtained as a drop from the tail. First search for the presence of parasites was performed 72 h post-infection and continued until trypomastigotes appeared in the blood every 8 h. At that time, fecal and blood samples were collected and used for isolation of total DNA using protocols described below. Blood samples were placed on a filter paper for ethanol evaporation and total DNA was isolated by incubation in a 200 μL volume with Chelex 100 (Sigma–Aldrich, St. Louis, MO, USA) (final concentration 5%) at 56 °C for 1 h, followed by boiling for 10 min. Before use, samples were centrifuged for 1 min at 15,000 rpm, and 2.5 μL of supernatant was used for PCR. DNA from fecal samples was isolated by a commercial stool kit (Qiagen, Venlo, Limburg, Netherlands). PCR and sequencing were performed for each sample at least in triplicate and, to avoid contamination, in a laboratory that does not work with trypanosomes, always with positive and negative controls included. A nested approach was used to amplify a region of the Internal Transcribed Spacer 1 (ITS-1) of Trypanosoma spp. (∼180-640 bp). Primers Tryp_3 (5′- TGCAATTATTGGTCGCGC -3′) and Tryp_4 (5′- CTTTGCTGCGTTCTT -3′) were used for the first round of PCR, while internal primers Tryp_1 (5′- AAGCCAAGTCATCCATCG -3′) and Tryp_2 (5′- TAGAGGAAGCAAAAG -3′) were used for the second round of PCR ( Adams et al., 2006 ). PCR was performed with Taq-polymerase (TopBio, Prague, Czech Republic) using the following program: 1 min at 95 °C, 35 cycles of 1 min at 94 °C, 1 min at 54 °C, 30 s at 72 °C, and 5 min at 72 °C, and PCR products were resolved in ethidium-bromide stained agarose gels. 2.3 Collection of primate samples Tissue samples were obtained from carcasses found mainly in Taï NP, Côte d'Ivoire. Furthermore, necropsy samples were collected from 93 wild primates from the Democratic Republic of Congo and Uganda. Additional 16 tissue samples were obtained from sanctuaries from Gambia, Guinea, Sierra Leone, Cameroon, Republic of Congo, Uganda, and two German zoos ( Table 1 ). Finally, 13 fecal samples were collected without disturbing the animals within 2–3 min after defecation and fixed in 96% ethanol. DNA extractions and PCR were performed in physically isolated facilities. DNA from tissue samples and feces was extracted using the DNAeasy tissue kits (Qiagen, Venlo, Limburg, Netherlands) and the EURx Gene Matrix Stool Kit (Roboklon), respectively. All amplicons were gel-purified using the Gel Extraction Kit (Qiagen, Venlo, Limburg, Netherlands) and cloned in TOPO TA Cloning (Invitrogen, Carlsbad, CA, USA) as specified by the manufacturer. Samples positive for ITS1 in nested PCR reactions were further analyzed for the presence of the ∼900 bp-long T. b. gambiense glycoprotein (TgSGP) gene, again via nested PCR, using outer primers 65-F (5′-GTGGCAATTACTAGCAATAGCG-3′) and 66-R (5′-GCCATCGTGCTTGCCGCTC-3′), and inner primers 61-F (5′-TCACGGCCATCAGACGGAGA-3′) and 62-R (5′-GGGCTCCTGCCTCAATTGCTGCA-3). Conditions were the same for both rounds: 2 min at 95 °C, 24 cycles of 1 min at 95 °C, 1 min at 55 °C, 2 min at 72 °C, and 2 min at 72 °C. As a positive control, DNA from blood of mouse experimentally infected with T. b. gambiense (LiTat 1.3) was used. 2.4 Phylogenetic analysis Highly polymorphic ITS1 sequences obtained by nested PCR from primate fecal and tissue samples were aligned with Clustal-X (ver. 2.0; gap opening penalty 12; gap extension penalty 5) and neighbor joining clustering with K2P distances was performed on the unmodified alignment using PAUP (4.0, beta version). The final alignment included 452 characters. The accession numbers of sequences retrieved from the GenBank ( Trypanosoma sp. ex Wildebeest JN673403 , for Trypanosoma theileri JX178185 , HQ664848 , and HQ664849 ) and used in phylogenetic reconstructions and ITS1 alignment are available from the authors upon request. 2.5 Trypanosome lysis in primate sera We used fresh sera from five non-human primate species (barbary macaque [ Macaca sylvanus ], mandrill [ Mandrillus sphinx ], spider monkey [ Ateles geoffroyi ], chimpanzee [ Pan troglodytes ], and Sumatran orangutan [ Pongo abelii ]) from the Plzeň, Liberec and Prague zoos, Czech Republic. At final concentration of 10% the sera were added to trypanosomatids cultivated in either HMI-9 or M199 medium with 10% fetal bovine serum (FBS), when concentration reached 5 × 10 5 cells/ml. The following isolates were used: T. b. brucei (STIB 920), T. b. rhodesiense (Etat 1.2 R variant), and T. b. gambiense (LiTat 1.3). Fresh normal human serum (NHS) obtained from a Central European individual and FBS were used as positive and negative controls, respectively. Trypanosomatids were cultivated in triplicate with the various primate sera for 24 h at 37 °C. After this period, resazurin (Sigma–Aldrich, St. Louis, MO, USA) was added to 1 ml of each culture (final concentration 58 pM) and cell viability was measured after 24 h (i.e. 48 hrs after the cells were mixed with primate sera) in triplicate using Tecan Infinite M200PRO under emission at 590 nm and excitation at 560 nm, and the results were analyzed using Anova. The relative survival was calculated as a ratio in a given serum compared to FBS. 3 Results 3.1 Detection of trypanosomes in feces of experimentally infected mice Using the nested PCR protocol, the ITS1 region of T. b. brucei , T. b. gambiense , T. b. rhodesiense and T. b. evansi was successfully amplified from DNA isolated from fecal samples of infected mice, with relevant controls being negative ( Fig. 1 ). Single abundant amplicons were of expected size (∼450 bp) and sequence ( Fig. 2 ). This experiment proved that under our experimental conditions, the flagellate DNA can be reliably detected in feces of laboratory mice using a 35 cycle PCR protocol. 3.2 Detection of trypanosomes in primates and phylogenetic analysis Overall, 132 samples composed of 119 various tissues and 13 feces, collected from 109 individual primates belonging to six species were subjected to diagnostic PCR with trypanosome-specific primers ( Tables 1 and 2 ). Nested PCR protocol established to amplify ITS1 from the fecal-derived DNA of laboratory-infected mice ( Fig. 1 ) was applied to DNA isolated from the tissue and fecal samples. The sequences of amplicons obtained from 17 primate tissue samples were subjected to BLAST analysis, aligned and grouped in a dendrogram, which showed that our sequences belong to the T. brucei group ( Fig. 2 , see below). Since in the experimentally infected mice, the ITS1-specific primers were able to amplify the target region of trypanosomes from the fecal samples, we have performed the same PCR on DNA isolated from fecal samples of wild chimpanzees. Out of 13 freshly collected feces, three were positive for trypanosome DNA ( Fig. 2 ), as confirmed by sequencing. DNA isolated from the same samples and analyzed independently in another laboratory by the same method lead to identical results. Next, we have generated an alignment from all ITS1 sequences (for detail see Table 2 ) amplified from the tissue samples of chimpanzees (number of samples = 12), sooty mangabeys (n = 1) and Western red colobus monkeys (n = 4), and also from the feces of chimpanzees (n = 3) ( Table 2 ). The dataset was complemented with relevant sequences available in GenBank, with Crithidia fasciculata used as outgroup. With two exceptions (TA11 and TA12, see Fig. 2 ), all newly obtained sequences fell into the T. brucei clade, as shown by a neighbor joining dendrogram ( Fig. 2 ). The T. brucei clade contains sequences from all known subspecies and is characterized by extremely short branches, which is not surprising given the known minimal differences among individual subspecies. A total of 15 sequences obtained from the tissues of chimpanzees, Western red colobus monkeys and a sooty mangabey fell into this clade, as well as three sequences obtained from the feces of three chimpanzees, providing strong evidence for the infections by the T. brucei group parasites ( Table 2 ; Fig. 2 ). All our attempts to amplify the TgSGP gene specific for T. b. gambiense ( Capewell et al., 2013b ) from the tissues and feces of apes positive for Trypanosoma ITS1 PCR assay invariably failed, although this gene can be easily amplified from feces and blood of mice infected with this T. brucei subspecies ( Supplementary data 1 ). Out of 18 positive samples belonging to the T. brucei clade, two samples, one from a spleen and second one from a lymph node of two post-mortem dissected wild Western chimpanzees originating from Taï NP, Côte d'Ivoire, fell outside of the T. brucei clade ( Table 2 ; Fig. 2 ). While TA 11 is almost identical with an unnamed Trypanosoma sp. (Access. No. JN673403 ) from a wildebeest ( Connochaetes ) captured in Serengeti NP, Tanzania, TA 12 from an adult female chimpanzee from Taï NP is strongly affiliated with T. theileri ( Fig. 2 ). Due to the limited information residing in the obtained ITS1 region, multiple attempts to amplify (fragments of) the SSU rRNA gene of the above-described pathogens were performed seeking to establish more precisely their species status. Unfortunately, these attempts almost invariably failed. However, in the case of the TA 12 tissue sample, a fragment of the SSU rRNA gene was successfully amplified which allowed us to assign the respective parasite with high confidence to T . theileri . 3.3 Trypanolytic capacity of primate sera Only mandrill and human sera (NHS) efficiently lysed T. b. brucei , with relative survival 0.101 ± 0.005 (mean ± SD; p  0.05) ( Supplementary data 2 ), while this pathogen also prospered in the other sera. The cultured T. b. rhodesiense cells were very efficiently lysed by the mandrill serum (0.138 ± 0.001; p  0.05) ( Supplementary data 2 ), while this pathogen also prospered in the other sera. The cultured T. b. rhodesiense cells were very efficiently lysed by the mandrill serum (0.138 ± 0.001; p < 0.001), with the sera of spider monkey and orangutan showing a moderate lysis (0.837 ± 0.054, p < 0.01 and 0.848 ± 0.067, p < 0.01 respectively). Under our conditions, NHS caused a very minor yet significant lysis as well (0.874 ± 0.075; p < 0.05). 4 Discussion Within the last decade it has become abundantly clear that African great apes play a significant role in evolution, emergence, and transmission of a range of infectious diseases important for human populations. Simian immunodeficiency viruses and apicomplexan parasites of the genus Plasmodium are among the best known examples ( Liu et al., 2010; Calvignac-Spencer et al., 2012; Liu et al., 2014 ). Although the wild chimpanzees and other African great apes live in regions endemic for African sleeping sickness, the research on their trypanosome infections has been neglected. This situation is mostly caused by the virtually insurmountable difficulties in obtaining blood samples from free-ranging animals and ethical unacceptability of experimental infections on captive animals. Interestingly, African primates exhibit a widely different degree of resistance against parasites of the T. brucei group ( Wheeler, 2010 ). In the literature, there are only very few notes on flagellates found in the blood of dead free-ranging chimpanzees, without any evidence that their death was a consequence of trypanosomiasis ( Hoare, 1972 ). Early experimental infections of chimpanzees with different subspecies of T. brucei led to controversial results: while T. b. brucei and T. b. rhodesiense resulted in fatal parasitemia ( Baker, 1962, 1968; Godfrey and Killick-Kendrick, 1967; Baker and Taylor, 1971 ), T. b. gambiense did not cause any obvious pathology despite the presence of trypomastigotes in the cerebro-spinal fluid ( Godfrey and Killick-Kendrick, 1967; Hoare, 1972 ). This suggests that chimpanzees might serve as reservoir hosts for T. b. gambiense . Moreover, recent results demonstrated the existence of trypanosomatid hybrids among some subspecies of the T. brucei group, which have various animal reservoirs and are capable of infecting humans ( Tait et al., 2011; Capewell et al., 2013a; Duffy et al., 2014; Echodu et al., 2015 ). Hence, addressing the question whether the free-ranging non-human primates may serve as a host or reservoir for some of these trypanosomes is relevant. Inspired by recent advances in our understanding of mechanisms of resistance to trypanosomes mediated by the TLF ( Raper and Friedman, 2013 ), we performed the experiments, in the frame of which three subspecies of the T. brucei group were incubated with sera from mandrill, barbary macaque, spider monkey, orangutan and chimpanzee. The obtained results on T. b. brucei and T. b. gambiense are in concordance with previous experiments of Lugli et al. (2004) , though we also demonstrated that the serum from a mandrill is able to also lyse T. b. gambiense . Furthermore, the chimpanzee sera failed to lyse T. b. brucei , T. b. rhodesiense and also T. b. gambiense . However, results obtained with the latter subspecies are in contrast with the early infection experiments ( Baker, 1962, 1968; Godfrey and Killick-Kendrick, 1967; Baker and Taylor, 1971 ), perhaps due to the fact that those have been conducted on animals originating from nature. This indicates that wild chimpanzees could have a (partial) resistance to T. b. gambiense , especially in the endemic regions. Our experiment confirmed the notion that chimpanzees are fully susceptible to T. b. brucei , against which humans and other primates developed efficient resistance ( Xong et al., 1998; Raper and Friedman, 2013; Uzureau et al., 2013; Thomson et al., 2014 ). These results suggest that the transmission of any of the T. brucei group flagellates into the blood of chimpanzees shall result into an infection. Inspired by successful detection of the Plasmodium DNA in primate feces ( Liu et al., 2010; Jirků et al., 2012 ), we proved the ability of ITS1-based PCR assay to amplify the DNA of all four T. brucei subspecies in feces of experimentally infected mice. PCR-based screening of the available set of fecal samples from wild chimpanzees revealed the presence of trypanosomes, and this finding was further supported by positive results from a number of tissue samples taken from dead chimpanzees. By allowing the detection of trypanosomes in feces, this assay overcomes the inaccessibility of blood and tissue samples from animals that are endangered in the wild. This detection method provides new opportunities for studies of trypanosomatid parasites in the African great apes, and potentially in other rare or highly endangered mammals. Our dendogram of the ITS1 region showed that absolute majority of the obtained sequences fell into the robust T. brucei clade, providing strong evidence that these infections were caused by one of its subspecies. However, since all attempts to amplify T. b. rhodesiense and T. b. gambiense -specific DNA regions failed, the subspecies status could not be established. In any case, the introduction of a PCR assay, capable of amplifying trypanosome DNA from the feces of great apes and potentially other infected mammals, will allow non-invasive large-scale screening for these important parasites. In summary, via feces- and tissue-based PCR assays, we have demonstrated that chimpanzees are frequently infected with trypanosomes from the T. brucei group and rarely also by other species. Future research might shed light on how chimpanzees control trypanosomiasis, and whether transmission occurs among them, other primates and humans. Appendix A Supplementary data The following are the supplementary data related to this article: Supplementary Data 1 Unsuccessful detection of TgSGP gene in primate samples positive for trypanosomes based on the ITS1 region; (A–J) samples from infected primates, but negative for TgSGP; (K) positive control; (L) negative control; (M) marker. Supplementary Data 2 Survival of trypanosomes in sera. Trypanosomes were cultivated in media with human, mandrill, barbary macaque, orangutan, spider monkey and chimpanzee sera, and their relative survival was estimated using the Alamar Blue assay. The survival was standardized using FBS, with NHS used as a positive control. The experiment was performed in triplicate, bars indicate average ± SD values.2

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8423223/

Subtle immunological differences in mRNA-1273 and BNT162b2 COVID-19 vaccine induced Fc-functional profiles

The successful development of several COVID-19 vaccines has substantially reduced morbidity and mortality in regions of the world where the vaccines have been deployed. However, in the wake of the emergence of viral variants, able to evade vaccine induced neutralizing antibodies, real world vaccine efficacy has begun to show differences across the mRNA platforms, suggesting that subtle variation in immune responses induced by the BNT162b2 and mRNA1273 vaccines may provide differential protection. Given our emerging appreciation for the importance of additional antibody functions, beyond neutralization, here we profiled the post-boost binding and functional capacity of the humoral response induced by the BNT162b2 and mRNA-1273 in a cohort of hospital staff. Both vaccines induced robust humoral immune responses to WT SARS-CoV-2 and VOCs. However, differences emerged across epitope-specific responses, with higher RBD- and NTD- specific IgA, as well as functional antibodies (ADNP and ADNK) in mRNA-1273 vaccine recipients. Additionally, RBD-specific antibody depletion highlighted the different roles of non-RBD-specific antibody effector function induced across the mRNA vaccines, providing novel insights into potential differences in protective immunity generated across these vaccines in the setting of newly emerging VOCs. Introduction: The unprecedented rapid development of multiple SARS-CoV-2 vaccines marked a breakthrough in vaccine development and provided hope for an end to the COVID-19 pandemic. However, rising numbers of breakthrough infections, driven by evolving variants of concern in the setting of waning immunity, have clearly illustrated the urgent need to define correlates of immunity. Preliminary immune correlates analyses have shown a strong relationship between neutralizing antibody levels and vaccine efficacy 1 . Yet, surprisingly, antibody binding titers provide even stronger surrogate of protection across vaccine platforms 2 – 4 , with protection observed prior to the evolution of neutralizing antibodies 4 , 5 , and has persisted even in the setting of waning neutralizing antibodies 6 . These data argue for a potential role for alternative protective antibody mechanisms of action. Beyond their role in binding and neutralization, antibodies mediate a wide array of additional protective immunological functions through their ability to recruit the immune system via Fc receptors and complement 7 . Fc-effector functions have been implicated in protection against multiple pathogens, including influenza 8 , anthrax 9 , HIV 10 , malaria 11 , and the Ebola virus 12 . Likewise, Fc-effector functions have been linked to protection against SARS-CoV-2 both following vaccination and administration of monoclonal therapeutics in animal models 13 – 15 . Importantly, Fc-effector function has been implicated in reducing the severity of disease rather than transmission, and thus likely to play a more critical role in vaccine attenuated disease, rather than simple blockade of infection. While accumulating data points to the ability of adenoviral platforms to evoke strong Fc-effector functions 16 , 17 that have been linked to protection against HIV or malaria 18 , less is known about the ability of newer vaccine platforms, including the mRNA vaccine, in eliciting these functions. Striking protection was observed in the phase 3 BNT162b2 4 and mRNA-1273 19 studies, with 94.1% and 95% vaccine efficacy observed at a time when the D614G strain was in circulation. Yet, despite the similar antibody titers and neutralizing antibody levels induced by these mRNA vaccines, emerging real world data have begun to point to differences in the real world efficacy of these two vaccines. Specifically, in the face of the Delta-variant, ~40% and ~75% efficacy has been noted in BNT162b2 and mRNA-1273 vaccinees 20 . Preliminary data in pregnant women have begun to point to differences in vaccine induced humoral immune responses elicited by the Pfizer/BioNTech (BNT162b2), and Moderna (mRNA-1273) vaccines 21 proposed to be driven by differences in vaccine dose, formulation, and/or the one week-delay in boosting 22 . However, whether similar differences exist in the non-pregnant vaccinees, particularly across VOCs, remains incompletely understood. Thus, here we compared the humoral response across the BNT162b2 and mRNA-1273 at the peak immunogenicity in a group of hospital staff. Both vaccines induced robust functional humoral immune responses, yet differences were noted in the vaccine induced antibody profiles across the vaccine groups, with higher RBD- and NTD- specific IgA, as well as functional antibodies (ADNP and ADNK) among mRNA-1273 immunized vaccinees. Both mRNA vaccines drove robust responses across the VOCs, able to bind to multiple FcRs, including the beta and delta variants. Moreover, RBD-specific antibody depletion highlighted the presence of non-RBD-specific antibody effector function deployed by both platforms, albeit at different levels, providing an explanation for differential sustained levels of Fc-effector function observed in the face of evolving VOCs. Methods: Study populations mRNA-vaccinated hospital staff: The Beth Israel Deaconess Medical Center institutional review board approved this study (#2021P000344) and the parent biorepository study (#2020P000361); participants provided written informed consent. We conducted a descriptive cohort study of hospital staff ≥18 years of age who were planning to receive an mRNA COVID-19 vaccine from December 2020 through February 2021 using samples collected in a larger hospital-wide, prospective data and tissue biorepository. Participants self-referred from flyers posted in the hospital vaccine clinics. All participants provided blood samples collected close to each vaccine dose and two to eight weeks after the second dose for the mRNA-1273 (Moderna) or BNT162b2 (Pfizer/BioNTech) COVID-19 vaccine. The analysis presented here includes non-pregnant individuals without immunosuppression medication use. To further characterize the study population, participants were asked to provide their race and ethnicity based on specified categories for each; they could select multiple race categories. Participants also reported if they had fever symptoms following either vaccine dose. Community-acquired COVID-19 individuals: Industry employees (Space Exploration Technologies Corp.) were volunteer tested for COVID-19, starting in April 2020. Participants completed a study survey including the collection of COVID-19 related symptoms. The cohort largely included mild-symptomatic infections 45 . Upon obtaining informed consent, blood samples were collected and used for immune profiling. The median age of the seropositive population was 32 years (range 19–62 years), and 84% were males. The enrolled participants were 66% White, 8% Asian, 6% More than one race, 2% Black, 1% American Indian/Alaska Native and 17% unknown. Volunteers were tested by PCR and for antibodies monthly. All antibody positive individuals were included in the study 45 . Both studies were approved by the Massachusetts General Brigham Healthcare (previously Partners Healthcare) Institutional Review Board. All participants provided written informed consent. Antigens Antigens used for Luminex based assays: SARS-CoV-2 D614G WT S (kindly provided by Erica Saphire, La Jolla Institute for Immunology), SARS-CoV-2 S1 (Sino Biological), SARS-CoV-2 S2 (Sino Biological) and SARS-CoV-2 RBD (kindly provided by Aaron Schmidt, Ragon Institute) as well as SARS-CoV-2 VOC,s such as Alpha B.1.1.7 S (LakePharma), Beta B.1.351 S (LakePharma), Gamma P1 S (LakePharma), Epsilon B.1.427 (kindly provided by Erica Saphire, La Jolla Institute for Immunology), Iota B.1.526 (LakePharma), Kappa B.1.617.1 S (Sino Biological) and Delta B.1.617.2 S (kindly provided by Erica Saphire, La Jolla Institute for Immunology) and Alpha B.1.1.7, Beta B.1.351, Gamma P1, Kappa B.1.617.2 and Delta B.1.617.2 RBDs (kindly provided by Florian Krammer, Icahn School of Medicine at Mount Sinai). Luminex profiling Serum samples were analyzed by customized Luminex assay to quantify the relative concentration of antigen-specific antibody isotypes, subclasses, and Fcγ-receptor (FcγR) binding profiles, as previously described 46 , 47 . Briefly, SARS-CoV-2 antigens were used to profile specific humoral immune responses. Antigens were coupled to magnetic Luminex beads (Luminex Corp) by carbodiimide-NHS ester-coupling (Thermo Fisher). Antigen-coupled microspheres were washed and incubated with plasma samples at an appropriate sample dilution (1:500 for IgG1 and all low affinity Fcγ- receptors, and 1:100 for all other readouts) for 2 hours at 37°C in 384-well plates (Greiner Bio-One). The high affinity FcR was not tested due to its minimal role in tuning antibody effector function 48 . Unbound antibodies were washed away, and antigen-bound antibodies were detected by using a PE-coupled detection antibody for each subclass and isotype (IgG1, IgG3, IgA1, and IgM; Southern Biotech), and Fcγ-receptors were fluorescently labeled with PE before addition to immune complexes (FcγR2a, FcγR3a; Duke Protein Production facility). After one hour of incubation, plates were washed, and flow cytometry was performed with an iQue (Intellicyt), and analysis was performed on IntelliCyt ForeCyt (v8.1). PE median fluorescent intensity (MFI) is reported as a readout for antigen-specific antibody titers. Antibody-dependent complement deposition (ADCD) Antibody-dependent complement deposition (ADCD) was conducted as previously described 49 . Briefly, SARS-CoV-2 antigens were coupled to magnetic Luminex beads (Luminex Corp) by carbodiimide-NHS ester-coupling (Thermo Fisher). Coupled beads were incubated for 2 hours at 37°C with serum samples (1:10 dilution) to form immune complexes and then washed to remove unbound immunoglobulins. In order to measure antibody-dependent deposition of C3, lyophilized guinea pig complement (Cedarlane) was diluted in gelatin veronal buffer with calcium and magnesium (GBV++) (Boston BioProducts) and added to immune complexes. Subsequently, C3 was detected with an anti-C3 fluorescein-conjugated goat IgG fraction detection antibody (Mpbio). The flow cytometry was performed with iQue (Intellicyt) and an S-Lab robot (PAA). ADCD was reported as the median of C3 deposition. Antibody-dependent cellular (ADCP) and neutrophil (ADNP) phagocytosis Antibody-dependent cellular phagocytosis (ADCP) and antibody-dependent neutrophil phagocytosis (ADNP) were conducted according to the previously described protocols 50 , 51 . In detail, SARS-CoV-2 antigens were biotinylated using EDC (Thermo Fisher) and Sulfo-NHS-LCLC biotin (Thermo Fisher) and coupled to yellow-green (505/515) fluorescent Neutravidinconjugated beads (Thermo Fisher), respectively. To form immune complexes, antigen-coupled beads were incubated for 2 hours at 37°C with 1:100 diluted serum samples and then washed to remove unbound immunoglobulins. For ADCP, the immune complexes were incubated for 16–18 hours with THP-1 cells (1.25×10 5 THP-1 cells/mL) and for ADNP for 1 hour with RBC-lyzed whole blood. Following the incubation, cells were fixed with 4% PFA. For ADNP, RBC-lyzed whole blood was washed, stained for CD66b+ (Biolegend) to identify neutrophils, and then fixed in 4% PFA. Flow cytometry was performed to identify the percentage of cells that had phagocytosed beads as well as the number of beads that had been phagocytosis (phagocytosis score = % positive cells × Median Fluorescent Intensity of positive cells/10000). The Flow cytometry was performed with 5 Laser LSR Fortessa Flow Cytometer and analysis was performed using FlowJo V10.7.1. Antibody-dependent NK cell degranulation Antibody dependent NK cell degranulation as described previously (49). Briefly, SARS-CoV-2 antigens were coated to 96-well ELISA at the protein concentration of 2 g/ml, incubated at 37°C for 2hrs and blocked with 5% BSA at 4°C overnight. NK cells were isolated from whole blood from healthy donors (by negative selection using RosetteSep (STEMCELL) then separated using a ficoll gradient. NK cells were rested overnight in media supplemented with IL-15. Serum samples were diluted at 1:25. After blocking, samples were added to coated plates and immune complexes were formed for two hours at 37°C. After the two hours, NK cells were prepared (antiCD107a– phycoerythrin (PE) – Cy5 (BD), brefeldin A (10 μg/ml) (Sigma), and GolgiStop (BD)), and added to each well. for 5 hours at 37°C. The cells were stained for surface markers using anti-CD3 PacBlue (BD), anti-CD16 APC-Cy5 (BD), and anti-CD56 PE-Cy7 (BD) and permeabilized. The flow cytometry was performed. NK cells were gates as CD3−, CD16+, CD56+ cells and NK cell activity was determined as the percent of NK cells positive for CD107a and MIP-1b. RBD-specific antibody depletion from polyclonal serum samples SARS-CoV-2 RBD-coated magnetic beads (ACROBiosystems) were prepared according to the manufacturer's protocol and resuspended in ultrapure water at 1 mg/ml concentration. Beads were washed three times in phosphate-buffered saline (PBS) with 0.05% bovine serum albumin (BSA) by using a magnet. Serum samples were incubated with beads, at 3:1 beads:serum ratio, rotating overnight at 4°C. A magnet was used to separate beads with surfaces-bounded RBD-specific antibodies from the RBD-specific antibodies-depleted supernatant. A mock depletion (pre-depletion samples) was performed by adding 150 μl of PBS + 0.05% BSA and incubating rotating overnight at 4°C. Functional assays were performed with pre- and post-depletion samples. Statistics Data analysis was performed using R version 4.0.2 (2020-06-22). Comparisons between vaccination arms were performed using Mann-Whitney U-test test followed by Benjamini Hochberg (BH) correction. Antigen responses (e.g., wild type to alpha) were compared using Wilcoxon-signed rank test followed by BH. Comparisons between RBD+ and RBD− samples were done using paired t-test. Prior to any multivariate analysis, all data were normalized using z-scoring. Multivariate classification models were trained to discriminate between individuals vaccinated with BNT162b2 and individuals vaccinated with mRNA-1273 using all the measured antibody responses. Models were built using a combination of the least absolute shrinkage and selection operator (LASSO) for feature selection and then classification using partial least square discriminant analysis (PLS-DA) with the LASSO-selected features using R package "ropls" version 1.20.0 (Thévenot et al., 2015) and "glmnet" version 4.0.2. Model accuracy was assessed using ten-fold cross-validation. For each test fold, LASSO-based feature selection was performed on logistic regression using the training set for that fold. LASSO was repeated 100 times, features selected at least 90 times out of 100 were identified as selected features. PLS-DA classifier was applied to the training set using the selected features and prediction accuracy was recorded. Selected features were ordered according to their Variable Importance in Projection (VIP) score and the first two latent variables (LVs) of the PLS-DA model were used to visualize the samples. A co-correlate network analysis was carried out to identify features that highly correlate with the LASSO selected features, and thus are potentially equally important for discriminating the vaccination arms. Correlations for the co-correlate network were performed using Spearman method followed by Benjamini-Hochberg multiple correction 52 . The co-correlate network was generated using R package "network" version 1.16.0 53 . All other figures were generated using ggplot2 54 . Study populations mRNA-vaccinated hospital staff: The Beth Israel Deaconess Medical Center institutional review board approved this study (#2021P000344) and the parent biorepository study (#2020P000361); participants provided written informed consent. We conducted a descriptive cohort study of hospital staff ≥18 years of age who were planning to receive an mRNA COVID-19 vaccine from December 2020 through February 2021 using samples collected in a larger hospital-wide, prospective data and tissue biorepository. Participants self-referred from flyers posted in the hospital vaccine clinics. All participants provided blood samples collected close to each vaccine dose and two to eight weeks after the second dose for the mRNA-1273 (Moderna) or BNT162b2 (Pfizer/BioNTech) COVID-19 vaccine. The analysis presented here includes non-pregnant individuals without immunosuppression medication use. To further characterize the study population, participants were asked to provide their race and ethnicity based on specified categories for each; they could select multiple race categories. Participants also reported if they had fever symptoms following either vaccine dose. Community-acquired COVID-19 individuals: Industry employees (Space Exploration Technologies Corp.) were volunteer tested for COVID-19, starting in April 2020. Participants completed a study survey including the collection of COVID-19 related symptoms. The cohort largely included mild-symptomatic infections 45 . Upon obtaining informed consent, blood samples were collected and used for immune profiling. The median age of the seropositive population was 32 years (range 19–62 years), and 84% were males. The enrolled participants were 66% White, 8% Asian, 6% More than one race, 2% Black, 1% American Indian/Alaska Native and 17% unknown. Volunteers were tested by PCR and for antibodies monthly. All antibody positive individuals were included in the study 45 . Both studies were approved by the Massachusetts General Brigham Healthcare (previously Partners Healthcare) Institutional Review Board. All participants provided written informed consent. mRNA-vaccinated hospital staff: The Beth Israel Deaconess Medical Center institutional review board approved this study (#2021P000344) and the parent biorepository study (#2020P000361); participants provided written informed consent. We conducted a descriptive cohort study of hospital staff ≥18 years of age who were planning to receive an mRNA COVID-19 vaccine from December 2020 through February 2021 using samples collected in a larger hospital-wide, prospective data and tissue biorepository. Participants self-referred from flyers posted in the hospital vaccine clinics. All participants provided blood samples collected close to each vaccine dose and two to eight weeks after the second dose for the mRNA-1273 (Moderna) or BNT162b2 (Pfizer/BioNTech) COVID-19 vaccine. The analysis presented here includes non-pregnant individuals without immunosuppression medication use. To further characterize the study population, participants were asked to provide their race and ethnicity based on specified categories for each; they could select multiple race categories. Participants also reported if they had fever symptoms following either vaccine dose. Community-acquired COVID-19 individuals: Industry employees (Space Exploration Technologies Corp.) were volunteer tested for COVID-19, starting in April 2020. Participants completed a study survey including the collection of COVID-19 related symptoms. The cohort largely included mild-symptomatic infections 45 . Upon obtaining informed consent, blood samples were collected and used for immune profiling. The median age of the seropositive population was 32 years (range 19–62 years), and 84% were males. The enrolled participants were 66% White, 8% Asian, 6% More than one race, 2% Black, 1% American Indian/Alaska Native and 17% unknown. Volunteers were tested by PCR and for antibodies monthly. All antibody positive individuals were included in the study 45 . Both studies were approved by the Massachusetts General Brigham Healthcare (previously Partners Healthcare) Institutional Review Board. All participants provided written informed consent. Antigens Antigens used for Luminex based assays: SARS-CoV-2 D614G WT S (kindly provided by Erica Saphire, La Jolla Institute for Immunology), SARS-CoV-2 S1 (Sino Biological), SARS-CoV-2 S2 (Sino Biological) and SARS-CoV-2 RBD (kindly provided by Aaron Schmidt, Ragon Institute) as well as SARS-CoV-2 VOC,s such as Alpha B.1.1.7 S (LakePharma), Beta B.1.351 S (LakePharma), Gamma P1 S (LakePharma), Epsilon B.1.427 (kindly provided by Erica Saphire, La Jolla Institute for Immunology), Iota B.1.526 (LakePharma), Kappa B.1.617.1 S (Sino Biological) and Delta B.1.617.2 S (kindly provided by Erica Saphire, La Jolla Institute for Immunology) and Alpha B.1.1.7, Beta B.1.351, Gamma P1, Kappa B.1.617.2 and Delta B.1.617.2 RBDs (kindly provided by Florian Krammer, Icahn School of Medicine at Mount Sinai). Luminex profiling Serum samples were analyzed by customized Luminex assay to quantify the relative concentration of antigen-specific antibody isotypes, subclasses, and Fcγ-receptor (FcγR) binding profiles, as previously described 46 , 47 . Briefly, SARS-CoV-2 antigens were used to profile specific humoral immune responses. Antigens were coupled to magnetic Luminex beads (Luminex Corp) by carbodiimide-NHS ester-coupling (Thermo Fisher). Antigen-coupled microspheres were washed and incubated with plasma samples at an appropriate sample dilution (1:500 for IgG1 and all low affinity Fcγ- receptors, and 1:100 for all other readouts) for 2 hours at 37°C in 384-well plates (Greiner Bio-One). The high affinity FcR was not tested due to its minimal role in tuning antibody effector function 48 . Unbound antibodies were washed away, and antigen-bound antibodies were detected by using a PE-coupled detection antibody for each subclass and isotype (IgG1, IgG3, IgA1, and IgM; Southern Biotech), and Fcγ-receptors were fluorescently labeled with PE before addition to immune complexes (FcγR2a, FcγR3a; Duke Protein Production facility). After one hour of incubation, plates were washed, and flow cytometry was performed with an iQue (Intellicyt), and analysis was performed on IntelliCyt ForeCyt (v8.1). PE median fluorescent intensity (MFI) is reported as a readout for antigen-specific antibody titers. Antibody-dependent complement deposition (ADCD) Antibody-dependent complement deposition (ADCD) was conducted as previously described 49 . Briefly, SARS-CoV-2 antigens were coupled to magnetic Luminex beads (Luminex Corp) by carbodiimide-NHS ester-coupling (Thermo Fisher). Coupled beads were incubated for 2 hours at 37°C with serum samples (1:10 dilution) to form immune complexes and then washed to remove unbound immunoglobulins. In order to measure antibody-dependent deposition of C3, lyophilized guinea pig complement (Cedarlane) was diluted in gelatin veronal buffer with calcium and magnesium (GBV++) (Boston BioProducts) and added to immune complexes. Subsequently, C3 was detected with an anti-C3 fluorescein-conjugated goat IgG fraction detection antibody (Mpbio). The flow cytometry was performed with iQue (Intellicyt) and an S-Lab robot (PAA). ADCD was reported as the median of C3 deposition. Antibody-dependent cellular (ADCP) and neutrophil (ADNP) phagocytosis Antibody-dependent cellular phagocytosis (ADCP) and antibody-dependent neutrophil phagocytosis (ADNP) were conducted according to the previously described protocols 50 , 51 . In detail, SARS-CoV-2 antigens were biotinylated using EDC (Thermo Fisher) and Sulfo-NHS-LCLC biotin (Thermo Fisher) and coupled to yellow-green (505/515) fluorescent Neutravidinconjugated beads (Thermo Fisher), respectively. To form immune complexes, antigen-coupled beads were incubated for 2 hours at 37°C with 1:100 diluted serum samples and then washed to remove unbound immunoglobulins. For ADCP, the immune complexes were incubated for 16–18 hours with THP-1 cells (1.25×10 5 THP-1 cells/mL) and for ADNP for 1 hour with RBC-lyzed whole blood. Following the incubation, cells were fixed with 4% PFA. For ADNP, RBC-lyzed whole blood was washed, stained for CD66b+ (Biolegend) to identify neutrophils, and then fixed in 4% PFA. Flow cytometry was performed to identify the percentage of cells that had phagocytosed beads as well as the number of beads that had been phagocytosis (phagocytosis score = % positive cells × Median Fluorescent Intensity of positive cells/10000). The Flow cytometry was performed with 5 Laser LSR Fortessa Flow Cytometer and analysis was performed using FlowJo V10.7.1. Antibody-dependent NK cell degranulation Antibody dependent NK cell degranulation as described previously (49). Briefly, SARS-CoV-2 antigens were coated to 96-well ELISA at the protein concentration of 2 g/ml, incubated at 37°C for 2hrs and blocked with 5% BSA at 4°C overnight. NK cells were isolated from whole blood from healthy donors (by negative selection using RosetteSep (STEMCELL) then separated using a ficoll gradient. NK cells were rested overnight in media supplemented with IL-15. Serum samples were diluted at 1:25. After blocking, samples were added to coated plates and immune complexes were formed for two hours at 37°C. After the two hours, NK cells were prepared (antiCD107a– phycoerythrin (PE) – Cy5 (BD), brefeldin A (10 μg/ml) (Sigma), and GolgiStop (BD)), and added to each well. for 5 hours at 37°C. The cells were stained for surface markers using anti-CD3 PacBlue (BD), anti-CD16 APC-Cy5 (BD), and anti-CD56 PE-Cy7 (BD) and permeabilized. The flow cytometry was performed. NK cells were gates as CD3−, CD16+, CD56+ cells and NK cell activity was determined as the percent of NK cells positive for CD107a and MIP-1b. RBD-specific antibody depletion from polyclonal serum samples SARS-CoV-2 RBD-coated magnetic beads (ACROBiosystems) were prepared according to the manufacturer's protocol and resuspended in ultrapure water at 1 mg/ml concentration. Beads were washed three times in phosphate-buffered saline (PBS) with 0.05% bovine serum albumin (BSA) by using a magnet. Serum samples were incubated with beads, at 3:1 beads:serum ratio, rotating overnight at 4°C. A magnet was used to separate beads with surfaces-bounded RBD-specific antibodies from the RBD-specific antibodies-depleted supernatant. A mock depletion (pre-depletion samples) was performed by adding 150 μl of PBS + 0.05% BSA and incubating rotating overnight at 4°C. Functional assays were performed with pre- and post-depletion samples. Statistics Data analysis was performed using R version 4.0.2 (2020-06-22). Comparisons between vaccination arms were performed using Mann-Whitney U-test test followed by Benjamini Hochberg (BH) correction. Antigen responses (e.g., wild type to alpha) were compared using Wilcoxon-signed rank test followed by BH. Comparisons between RBD+ and RBD− samples were done using paired t-test. Prior to any multivariate analysis, all data were normalized using z-scoring. Multivariate classification models were trained to discriminate between individuals vaccinated with BNT162b2 and individuals vaccinated with mRNA-1273 using all the measured antibody responses. Models were built using a combination of the least absolute shrinkage and selection operator (LASSO) for feature selection and then classification using partial least square discriminant analysis (PLS-DA) with the LASSO-selected features using R package "ropls" version 1.20.0 (Thévenot et al., 2015) and "glmnet" version 4.0.2. Model accuracy was assessed using ten-fold cross-validation. For each test fold, LASSO-based feature selection was performed on logistic regression using the training set for that fold. LASSO was repeated 100 times, features selected at least 90 times out of 100 were identified as selected features. PLS-DA classifier was applied to the training set using the selected features and prediction accuracy was recorded. Selected features were ordered according to their Variable Importance in Projection (VIP) score and the first two latent variables (LVs) of the PLS-DA model were used to visualize the samples. A co-correlate network analysis was carried out to identify features that highly correlate with the LASSO selected features, and thus are potentially equally important for discriminating the vaccination arms. Correlations for the co-correlate network were performed using Spearman method followed by Benjamini-Hochberg multiple correction 52 . The co-correlate network was generated using R package "network" version 1.16.0 53 . All other figures were generated using ggplot2 54 . Results: Enrollment Seventy-three participants, 28 receiving mRNA-1273 and 45 receiving BNT162b2, were included from the hospital-wide biorepository of vaccinated individuals who received an mRNA COVID19 vaccine and had serum available for analysis following their second vaccine dose ( Table 1 ). The vaccines were delivered intramuscularly, 30 μg of BNT162b2 and 100 μg of mRNA-127 were delivered, three and four weeks apart, respectively. Samples were obtained a median (interquartile range, IQR) of 19 (15, 26) days after the second vaccine dose. Prior SARS-CoV-2 infection (mild disease) was diagnosed in 7% of mRNA-1273 vaccinated and 2% of BNT162b2 vaccinated. After the second dose, fever was reported in 12 (48%) mRNA-1273 vaccinated and 19 (45%) of BNT162b2 vaccinated participants ( Table 1 ). mRNA-1273 and BNT162b2 COVID-19 vaccines induce robust WT-antibody Fc-profiles The two approved mRNA vaccines clearly induce robust antibody titers and neutralization 23 , 24 , however, real world efficacy data has begun to show differences across the vaccines in their ability to resist delta viral infection 20 . Specifically, approximately 42% protection has been noted following the BNT162b2 vaccine, and 76% efficacy has been observed with the mRNA-1273 vaccine. Recent profiling in pregnant women highlighted significant differences in the Fc-quality of the humoral immune responses induced by these two vaccines 25 , potentially related to differences in dosage, formulation, the timing of the boost, or mRNA design. However, whether these same immunological changes occur in non-pregnant vaccinees and vary across emerging variants of concern (VOCs) remains incompletely understood. Thus, here we sought to determine whether the two authorized COVID-19 mRNA vaccines elicit similar Fc profiles in a cohort of 73 health care workers vaccinated with either of the vaccines. The vaccines were delivered intramuscularly, 30 μg of BNT162b2 and 100 μg of mRNA-127 were delivered, three and four weeks apart, respectively, and blood was collected two weeks post the final immunization. The wild-type SARS-CoV-2 RBD-, N-terminal domain (NTD)-, S-, S1- and S2-specific antibody titers, Fc-receptor binding, and Fc-functions were analyzed. Robust vaccine induced antibody responses were observed across both the mRNA-1273 ( n = 28 ) and BNT162b2 ( n = 45 ) vaccines ( Figure. 1A ), marked by slightly higher responses in the mRNA-1273 vaccinees. Univariate comparisons across each antigen and Fc-profile measurement highlighted the presence of equivalent IgG and IgM binding titers, but higher levels of IgA-binding titers elicited by the mRNA-1273 vaccine, particularly to the Spike, RBD, NTD, and S1 domains ( Figure. 1B ). Moreover, robust and largely equivalent cross Fc-receptor binding was observed across both vaccines, with the exception of enhanced NTD-specific Fc-receptor binding antibodies induced by the mRNA-1273 vaccine. Similarly, equivalent levels of antibody dependent complement deposition (ADCD) and antibody dependent cellular phagocytosis by monocytes (ADCP) were observed across the two vaccine groups at peak immunogenicity. Conversely, mRNA-1273 vaccinated individuals exhibited significantly higher levels of antibody dependent neutrophil phagocytosis (ADNP) and antibody dependent NK cell activation (degranulation: CD107a, cytokine secretion: IFN, and chemokine secretion: MIP1) ( Figure.1C ). Given the univariate differences, we next aimed to define whether differences existed in the overall multivariate vaccine profile across the two vaccine groups. Thus, we used the least absolute shrinkage and selection operator (LASSO) feature selection to initially reduce all antibody features to a minimal set which represented the overall variation in the antibody profiles and to avoid over-fitting, followed by classification using partial least squares discriminant analysis (PLS-DA). Separation was observed across the two different mRNA vaccine profiles ( Figure.1D ), marked largely by augmented responses in the mRNA-1273 vaccine induced immune response. Specifically, five features were selectively enhanced in the mRNA-1273 vaccine profiles, including RBD-specific IgA1 and IgG2, as well as NTD-specific IgA1, FcγR2A, and FcγR2B. Given the highly correlated nature of the vaccine induced humoral immune response, a correlation network analysis was built between LASSO-selected features and the overall immune response to define the additional features that may shift differentially across the vaccine profiles ( Figure.1E ). Three clusters appeared, bearing elevated IgA response across all antigenic determinants, a small network of IgG2 responses, and a large network of Fc-receptor binding antibody responses across multiple antigenic targets all enriched among mRNA-1273 immunized individuals. These data point to robust humoral immune responses induced by both mRNA platforms, but that do diverge with enhanced epitope spreading, IgA immunity, and specific antibody effector functions in mRNA-1273 immunized individuals. mRNA-1273 and BNT162b2 vaccines induce FcR-binding responses to multiple VOCs. Despite the remarkable efficacy of the mRNA vaccines against the original SARS-CoV-2 variant, waves of variants have emerged that include amino acid substitutions that significantly diminish neutralizing antibody activity 26 – 28 . Among the variants of concern (VOCs), the mRNA vaccines appear to neutralize Alpha (B.1.1.7) 29 and Gamma (P.1) with only a minimal loss of activity, but exhibit compromised neutralizing activity against the beta (B.1.351) variant 30 31 . Yet, whether Fc responses were equally affected across the VOCs remains unclear. Both mRNA-1273 and BNT162b2 vaccine-induced antibodies bound equally well across the Alpha B.1.1.7, Beta B.1.351, and Gamma P.1 VOCs ( Figure.2A ). Interestingly, IgM titers were higher in BNT162b2 vaccinated individuals to the Beta and Gamma VOCs. Additionally, a trend towards higher IgG1-binding titers was also noted in the BNT162b2 immunized individuals across all 3 VOCs. Conversely, IgA responses were amplified in the mRNA-1273 immunized individuals ( Figure. 2A ). However, Fc receptor-binding antibodies were induced by both vaccines to all 3 VOCs at equivalent levels. Along the same lines, both antibody-dependent monocyte (ADCP) and neutrophil (ADNP) phagocytosis were largely equivalent across the variants ( Figure.2B ), highlighting the robust Fc binding and functional profiles across VOCs elicited by both mRNA platforms. Yet, despite these univariate results, we finally aimed to ask if any multivariate differences could be observed across the two mRNA platforms in their VOC response ( Figure.2C ). The LASSO/PLSDA revealed separation in the Fc-profiles induced to the VOCs between the mRNA-1273 and BNT162b vaccinated individuals ( Figure.2C ). The profile was marked by higher IgM-Beta levels in BNT162b vaccinated individuals. Conversely, higher levels of IgA/IgG2 responses were observed to alpha in mRNA-1273 vaccinated individuals who also had higher levels of FCR2A and FCR2B NTD-binding antibodies. The extended LASSO co-correlate network further highlighted the presence of IgG2, IgA, and IgM-only networks across multiple VOCs, suggesting that isotype biased selection across the mRNA platform includes reactivities across VOCs. Additionally, a large network of highly functional pan VOC and epitope responses were observed in the mRNA1273 profile, marked by an enrichment of NTD-specific antibody responses. Thus, the two mRNA platforms elicit overall similar levels of functional antibodies to the VOCs, with an IgM/IgG biased profile induced by BNT162b2 and a more class switched IgA/IgG-driven profile induced by mRNA-1273. mRNA-1273 and BNT162b2 vaccines induce more robust RBD and full Spike VOC-targeting Fc-functional antibodies compared to natural infection Alarmingly, several variants, including the Alpha (B.1.1.7), Beta (B.1.351), Gamma (P1), and Delta (B.1.617.2) VOCs, have begun to breakthrough natural 32 and vaccine induced immune responses 26 , causing large numbers of outbreaks. The unexpected significant reduction of effectiveness of mRNA vaccines against VOCs, especially delta (B.1617.2) variant, is emerging 20 , albeit with the majority of breakthroughs remaining largely non-lethal 1 . Yet, differential real-world efficacy against the Delta VOC 20 points to a nuanced immune response to delta. Thus, we next aimed to compare the cross-VOC antibody Fc-profiles targeting both VOC RBDs or full Spike antigens across a subset of the vaccinees and a group of mild-community acquired convalescent individuals. Antibody profiles were compared across mRNA-1273 ( n = 16 ) and BNT162b ( n = 15 ) vaccines and 10 convalescents ( Figure.3 ). mRNA vaccine induced IgG1 responses were higher than convalescent responses to the original variant RBD (WT), Alpha (B.1.1.7), Beta (B.1.351), Kappa (B.1.617.1), and Delta (B.1.617.2), but were lower for the gamma (P.1) RBD. Similar patterns were observed across all Fc-receptor RBD-binding antibodies induced by the mRNA vaccines, that all showed superior binding to FcRs compared to convalescent antibodies that bound poorly to all RBD VOCs compared to the WT variant ( Figure.3A ). Interestingly, slightly higher antibody binding was noted across RBD VOCs for mRNA-1273 immunized individuals compared to BNT162b2, albeit the pattern of recognition was the same. Conversely, IgG1 and IgG3 binding patterns to S-specific antibodies revealed enhanced mRNA-spike specific binding to nearly all VOC-full Spike antigens, except the kappa variant, compared to WT Spike-specific antibody binding ( Figure.3B ). Importantly, all Spike-specific binding IgG responses were lower in convalescents compared to mRNA vaccinees, with slightly higher IgG1 and IgG3 binding noted to nearly all VOC Spikes. In contrast, Fc-receptor binding antibodies exhibited equivalent VOC-recognition across both mRNA vaccine platforms, highlighting potential functional differences in RBD- and Spike-specific antibodies induced across the platforms. Thus, despite the more variable FcR-binding profiles to RBD VOCs, stable FcRbinding was noted to most full spike VOCs. Given the persistent protection against delta in newly/recently vaccinated individuals 1 , but enhanced breakthrough over time, these data may suggest that the presence of broad non-RBD-specific functional immunity may be key to protection, given that VOC-RBDs clearly are able to exploit RBD-specific Fc-vulnerabilities even across vaccine platforms. RBD-specific antibody depletion influences antibody-mediated monocyte and neutrophil phagocytosis The data above suggested that potential differences in RBD and Spike specific contributions to polyclonal antibody Fc-binding profiles and function. Thus, to address this possibility, RBD-specific antibodies were depleted from the polyclonal serum of our vaccinees and convalescent samples ( Figure. 4 ) and tested for opsonophagocytic functions linked to the natural resolution of infection 33 , 34 . Specifically, monocyte phagocytosis (ADCP, Figure. 4A ) or neutrophil phagocytosis (ADNP, Figure. 4B ) were evaluated across VOCs. RBD depletion resulted in a significant loss of ADCP against the WT variant Spike ( Figure. 4A , bottom left). However, RBD depletion did not affect the ADCP response in convalescents to all other Spike VOCs, arguing that antibodies in natural infection that drive function largely target areas outside of the RBD. Conversely, RBD depletion resulted in slightly reduced ADCP in BNT162b2 immunized individuals to the WT, beta, and epsilon variants, but not to the alpha and gamma variants. Similarly, RBD-depletion resulted in reduced Beta and Gamma ADCP in mRNA-1273 immunized individuals. These data argue for variable, but low-level alterations in RBD-specific ADCP activity across VOC Spikes following natural infection, and differentially following mRNA vaccination, suggesting that non-RBD-specific antibodies may continue to drive opsonophagocytic control of the virus even in the setting of profound changes in the RBD that may knock out neutralization. In contrast to ADCP, more variation was observed in neutrophil phagocytosis (ADNP) with RBD depletion. Interestingly, RBD depletion led to increased ADNP against the WT and Epsilon variants, suggesting that RBD-specific antibodies may, in fact, block this activity following the natural resolution of disease. Conversely, RBD depletion led to reduced ADNP to the alpha, beta, epsilon, and iota variants in BNT162b2 immunized individuals. Similarly, RBD depletion reduced ADNP to WT, alpha, beta, gamma, and iota in mRNA-1273 vaccinated individuals. Interestingly, the RBD depletion led to a more profound loss of ADNP in mRNA1273 immunized individuals compared to BNT162b2 immunized individuals, marking significant differences in the functional activity of particular sub-populations of antibodies elicited by these vaccines. Thus, given the higher levels of mRNA-1273 induced pan-VOC RBD specific immunity, that appears to contribute largely to ADNP activity, these data point to the possible existence of different epitope-specific functional correlates of immunity elicited across the two vaccine platforms and natural infection. Thus, these data collectively show the robust induction of functional antibody responses that differ in their Fc-biology, following mRNA-1273 and BNT162b2 vaccination, marked by differences in the overall isotype/subclass, Fc-receptor binding profiles, and epitope specific functions across VOCs, providing some potential explanation for differences in persistent protection afforded against by this newly emerging vaccine platform. Enrollment Seventy-three participants, 28 receiving mRNA-1273 and 45 receiving BNT162b2, were included from the hospital-wide biorepository of vaccinated individuals who received an mRNA COVID19 vaccine and had serum available for analysis following their second vaccine dose ( Table 1 ). The vaccines were delivered intramuscularly, 30 μg of BNT162b2 and 100 μg of mRNA-127 were delivered, three and four weeks apart, respectively. Samples were obtained a median (interquartile range, IQR) of 19 (15, 26) days after the second vaccine dose. Prior SARS-CoV-2 infection (mild disease) was diagnosed in 7% of mRNA-1273 vaccinated and 2% of BNT162b2 vaccinated. After the second dose, fever was reported in 12 (48%) mRNA-1273 vaccinated and 19 (45%) of BNT162b2 vaccinated participants ( Table 1 ). mRNA-1273 and BNT162b2 COVID-19 vaccines induce robust WT-antibody Fc-profiles The two approved mRNA vaccines clearly induce robust antibody titers and neutralization 23 , 24 , however, real world efficacy data has begun to show differences across the vaccines in their ability to resist delta viral infection 20 . Specifically, approximately 42% protection has been noted following the BNT162b2 vaccine, and 76% efficacy has been observed with the mRNA-1273 vaccine. Recent profiling in pregnant women highlighted significant differences in the Fc-quality of the humoral immune responses induced by these two vaccines 25 , potentially related to differences in dosage, formulation, the timing of the boost, or mRNA design. However, whether these same immunological changes occur in non-pregnant vaccinees and vary across emerging variants of concern (VOCs) remains incompletely understood. Thus, here we sought to determine whether the two authorized COVID-19 mRNA vaccines elicit similar Fc profiles in a cohort of 73 health care workers vaccinated with either of the vaccines. The vaccines were delivered intramuscularly, 30 μg of BNT162b2 and 100 μg of mRNA-127 were delivered, three and four weeks apart, respectively, and blood was collected two weeks post the final immunization. The wild-type SARS-CoV-2 RBD-, N-terminal domain (NTD)-, S-, S1- and S2-specific antibody titers, Fc-receptor binding, and Fc-functions were analyzed. Robust vaccine induced antibody responses were observed across both the mRNA-1273 ( n = 28 ) and BNT162b2 ( n = 45 ) vaccines ( Figure. 1A ), marked by slightly higher responses in the mRNA-1273 vaccinees. Univariate comparisons across each antigen and Fc-profile measurement highlighted the presence of equivalent IgG and IgM binding titers, but higher levels of IgA-binding titers elicited by the mRNA-1273 vaccine, particularly to the Spike, RBD, NTD, and S1 domains ( Figure. 1B ). Moreover, robust and largely equivalent cross Fc-receptor binding was observed across both vaccines, with the exception of enhanced NTD-specific Fc-receptor binding antibodies induced by the mRNA-1273 vaccine. Similarly, equivalent levels of antibody dependent complement deposition (ADCD) and antibody dependent cellular phagocytosis by monocytes (ADCP) were observed across the two vaccine groups at peak immunogenicity. Conversely, mRNA-1273 vaccinated individuals exhibited significantly higher levels of antibody dependent neutrophil phagocytosis (ADNP) and antibody dependent NK cell activation (degranulation: CD107a, cytokine secretion: IFN, and chemokine secretion: MIP1) ( Figure.1C ). Given the univariate differences, we next aimed to define whether differences existed in the overall multivariate vaccine profile across the two vaccine groups. Thus, we used the least absolute shrinkage and selection operator (LASSO) feature selection to initially reduce all antibody features to a minimal set which represented the overall variation in the antibody profiles and to avoid over-fitting, followed by classification using partial least squares discriminant analysis (PLS-DA). Separation was observed across the two different mRNA vaccine profiles ( Figure.1D ), marked largely by augmented responses in the mRNA-1273 vaccine induced immune response. Specifically, five features were selectively enhanced in the mRNA-1273 vaccine profiles, including RBD-specific IgA1 and IgG2, as well as NTD-specific IgA1, FcγR2A, and FcγR2B. Given the highly correlated nature of the vaccine induced humoral immune response, a correlation network analysis was built between LASSO-selected features and the overall immune response to define the additional features that may shift differentially across the vaccine profiles ( Figure.1E ). Three clusters appeared, bearing elevated IgA response across all antigenic determinants, a small network of IgG2 responses, and a large network of Fc-receptor binding antibody responses across multiple antigenic targets all enriched among mRNA-1273 immunized individuals. These data point to robust humoral immune responses induced by both mRNA platforms, but that do diverge with enhanced epitope spreading, IgA immunity, and specific antibody effector functions in mRNA-1273 immunized individuals. mRNA-1273 and BNT162b2 vaccines induce FcR-binding responses to multiple VOCs. Despite the remarkable efficacy of the mRNA vaccines against the original SARS-CoV-2 variant, waves of variants have emerged that include amino acid substitutions that significantly diminish neutralizing antibody activity 26 – 28 . Among the variants of concern (VOCs), the mRNA vaccines appear to neutralize Alpha (B.1.1.7) 29 and Gamma (P.1) with only a minimal loss of activity, but exhibit compromised neutralizing activity against the beta (B.1.351) variant 30 31 . Yet, whether Fc responses were equally affected across the VOCs remains unclear. Both mRNA-1273 and BNT162b2 vaccine-induced antibodies bound equally well across the Alpha B.1.1.7, Beta B.1.351, and Gamma P.1 VOCs ( Figure.2A ). Interestingly, IgM titers were higher in BNT162b2 vaccinated individuals to the Beta and Gamma VOCs. Additionally, a trend towards higher IgG1-binding titers was also noted in the BNT162b2 immunized individuals across all 3 VOCs. Conversely, IgA responses were amplified in the mRNA-1273 immunized individuals ( Figure. 2A ). However, Fc receptor-binding antibodies were induced by both vaccines to all 3 VOCs at equivalent levels. Along the same lines, both antibody-dependent monocyte (ADCP) and neutrophil (ADNP) phagocytosis were largely equivalent across the variants ( Figure.2B ), highlighting the robust Fc binding and functional profiles across VOCs elicited by both mRNA platforms. Yet, despite these univariate results, we finally aimed to ask if any multivariate differences could be observed across the two mRNA platforms in their VOC response ( Figure.2C ). The LASSO/PLSDA revealed separation in the Fc-profiles induced to the VOCs between the mRNA-1273 and BNT162b vaccinated individuals ( Figure.2C ). The profile was marked by higher IgM-Beta levels in BNT162b vaccinated individuals. Conversely, higher levels of IgA/IgG2 responses were observed to alpha in mRNA-1273 vaccinated individuals who also had higher levels of FCR2A and FCR2B NTD-binding antibodies. The extended LASSO co-correlate network further highlighted the presence of IgG2, IgA, and IgM-only networks across multiple VOCs, suggesting that isotype biased selection across the mRNA platform includes reactivities across VOCs. Additionally, a large network of highly functional pan VOC and epitope responses were observed in the mRNA1273 profile, marked by an enrichment of NTD-specific antibody responses. Thus, the two mRNA platforms elicit overall similar levels of functional antibodies to the VOCs, with an IgM/IgG biased profile induced by BNT162b2 and a more class switched IgA/IgG-driven profile induced by mRNA-1273. mRNA-1273 and BNT162b2 vaccines induce more robust RBD and full Spike VOC-targeting Fc-functional antibodies compared to natural infection Alarmingly, several variants, including the Alpha (B.1.1.7), Beta (B.1.351), Gamma (P1), and Delta (B.1.617.2) VOCs, have begun to breakthrough natural 32 and vaccine induced immune responses 26 , causing large numbers of outbreaks. The unexpected significant reduction of effectiveness of mRNA vaccines against VOCs, especially delta (B.1617.2) variant, is emerging 20 , albeit with the majority of breakthroughs remaining largely non-lethal 1 . Yet, differential real-world efficacy against the Delta VOC 20 points to a nuanced immune response to delta. Thus, we next aimed to compare the cross-VOC antibody Fc-profiles targeting both VOC RBDs or full Spike antigens across a subset of the vaccinees and a group of mild-community acquired convalescent individuals. Antibody profiles were compared across mRNA-1273 ( n = 16 ) and BNT162b ( n = 15 ) vaccines and 10 convalescents ( Figure.3 ). mRNA vaccine induced IgG1 responses were higher than convalescent responses to the original variant RBD (WT), Alpha (B.1.1.7), Beta (B.1.351), Kappa (B.1.617.1), and Delta (B.1.617.2), but were lower for the gamma (P.1) RBD. Similar patterns were observed across all Fc-receptor RBD-binding antibodies induced by the mRNA vaccines, that all showed superior binding to FcRs compared to convalescent antibodies that bound poorly to all RBD VOCs compared to the WT variant ( Figure.3A ). Interestingly, slightly higher antibody binding was noted across RBD VOCs for mRNA-1273 immunized individuals compared to BNT162b2, albeit the pattern of recognition was the same. Conversely, IgG1 and IgG3 binding patterns to S-specific antibodies revealed enhanced mRNA-spike specific binding to nearly all VOC-full Spike antigens, except the kappa variant, compared to WT Spike-specific antibody binding ( Figure.3B ). Importantly, all Spike-specific binding IgG responses were lower in convalescents compared to mRNA vaccinees, with slightly higher IgG1 and IgG3 binding noted to nearly all VOC Spikes. In contrast, Fc-receptor binding antibodies exhibited equivalent VOC-recognition across both mRNA vaccine platforms, highlighting potential functional differences in RBD- and Spike-specific antibodies induced across the platforms. Thus, despite the more variable FcR-binding profiles to RBD VOCs, stable FcRbinding was noted to most full spike VOCs. Given the persistent protection against delta in newly/recently vaccinated individuals 1 , but enhanced breakthrough over time, these data may suggest that the presence of broad non-RBD-specific functional immunity may be key to protection, given that VOC-RBDs clearly are able to exploit RBD-specific Fc-vulnerabilities even across vaccine platforms. RBD-specific antibody depletion influences antibody-mediated monocyte and neutrophil phagocytosis The data above suggested that potential differences in RBD and Spike specific contributions to polyclonal antibody Fc-binding profiles and function. Thus, to address this possibility, RBD-specific antibodies were depleted from the polyclonal serum of our vaccinees and convalescent samples ( Figure. 4 ) and tested for opsonophagocytic functions linked to the natural resolution of infection 33 , 34 . Specifically, monocyte phagocytosis (ADCP, Figure. 4A ) or neutrophil phagocytosis (ADNP, Figure. 4B ) were evaluated across VOCs. RBD depletion resulted in a significant loss of ADCP against the WT variant Spike ( Figure. 4A , bottom left). However, RBD depletion did not affect the ADCP response in convalescents to all other Spike VOCs, arguing that antibodies in natural infection that drive function largely target areas outside of the RBD. Conversely, RBD depletion resulted in slightly reduced ADCP in BNT162b2 immunized individuals to the WT, beta, and epsilon variants, but not to the alpha and gamma variants. Similarly, RBD-depletion resulted in reduced Beta and Gamma ADCP in mRNA-1273 immunized individuals. These data argue for variable, but low-level alterations in RBD-specific ADCP activity across VOC Spikes following natural infection, and differentially following mRNA vaccination, suggesting that non-RBD-specific antibodies may continue to drive opsonophagocytic control of the virus even in the setting of profound changes in the RBD that may knock out neutralization. In contrast to ADCP, more variation was observed in neutrophil phagocytosis (ADNP) with RBD depletion. Interestingly, RBD depletion led to increased ADNP against the WT and Epsilon variants, suggesting that RBD-specific antibodies may, in fact, block this activity following the natural resolution of disease. Conversely, RBD depletion led to reduced ADNP to the alpha, beta, epsilon, and iota variants in BNT162b2 immunized individuals. Similarly, RBD depletion reduced ADNP to WT, alpha, beta, gamma, and iota in mRNA-1273 vaccinated individuals. Interestingly, the RBD depletion led to a more profound loss of ADNP in mRNA1273 immunized individuals compared to BNT162b2 immunized individuals, marking significant differences in the functional activity of particular sub-populations of antibodies elicited by these vaccines. Thus, given the higher levels of mRNA-1273 induced pan-VOC RBD specific immunity, that appears to contribute largely to ADNP activity, these data point to the possible existence of different epitope-specific functional correlates of immunity elicited across the two vaccine platforms and natural infection. Thus, these data collectively show the robust induction of functional antibody responses that differ in their Fc-biology, following mRNA-1273 and BNT162b2 vaccination, marked by differences in the overall isotype/subclass, Fc-receptor binding profiles, and epitope specific functions across VOCs, providing some potential explanation for differences in persistent protection afforded against by this newly emerging vaccine platform. Discussion: Despite the remarkable protective immunity observed in the BNT162b2 and mRNA-1273 vaccine phase 3 studies against the original SARS-CoV-2 variant, breakthrough infections are on the rise globally among vaccinees 35 . Yet, although the rise of breakthrough infections, severe disease, hospitalization, and death remain low in most populations, fatality rates have begun to emerge in the elderly, prompting discussions on additional vaccine boost 36 . While both mRNA COVID-19 vaccines induced comparable and robust antibody titers and neutralization, emerging data point to variable levels of real-world vaccine efficacy across the platforms potentially linked to differences in the formulation, design, boosting intervals, and dose 20 . Thus, while both vaccines induced robust antibody titers and neutralization, understanding immunological differences across the vaccines may provide critical insights on immune correlates of protection to guide next generation vaccine design and to guide boosting. Among the proposed non-neutralizing antibody immune mechanisms of protection, T cells have been proposed as a critical arm in the control of SARS-CoV-2 due to their known role in the battle against many viruses 37 , 38 . Yet, additional mechanisms, such as the role of antibody-mediated effector function, have also been shown to play a critical role in vaccine-mediated protection against a broad array of pathogens 7 , including SARS-CoV-2 39 . Thus, here we deeply probed the functional humoral immune response induced by distinct mRNA vaccine platforms and probed their Fc-functional performance across VOCs, demonstrating robust Fc-functional responses, albeit distinct, induced by both the BNT162b2 and mRNA-1273 vaccines, that elicit Fc-effector functions against most VOCs, including the beta and delta variants, despite the documented loss of neutralizing activity. These data point to a potential role for vaccine induced Fc-effector function in mRNA vaccine induced protection against disease mediated by VOCs, but also to nuances in the functional response to VOCs that may contribute to real world efficacy differences across the platforms. While no difference in neutralizing activity has been reported across the BNT162b2 and mRNA1273 vaccines 24 , some differences in isotype/subclass and Fc-functions were noted across the two platforms. Consistent with previous observations in pregnant women 25 , elevated levels of IgA were noted following mRNA-1273, accompanied by higher levels of ADNP and NK cell activation. Conversely, an IgM/IgG bias was noted in BNT162b2 vaccinated individuals to VOCs, pointing to differences in class switching across the mRNA platforms. Whether these differences contribute to different efficacy, particularly over time as the response wanes, remains unclear but will be addressed in long-term follow-up breakthrough studies. Moreover, whether these differences are related to differences in lipid nanoparticle composition, mRNA dose, and delay in boosting remains unclear, but highlights the potential of mRNA vaccines to drive "tunable" Fc effector function, that may be selectively shaped to achieve enhanced and selective control over particular target pathogens and non-infectious diseases in the future. The majority of mutations in the VOCs occur in N-terminal and RBD domains 40 , which play a critical role in enhancing binding to ACE2. Given that neutralizing antibodies target these same sites, aimed at occluding ACE2-access, or aimed at compromising RBD-structure to prevent ACE2-interactions, these same mutations inadvertently compromise neutralizing antibody activity. Conversely, Fc-functional antibodies can target the whole surface of the Spike antigen, and thus are not compromised in the same manner as neutralizing antibodies by individual or clusters of mutations found to alter ACE2-binding. In fact, while a large part of neutralizing antibodies targets the RBD, the RBD depletion did not knock out Fc-effector function in convalescent individuals or BNT162b2 and mRNA-1273. Instead, vaccine induced RBD-specific antibodies contributed minimally to ADCP mediated activity across the VOCs, however RBD-specific antibodies contributed more to ADNP activity across the VOCs, suggesting that epitope-specific functional programming likely occurs across the mRNA vaccines, with a more prominent focus of ADNP on the RBD in mRNA-1273 immunized individuals. Moreover, depletion of RBD-specific antibodies from convalescent plasma resulted in improved ADNP to some variant Spike antigens, suggesting that RBD-specific antibodies may block Fc-effector function in natural infection, either by blocking access of additional functional antibodies, or due to altered or perturbed Fc glycosylation induced in natural infection. Thus, these data point to the vaccine and infection-induced differences in Fc-programming at an epitope-specific level, which may play a critical part in the level of protection against disease severity across naturally immune and vaccinated populations that have been observed across breakthrough cases globally. Thus, understanding the relationship between non-neutralizing antibody functions, epitope specificity, and clinical protection may provide unexpected insights to plan for rational boosting efforts to stop the continuous emergence of VOC. The spread of the delta VOC has raised concerns globally about vaccine efficacy and the need for boosting. While hospitalizations remain consistently high in regions of the globe where vaccine deployment has been slow, most infections in previously vaccinated individuals do not require hospitalization. These data suggest that while both naturally and vaccine induced antibodies gradually lose the capability of preventing transmission against VOCs, vaccine antibodies may still provide some barrier to severity of disease. The data here demonstrate the ability of the BNT162b2 and mRNA-1273 vaccines to induce robust Fc-effector functions, previously linked to the resolution of severe disease in unvaccinated individuals 34 , 41 , 42 , across many VOCs, despite the reported loss of neutralization. Thus, despite increased breakthroughs in BNT162b2, the quality of the recall, an anamnestic response may be sufficient to respond to SARS-CoV-2 and drive control and clearance of infection. However, understanding the differences in transmission blocking activity across the BNT162b2 and mRNA-1273 may provide new clues for the redesign of vaccines able to provide a longer barrier of protection against the virus, which may be required to slow the rates of evolution of variants of concern. Linked to our emerging appreciation for the role of Fc-effector function in protection from infection/disease in non-human primates 13 , 43 , hamsters 14 , and mice 44 , upcoming immune correlates analyses and breakthrough studies will provide a concrete recognition for the role of Fc-effector function in protection against SARS-CoV-2. Yet, ultimately, the data here point to the potential nuanced differences in the quality of the humoral immune response induced across the two authorized mRNA vaccine technologies, able to broadly harness multiple Fc-effector functions, and tune these functions differentially to distinct epitopes and innate immune cell types depending on dose, formulation, or booster-timing. Competing interests: G.A. is a founder and equity holder for Seromyx Systems Inc., an employee and equity holder for Leyden Labs, and has received financial support from Abbvie, BioNtech, GSK, Gilead, Merck, Moderna, Novartis, Pfizer, and Sanofi. D.D., J.R., A.S.M, and E.R.M. are employees of Space Exploration Technologies Corp. All other authors have declared that no conflict of interest exists.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6873190/

The transcription factor PU.1 mediates enhancer-promoter looping that is required for IL-1β eRNA and mRNA transcription in mouse melanoma and macrophage cell lines

The DNA-binding protein PU.1 is a myeloid lineage–determining and pioneering transcription factor due to its ability to bind "closed" genomic sites and maintain "open" chromatin state for myeloid lineage–specific genes. The precise mechanism of PU.1 in cell type–specific programming is yet to be elucidated. The melanoma cell line B16BL6, although it is nonmyeloid lineage, expressed Toll-like receptors and activated the transcription factor NF-κB upon stimulation by the bacterial cell wall component lipopolysaccharide. However, it did not produce cytokines, such as IL-1β mRNA. Ectopic PU.1 expression induced remodeling of a novel distal enhancer (located ∼10 kbp upstream of the IL-1β transcription start site), marked by nucleosome depletion, enhancer-promoter looping, and histone H3 lysine 27 acetylation (H3K27ac). PU.1 induced enhancer-promoter looping and H3K27ac through two distinct PU.1 regions. These PU.1-dependent events were independently required for subsequent signal-dependent and co-dependent events: NF-κB recruitment and further H3K27ac, both of which were required for enhancer RNA (eRNA) transcription. In murine macrophage RAW264.7 cells, these PU.1-dependent events were constitutively established and readily expressed eRNA and subsequently IL-1β mRNA by lipopolysaccharide stimulation. In summary, this study showed a sequence of epigenetic events in programming IL-1β transcription by the distal enhancer priming and eRNA production mediated by PU.1 and the signal-dependent transcription factor NF-κB.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2721800/

Host-Derived Tumor Endothelial Marker 8 (TEM8) Promotes the Growth of Melanoma

TEM8 was initially identified as a gene overexpressed in the vasculature of human tumors and was subsequently identified as an anthrax toxin receptor. To assess the functional role of TEM8, we disrupted the TEM8 gene in mice by targeted homologous recombination. TEM8 −/− mice were viable and reached adulthood without defects in physiological angiogenesis. However, histopathologic analysis revealed an excess of extracellular matrix (ECM) in several tissues including the ovaries, uterus, skin, and periodontal ligament of the incisors, the latter resulting in dental dysplasia. When challenged with B16 melanoma, tumor growth was delayed in TEM8 −/− mice while the growth of other tumors, such as Lewis lung carcinoma, was unaltered. These studies demonstrate that host-derived TEM8 promotes the growth of certain tumors and suggest that TEM8 antagonists may have utility in the development of new anti-cancer therapies.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4309704/

Global Transcriptional Response of Clostridium difficile Carrying the ϕCD38-2 Prophage

Clostridium difficile is one of the most dangerous pathogens in hospital settings. Most strains of C. difficile carry one or more prophages, and some of them, like ϕCD38-2 and ϕCD119, can influence the expression of toxin genes. However, little is known about the global host response in the presence of a given prophage. In order to fill this knowledge gap, we used high-throughput RNA sequencing (RNA-seq) to conduct a genome-wide transcriptomic analysis of the epidemic C. difficile strain R20291 carrying the ϕCD38-2 prophage. A total of 39 bacterial genes were differentially expressed in the R20291 lysogen, 26 of them being downregulated. Several of the regulated genes encode transcriptional regulators and phosphotransferase system (PTS) subunits involved in glucose, fructose, and glucitol/sorbitol uptake and metabolism. ϕCD38-2 also upregulated the expression of a group of regulatory genes located in phi-027, a resident prophage common to most ribotype 027 isolates. The most differentially expressed gene was that encoding the conserved phase-variable cell wall protein CwpV, which was upregulated ∼20-fold in the lysogen. Quantitative PCR and immunofluorescence showed that the increased cwpV expression results from a greater proportion of cells actively transcribing the gene. Indeed, ∼95% of lysogenic cells express cwpV , as opposed to only ∼5% of wild-type cells. Furthermore, the higher proportion of cells expressing cwpV results from a higher frequency of recombination of the genetic switch controlling phase variation, which we confirmed to be dependent on the host-encoded recombinase RecV. In summary, ϕCD38-2 interferes with phase variation of the surface protein CwpV and the expression of metabolic genes.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5554066/

Aortic Adventitial Fibroblast Sensitivity to Mitogen Activated Protein Kinase Inhibitors Depends on Substrate Stiffness

Adventitial fibroblasts (AFs) are key determinants of arterial function and critical mediators of arterial disease progression. The effects of altered stiffness, particularly those observed across individuals during normal vascular function, and the mechanisms by which AFs respond to altered stiffness, are not well understood. To study the effects of matrix stiffness on AF phenotype, cytokine production, and the regulatory pathways utilized to interpret basic cell-matrix interactions, human aortic AFs were grown in 5%, 7.5%, and 10% (w/v%) PEG-based hydrogels with Young's moduli of 1.2, 3.3, and 9.6 kPa, respectively. In 5% gels, AFs had higher proliferation rates, elevated monocyte chemoattractant protein-1 secretion, and enhanced monocyte recruitment. Significantly more AFs were α-smooth muscle actin positive in 7.5% gels, indicating myofibroblast development. AFs in 10% gels had low proliferation rates but produced high levels of interleukin-6 and vascular endothelial growth factor-A. Importantly, these modulus-dependent changes in AF phenotype were accompanied by alterations in the mitogen-activated protein kinase (MAPK) pathways contributing to the production of cytokines. These data indicate that complex cell regulatory changes occur with altered tissue stiffness and suggest that therapeutics affecting MAPK pathways may have altered effects on AFs depending on substrate stiffness.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6243247/

Caspase-8 induces cleavage of gasdermin D to elicit pyroptosis during Yersinia infection

Significance Here we demonstrate that Yersinia YopJ-induced murine macrophage death involves caspase-8–induced cleavage of both gasdermin D (GSDMD) and gasdermin E (GSDME). The ensuing cell death is rapid, morphologically is similar to pyroptosis, and induces IL-1 release. Recently, both GSDMD and GSDME were reported to be critical effectors of caspase-1/11–driven pyroptosis and caspase-3–dependent secondary necrosis, which prompted the redefinition of pyroptosis as cell death-mediated by gasdermin activation. Our work extends these studies and shows that activation of caspase-8 in the context of TAK1 inhibition results in cleavage of both GSDMD and GSDME, leading to pyroptotic-like cell death. Further study will be needed to determine whether caspase-8 cleaves GSDMD directly or via intermediate substrates.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8403129/

Spatial regulation of protein A in Staphylococcus aureus

Summary Surface proteins of Staphylococcus aureus play vital roles in bacterial physiology and pathogenesis. Recent work suggests that surface proteins are spatially regulated by a YSIRK/GXXS signal peptide that promotes cross-wall targeting at the mid-cell, though the mechanisms remain unclear. We previously showed that protein A (SpA), a YSIRK/GXXS protein and key staphylococcal virulence factor, mis-localizes in a ltaS mutant deficient in lipoteichoic acid (LTA) production. Here, we identified that SpA contains another cross-wall targeting signal, the LysM domain, which, in addition to the YSIRK/GXXS signal peptide, significantly enhances SpA cross-wall targeting. We show that LTA synthesis, but not LtaS, is required for SpA septal anchoring and cross-wall deposition. Interestingly, LTA is predominantly found at the peripheral cell membrane and is diminished at the septum of dividing staphylococcal cells, suggesting a restriction mechanism for SpA septal localization. Finally, we show that D-alanylation of LTA abolishes SpA cross-wall deposition by disrupting SpA distribution in the peptidoglycan layer without altering SpA septal anchoring. Our study reveals that multiple factors contribute to the spatial regulation and cross-wall targeting of SpA via different mechanisms, which coordinately ensures efficient incorporation of surface proteins into the growing peptidoglycan during the cell cycle.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4740342/

Whole Genome Analysis of Injectional Anthrax Identifies Two Disease Clusters Spanning More Than 13 Years

Background Anthrax is a rare disease in humans but elicits great public fear because of its past use as an agent of bioterrorism. Injectional anthrax has been occurring sporadically for more than ten years in heroin consumers across multiple European countries and this outbreak has been difficult to trace back to a source. Methods We took a molecular epidemiological approach in understanding this disease outbreak, including whole genome sequencing of Bacillus anthracis isolates from the anthrax victims. We also screened two large strain repositories for closely related strains to provide context to the outbreak. Findings Analyzing 60 Bacillus anthracis isolates associated with injectional anthrax cases and closely related reference strains, we identified 1071 Single Nucleotide Polymorphisms (SNPs). The synapomorphic SNPs (350) were used to reconstruct phylogenetic relationships, infer likely epidemiological sources and explore the dynamics of evolving pathogen populations. Injectional anthrax genomes separated into two tight clusters: one group was exclusively associated with the 2009–10 outbreak and located primarily in Scotland, whereas the second comprised more recent (2012–13) cases but also a single Norwegian case from 2000. Interpretation Genome-based differentiation of injectional anthrax isolates argues for at least two separate disease events spanning > 12 years. The genomic similarity of the two clusters makes it likely that they are caused by separate contamination events originating from the same geographic region and perhaps the same site of drug manufacturing or processing. Pathogen diversity within single patients challenges assumptions concerning population dynamics of infecting B. anthracis and host defensive barriers for injectional anthrax. Funding This work was supported by the United States Department of Homeland Security grant no. HSHQDC-10-C-00,139 and via a binational cooperative agreement between the United States Government and the Government of Germany. This work was supported by funds from the German Ministry of Defense (Sonderforschungsprojekt 25Z1-S-431,214). Support for sequencing was also obtained from Illumina, Inc. These sources had no role in the data generation or interpretation, and had not role in the manuscript preparation. Panel 1: Research in Context Systematic Review We searched PubMed for any article published before Jun. 17, 2015, with the terms " Bacillus anthracis " and "heroin", or "injectional anthrax". Other than our previously published work ( Price et al., 2012 ), we found no other relevant studies on elucidating the global phylogenetic relationships of B. anthracis strains associated with injectional anthrax caused by recreational heroin consumption of spore-contaminated drug. There were, however, publically available genome sequences of two strains involved ( Price et al., 2012 , Grunow et al., 2013 ) and the draft genome sequence of Bacillus anthracis UR-1, isolated from a German heroin user ( Ruckert et al., 2012 ) with only limited information on the genotyping of closely related strains ( Price et al., 2012 , Grunow et al., 2013 ). Lay Person Interpretation Injectional anthrax has been plaguing heroin drug users across Europe for more than 10 years. In order to better understand this outbreak, we assessed genomic relationships of all available injectional anthrax strains from four countries spanning a > 12 year period. Very few differences were identified using genome-based analysis, but these differentiated the isolates into two distinct clusters. This strongly supports a hypothesis of at least two separate anthrax spore contamination events perhaps during the drug production processes. Identification of two events would not have been possible from standard epidemiological analysis. These comprehensive data will be invaluable for classifying future injectional anthrax isolates and for future geographic attribution. Background Anthrax is a rare disease in humans but elicits great public fear because of its past use as an agent of bioterrorism. Injectional anthrax has been occurring sporadically for more than ten years in heroin consumers across multiple European countries and this outbreak has been difficult to trace back to a source. Methods We took a molecular epidemiological approach in understanding this disease outbreak, including whole genome sequencing of Bacillus anthracis isolates from the anthrax victims. We also screened two large strain repositories for closely related strains to provide context to the outbreak. Findings Analyzing 60 Bacillus anthracis isolates associated with injectional anthrax cases and closely related reference strains, we identified 1071 Single Nucleotide Polymorphisms (SNPs). The synapomorphic SNPs (350) were used to reconstruct phylogenetic relationships, infer likely epidemiological sources and explore the dynamics of evolving pathogen populations. Injectional anthrax genomes separated into two tight clusters: one group was exclusively associated with the 2009–10 outbreak and located primarily in Scotland, whereas the second comprised more recent (2012–13) cases but also a single Norwegian case from 2000. Interpretation Genome-based differentiation of injectional anthrax isolates argues for at least two separate disease events spanning > 12 years. The genomic similarity of the two clusters makes it likely that they are caused by separate contamination events originating from the same geographic region and perhaps the same site of drug manufacturing or processing. Pathogen diversity within single patients challenges assumptions concerning population dynamics of infecting B. anthracis and host defensive barriers for injectional anthrax. Funding This work was supported by the United States Department of Homeland Security grant no. HSHQDC-10-C-00,139 and via a binational cooperative agreement between the United States Government and the Government of Germany. This work was supported by funds from the German Ministry of Defense (Sonderforschungsprojekt 25Z1-S-431,214). Support for sequencing was also obtained from Illumina, Inc. These sources had no role in the data generation or interpretation, and had not role in the manuscript preparation. Panel 1: Research in Context Systematic Review We searched PubMed for any article published before Jun. 17, 2015, with the terms " Bacillus anthracis " and "heroin", or "injectional anthrax". Other than our previously published work ( Price et al., 2012 ), we found no other relevant studies on elucidating the global phylogenetic relationships of B. anthracis strains associated with injectional anthrax caused by recreational heroin consumption of spore-contaminated drug. There were, however, publically available genome sequences of two strains involved ( Price et al., 2012 , Grunow et al., 2013 ) and the draft genome sequence of Bacillus anthracis UR-1, isolated from a German heroin user ( Ruckert et al., 2012 ) with only limited information on the genotyping of closely related strains ( Price et al., 2012 , Grunow et al., 2013 ). Lay Person Interpretation Injectional anthrax has been plaguing heroin drug users across Europe for more than 10 years. In order to better understand this outbreak, we assessed genomic relationships of all available injectional anthrax strains from four countries spanning a > 12 year period. Very few differences were identified using genome-based analysis, but these differentiated the isolates into two distinct clusters. This strongly supports a hypothesis of at least two separate anthrax spore contamination events perhaps during the drug production processes. Identification of two events would not have been possible from standard epidemiological analysis. These comprehensive data will be invaluable for classifying future injectional anthrax isolates and for future geographic attribution. 1 Introduction Anthrax is a common disease in livestock and usually only infects humans that come into contact with infected animals or their products ( Mock & Fouet, 2001 ). High hygienic standards, knowledge of the disease, and effective surveillance make human anthrax infections extremely rare in industrialized countries. Most commonly, human anthrax presents as a cutaneous lesion due to spore deposition from handling contaminated animal hair, meat, or skin. The three other recognized anthrax types are gastrointestinal, inhalational, and injectional anthrax, which have unique clinical presentations dependent on the infective route. Injectional anthrax (IA), which was only recently recognized, is associated with recreational drug use when Bacillus anthracis spores have contaminated heroin ( Berger et al., 2014 ). Injectional drug consumers may have preexisting health issues that exacerbate their sensitivity to anthrax and victims may be reluctant to seek medical attention. The presentation of IA can be particularly gruesome with extensive tissue necrosis emanating from the injection site ( Booth et al., 2010 , Russell et al., 2013 ). While the first case of IA was described in 2000 from Norway ( Ringertz et al., 2000 ), it is remarkable that no additional cases of IA were documented over the following 10 years. The largest subsequent cluster of IA cases occurred in 2009–10 and predominantly centered in Scotland ( Team, 2011 ). Other more sporadic cases have been observed in England, Wales, Scotland, Germany, Denmark, and France in 2012–13 ( Hanczaruk et al., 2014 ). It is clear that not all drug users consuming a particular batch of heroin have contracted anthrax and attempts to directly detect or isolate B. anthracis from heroin have failed ( Team, 2011 ). How and why heroin was contaminated with B. anthracis spores has been difficult to address due to the illicit nature of the drug, but likely occurred during its distribution. B. anthracis evolved as a clonal lineage from a B. cereus -like ancestor with increased pathogenicity due to the acquisition of two virulence plasmids and chromosomal mutations ( Mock and Fouet, 2001 , Okinaka et al., 1999 ). The clonal nature of B. anthracis and, perhaps, slow evolutionary rates have resulted in a highly homogeneous species with little strain-to-strain diversity. Nevertheless, Single Nucleotide Polymorphism (SNP) variation has been readily discovered through the use of whole genome sequencing and the population structure has been defined to a high level at both a local and global scale ( Pearson et al., 2004 , Van Ert et al., 2007 , Read et al., 2002 ). In a previous analysis of the Scottish IA outbreak, a single whole genome sequence was used to identify clade-specific SNPs ( Price et al., 2012 ). Diagnostic assays targeting some of these SNPs clustered all IA specimens, including those from 2012 ( Price et al., 2012 , Grunow et al., 2013 ), into a single subtype (defined by SNP1173928 and SNP1053700) nested within a clade that contained two closely affiliated strains from infected humans (Turkey) ( Price et al., 2012 ). Given a suspected origin in Afghanistan, a major producer of heroin destined for Western Europe, these results led to the conclusion that the heroin may have been contaminated along transportation routes through the Middle East and not at the production source. Importantly, it was discovered that European IA cases could be differentiated by a VNTR allelic difference, raising questions about the epidemiological significance of the differentiation. However, as only IA strains appeared to be in a common clade within the comparison group, a single continuing contamination event and persistence of contaminated heroin circulating in Europe provided the most parsimonious description of this outbreak ( Hanczaruk et al., 2014 ). Detailed knowledge of heroin production practices, transportation, processing, and potential sources of spore contamination is lacking, which makes understanding and controlling IA challenging. Here, we greatly expand upon our knowledge on the genetic make-up and relationships of IA-associated B. anthracis by analyzing whole genome sequences (WGS) from 58 newly sequenced isolates and two published isolates from 36 different IA patients and seven WGS from phylogenetically affiliated strains. We find that IA cases from the years 2000 to 2013 can be clustered into two distinct genetic groups that overlap in time and space, suggesting distinct contamination events, albeit with very similar but unique strains of B. anthracis that likely originated from a common geographic region. 2 Methods B. anthracis strains . We analyzed the genomes of 67 B. anthracis isolates, including 60 associated with IA cases. Seven additional strains were not from IA cases but were phylogenetically affiliated. The more distantly related Ames Ancestor is a very high quality finished genome and was used as a reference. These strains are listed in Table S1 along with their original sources and NCBI accession numbers for each WGS. All isolates represented residual material from clinical diagnostic procedures and were de-identified. As such, they are exempt from the human subjects requirements for patient enrollment. 2.1 DNA Isolation and WGS Generation B. anthracis isolates from IA cases and the reference strain collections were cultured on blood agar plates (bioMérieux) and incubated for 16–24 h at 37 °C and then used for DNA purification. DNA was harvested from single colonies, either directly or following liquid media culturing, using the QIAamp UCP Pathogen Mini Kit or Blood and Tissue kit (Qiagen, Valencia, CA) with an enzymatic (lysozyme) or mechanical pre-lysis step. DNA library preparations for WGS were performed using standard methods and performed on an Illumina MiSeq instrument (see Supplemental Methods for details). 2.2 SNP Analysis of WGS Reference genome assemblies were downloaded from GenBank ( Benson et al., 2012 ). Paired-end reads were aligned against the Ames Ancestor genome (NC_007530) using BWA-MEM ( Li, 2013 ) and SNPs were called with the UnifiedGenotyper method in GATK ( DePristo et al., 2011 , McKenna et al., 2010 ); any SNPs that had a coverage less than 6X depth and majority allele frequencies  13 years is astounding and difficult to reconcile with a single contamination event. It has been argued that heroin supply and inventories would not last for 13 years, especially given the apparent 10-year lapse in G-II IA cases. We therefore suggest that there is a single contaminating source for the G-II cases but that there have been two or more (unobserved) contamination events.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3061125/

Electroporation and lysis of marine microalga Karenia brevis for RNA extraction and amplification

We describe here a simple device for dielectrophoretic concentration of marine microalga Karenia brevis non-motile cells, followed by electric field-mediated lysis for RNA extraction. The lysate was purified using magnetic beads and pure RNA extracted. RNA quality was assessed off-chip by nucleic acid sequence-based amplification and the optimum conditions for lysis were determined. This procedure will form part of an integrated microfluidic system that is being developed with sub-systems for performing cell concentration and lysis, RNA extraction/purification and real-time quantitative RNA detection. The integrated system and its components could be used for a large range of applications including in situ harmful algal bloom detection, transcriptomics and point-of-care diagnostics.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8546559/

Technology Advances, High-Risk Research, and a Safe Way Forward

ABSTRACT The human and economic toll of the coronavirus disease 2019 (COVID-19) pandemic and the unknowns regarding the origins of the virus, with a backdrop of enormous advances in technologies and human understanding of molecular virology, have raised global concerns about the safety of the legitimate infectious disease research enterprise. We acknowledge the safety and security risks resulting from the broad availability of tools and knowledge, tools and knowledge that can be exploited equally for good or harm. The last 2 decades have shown us that the risks are real. They have also shown us that more traditional top-down regulations alone are not the answer. We encourage government to be thoughtful and nuanced in dealing with this significant challenge and to carefully consider human factors and the important role of organizational-level leadership before simply layering an additional bureaucratic burden on the enterprise without understanding value and cost. COMMENTARY Uncertainties still surround the origins of the coronavirus disease 2019 (COVID-19) pandemic, which is believed to have sickened more than 200 million, killed more than 4 million, and terribly disrupted economies worldwide. The pandemic continues to rage, now with variant viruses and in third and fourth waves. A mix of responsible and irresponsible journalism and political bias confuse and even divide populations with simplistic explanations focusing on "lab leak" or "jump from nature." Some suggest the possibility of laboratory manipulations of the virus. One of the expected calls, among the many, is for more regulation of our own infectious disease life sciences enterprise. We in the United States have a multiple-decade history of high-containment infectious disease research without catastrophic outcomes. We have responded to several minor and even potential safety and security incidents—resulting from carelessness, poor judgment, and even crime—with crude bureaucratic regulatory tools that often negatively impact the vast majority who are careful scientists. We do not have good metrics for regulatory effectiveness; however, the absence of laboratory-acquired disease and the lack of serious breaches of biocontainment attest to the inherent safety of modern U.S. high-containment facilities. Not surprisingly, the advanced knowledge and technologies available in our laboratories today are equally powerful in the hands of "good" and "bad" scientists, competent and incompetent. We do not doubt the potential for great harm. Thus, we need to be agile in adapting to new potential risks associated with technological advances while maintaining our ability to make the scientific progress that clearly benefits society. It is not that we have not considered surprises from nature and even from our labs. We adopted the Select Agent Rule nationally in 1996 and expanded it in 2002 all in response to "security" incidents. When there were few new security incidents, we made lab safety the focus. In 2004, after 9/11 and the "anthrax letters" wake-up calls, the National Academies of Sciences, Engineering, and Medicine (NASEM) published the "Fink report," Biotechnology Research in an Age of Terrorism ( 1 ), which gave us common-sense examples of the kinds of harm that could come from our research enterprise if we are not watchful and careful. While an important first step in developing our thinking, it took us down the road of "dual-use research of concern," highlighting the technologies more than the human behavior, typically the cause of "misuse" when it does occur. The Fink report recommended that a national board be established which would serve as an advisory body on these matters. In 2005, the National Science Advisory Board for Biosecurity (NSABB) was put in place and quickly dealt with security questions related to the publication of the sequence of the recovered 1918 pandemic flu strain. In 2007, the NSABB Framework for Oversight of Dual Use Life Science Research ( 2 ) established guidelines for evaluating potentially dangerous research "from hypothesis to publication." In 2008, the Department of Army borrowed from the nuclear enterprise to establish a Surety Program (Safety, Security, Agent accountability and Personnel reliability), some of which is ill-suited to biology. Also, in 2008/2009, four national committees ( 3 ) considered how to deal with the "insider threat," a similar challenge to the life sciences community. Just a few years later, the NIH funded two scientists who successfully (and safely) engineered H5N1 flu virus, making it aerosol transmissible in ferrets; this was high-risk research. The NSABB and international colleagues helped journals deal with the issue of publication of those potentially abusable data, eventually voting to allow it. (Even that exercise was somewhat artificial, because findings from academic research done in a free society like ours cannot really be "put back in the bottle.") In 2016, the NSABB published Recommendations for the Evaluation and Oversight of Proposed Gain-of-Function Research ( 4 ), considering the period from "research design to approval for funding." Then, in 2017, the White House Office of Science and Technology Policy crafted Recommended Policy Guidance for Potential Pandemic Pathogen Care and Oversight ( 5 ) based on the NSABB recommendations and established a Department of Health and Human Services (DHHS) committee to review and manage such potentially risky research. Questions regarding the diligence with which the NIH subsequently applied these new oversight tools and the transparency associated with the DHHS review process exist. To date, much of the effort to minimize the risk of misuse of modern technology and its information product has been undertaken by journals and shared by the researchers themselves. While an important piece of the puzzle's solution, their impact occurs primarily after the research has been done and is further limited to consideration of what is deemed appropriate for sharing by the researchers. While it is important for publishers to take careful note of dual-use research of concern (DURC) issues, their engagement is far too late in the process. The guidelines from the NSABB in 2007 and 2016/2017 make it clear that the early period around hypothesis development is where oversight should begin. That will require organizational leadership with integrity and a strong sense of responsibility, which is then embraced by all involved, from students and technicians to principal investigators (PIs) and program leaders. If done well, many proposals to conduct "dangerous research" will never even make it to a NIH study section. Now, with the COVID-19 pandemic swirling around us all, political scientists, arms control experts, and some biologists, many of whom have never served in a high-containment laboratory, are calling for more regulation of the enterprise. Others just want the U.S. government, or even international bodies, to crack down on what they call "gain-of-function" research. What is needed? What will work? What will be the costs? Some, having observed the bureaucratic implementation of the Select Agent Rule and other regulations, call for nuance by the government. But is the government capable of nuance? We have seen detailed lists of technologies for which there is risk of harm, although the vast majority will be used for good. It is sad that in this era we need to speculate how someone might abuse almost every sector and tool of biotechnology—well beyond the Fink seven deadly sins—for harm. There have even been calls for international regulatory and monitoring mechanisms, to among other things, supervise research on dangerous pathogens. Leaders of high-containment labs have responsibly managed work with dangerous pathogens for decades. Are we now to assume that where evil can be applied, it will be, unless there is top-down government intervention regarding use of these technologies? We do not believe that more regulations will necessarily equate to a safer future. There will always be sound, responsible leadership and healthy organizational cultures in most high-containment laboratories around the globe, whether there is big government oversight or not. The highest containment level, biosafety level 4 (BSL-4) labs, now more than 50 globally, often have better oversight and management than the much more numerous BSL-3 and BSL-2 labs. Nor are we suggesting that the "potential" for harm from nature or the lab is not enormous. The COVID-19 pandemic has demonstrated the tragic consequences of a novel pathogen on society. We only hope that careful thought is given to both the benefits and the potential unintended consequences that might result from additional regulatory interventions applied at this point in our enterprise history. We must think before we regulate and consider cost and benefit. We have previously extolled the virtues of sound leadership at the institute level to deal with traditional laboratory safety and security, including the "insider threat" ( 6 ). We believe that the primary responsibility for conducting safe and secure research with highly hazardous pathogens broadly should be at the organizational level, accepted as a responsibility by the leadership of the institute or laboratory conducting the work. Near the bench rather than the Beltway! History has shown that successful leaders lead with quality science, emphasis on safety, vision, education, responsibility, accountability, honesty, transparency, and ethics. Such leaders invariably have an open-door policy and build networks of regular and open communication within their organization and with other institutes and their scientists. It is not difficult for such an enlightened and selfless leader to develop an institutional culture of accountability and trust based on those basic principles. It has worked in the past; it will work today. Yet our experience suggests it is becoming progressively more difficult to develop, trust and reward enlightened leaders who just "do the right thing." For the most part, there will always be humans with significant understanding of the risks and benefits, and the moral courage, to lead and operate safely in the infectious disease research enterprise no matter where technologies and knowledge take us. But there will also always be some who are careless or thoughtless or those who take excessive risk for their own reasons. Unfortunately, it is apparently easier to add regulations than it is to ensure that humans are responsible for their actions. Much of the cry is for more regulatory oversight at the bench level, the approach we have taken in the past. This risks making legitimate work more difficult and expensive for the vast majority of the enterprise's scientists and leaders who do take full responsibility for their actions and publications. Yet, the specific examples we have of the most high-risk research conducted in recent years suggest that "go-no go" decisions are often made at the government funding agency level. The concept of educating scientists about the importance of individual and corporate responsibility and, particularly, further developing sound leaders and a culture of responsible science and expecting excellence from them, will likely be supported by those at the bench who understand…. but what will happen to implement the call from a few for nuanced oversight? The four studies in 2008, mentioned above, made similar proposals supporting "the importance of leadership, cultures of responsibility, accountability, trust, values and moral obligations." There has been little apparent effort by government to implement those recommendations made more than 10 years ago. Let us hope this time it will be different. New regulations, almost by definition, cost time and money…and can reduce the pool of young scientists interested in the field. Neither regulation nor leadership alone are a 100% solution, but sound leadership "doing the right thing" in our labs costs nothing. Let us do all we can to support a safe, secure, and productive enterprise while not hobbling our researchers unnecessarily with additional burdensome oversight that does little to make society safer but hampers our progress in addressing the disease threats of today and tomorrow. We are in a serious global competition.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7982296/

Zoonotic disease preparedness in sub-Saharan African countries

Background The emergence of high consequence pathogens such as Ebola and SARS-CoV-2, along with the continued burden of neglected diseases such as rabies, has highlighted the need for preparedness for emerging and endemic infectious diseases of zoonotic origin in sub-Saharan Africa (SSA) using a One Health approach. To identify trends in SSA preparedness, the World Health Organization (WHO) Joint External Evaluation (JEE) reports were analysed. JEEs are voluntary, collaborative processes to assess country's capacities to prevent, detect and rapidly respond to public health risks. This report aimed to analyse the JEE zoonotic disease preparedness data as a whole and identify strengths and weaknesses. Methods JEE zoonotic disease preparedness scores for 44 SSA countries who had completed JEEs were analysed. An overall zoonotic disease preparedness score was calculated as an average of the sum of all the SSA country zoonotic disease preparedness scores and compared to the overall mean JEE score. Zoonotic disease preparedness indicators were analysed and data were collated into regions to identify key areas of strength. Results The mean 'Zoonotic disease' preparedness score (2.35, range 1.00–4.00) was 7% higher compared to the mean overall JEE preparedness score (2.19, range 1.55–3.30), putting 'Zoonotic Diseases' 5th out of 19 JEE sub-areas for preparedness. The average scores for each 'Zoonotic Disease' category were 2.45 for 'Surveillance Systems', 2.76 for 'Veterinary Workforce' and 1.84 for 'Response Mechanisms'. The Southern African region scored highest across the 'Zoonotic disease' categories (2.87). A multisectoral priority zoonotic pathogens list is in place for 43% of SSA countries and 70% reported undertaking national surveillance on 1–5 zoonotic diseases. 70% of SSA countries reported having public health training courses in place for veterinarians and 30% had veterinarians in all districts (reported as sufficient staffing). A multisectoral action plan for zoonotic outbreaks was in place for 14% countries and 32% reported having an established inter-agency response team for zoonotic outbreaks. The zoonotic diseases that appeared most in reported country priority lists were rabies and Highly Pathogenic Avian Influenza (HPAI) (both 89%), anthrax (83%), and brucellosis (78%). Conclusions With 'Zoonotic Diseases' ranking 5th in the JEE sub-areas and a mean SSA score 7% greater than the overall mean JEE score, zoonotic disease preparedness appears to have the attention of most SSA countries. However, the considerable range suggests that some countries have more measures in place than others, which may perhaps reflect the geography and types of pathogens that commonly occur. The category 'Response Mechanisms' had the lowest mean score across SSA, suggesting that implementing a multisectoral action plan and response team could provide the greatest gains. Background The emergence of high consequence pathogens such as Ebola and SARS-CoV-2, along with the continued burden of neglected diseases such as rabies, has highlighted the need for preparedness for emerging and endemic infectious diseases of zoonotic origin in sub-Saharan Africa (SSA) using a One Health approach. To identify trends in SSA preparedness, the World Health Organization (WHO) Joint External Evaluation (JEE) reports were analysed. JEEs are voluntary, collaborative processes to assess country's capacities to prevent, detect and rapidly respond to public health risks. This report aimed to analyse the JEE zoonotic disease preparedness data as a whole and identify strengths and weaknesses. Methods JEE zoonotic disease preparedness scores for 44 SSA countries who had completed JEEs were analysed. An overall zoonotic disease preparedness score was calculated as an average of the sum of all the SSA country zoonotic disease preparedness scores and compared to the overall mean JEE score. Zoonotic disease preparedness indicators were analysed and data were collated into regions to identify key areas of strength. Results The mean 'Zoonotic disease' preparedness score (2.35, range 1.00–4.00) was 7% higher compared to the mean overall JEE preparedness score (2.19, range 1.55–3.30), putting 'Zoonotic Diseases' 5th out of 19 JEE sub-areas for preparedness. The average scores for each 'Zoonotic Disease' category were 2.45 for 'Surveillance Systems', 2.76 for 'Veterinary Workforce' and 1.84 for 'Response Mechanisms'. The Southern African region scored highest across the 'Zoonotic disease' categories (2.87). A multisectoral priority zoonotic pathogens list is in place for 43% of SSA countries and 70% reported undertaking national surveillance on 1–5 zoonotic diseases. 70% of SSA countries reported having public health training courses in place for veterinarians and 30% had veterinarians in all districts (reported as sufficient staffing). A multisectoral action plan for zoonotic outbreaks was in place for 14% countries and 32% reported having an established inter-agency response team for zoonotic outbreaks. The zoonotic diseases that appeared most in reported country priority lists were rabies and Highly Pathogenic Avian Influenza (HPAI) (both 89%), anthrax (83%), and brucellosis (78%). Conclusions With 'Zoonotic Diseases' ranking 5th in the JEE sub-areas and a mean SSA score 7% greater than the overall mean JEE score, zoonotic disease preparedness appears to have the attention of most SSA countries. However, the considerable range suggests that some countries have more measures in place than others, which may perhaps reflect the geography and types of pathogens that commonly occur. The category 'Response Mechanisms' had the lowest mean score across SSA, suggesting that implementing a multisectoral action plan and response team could provide the greatest gains. Study highlights 'Veterinary Workforce' was the strongest 'Zoonotic Disease' category across SSA 'Response Mechanisms' was the weakest category, and improving communication between clinical, veterinary and environmental sectors could improve zoonotic disease responses Southern Africa had the highest mean score for all 'Zoonotic Disease' categories, suggesting that other countries could adapt their strengths to their own situations The 5 most-cited zoonoses on SSA priority pathogen lists are rabies, Highly pathogenic avian influenza (HPAI), anthrax, brucellosis and bovine tuberculosis, all of which are 'neglected' diseases Introduction Many human infectious diseases have originated from animals. It is thought that 60% of currently known human infectious diseases [ 1 ] and as much as 75% of emerging infectious diseases are of zoonotic origin, or the result of a spill over event which then established itself in humans [ 1 , 2 ]. Whilst many high income countries (HICs) have successfully reduced or eradicated zoonoses, often utilizing expensive interventions, the heaviest burden of zoonotic diseases now often falls on low and middle-income countries (LMICs), who historically have the poorest healthcare infrastructure and rely most heavily on livestock economically [ 3 ]. This is especially important in sub-Saharan Africa (SSA), where reliance on livestock and bush meat can be high [ 4 ]. The fact that, in many of the regions affected by zoonotic diseases, livestock play a vital role both as a cash reserve and source of income in poor communities makes it not only medically, but also economically, important to ensure animals are kept healthy [ 4 ]. The environmental and social change our expanding population is causing is affecting the way infectious diseases spread across the globe. The increase in air travel to and from high-burden areas such as South America, Africa and Asia is likely to increase the introduction of vector-borne pathogens to new regions, whilst the abundance of human-commensal species, such as rats, is increasing alongside urbanisation and deforestation, raising the opportunity for human transmission of zoonoses [ 5 ]. As, by definition, zoonotic diseases are infectious diseases that are naturally transmitted from vertebrate animals to humans and vice versa [ 6 ], the need for a One Health approach cannot be overstated. Limiting or eliminating the transfer of pathogens between animals, the environment and humans, and therefore reducing the risk of them becoming an emerging infectious disease is of vital importance. The value of zoonotic pathogen surveillance is beginning to gain traction in Africa, although it is often difficult to identify whether data are collected, never mind what trends may exist [ 7 – 9 ]. The World Health Organization (WHO) notes that the steps to achieve a One Health approach should include: promoting the concept of a multisectoral approach and developing integrated control packages, raising the profile of neglected zoonotic diseases, collecting surveillance data to identify those at risk, and investing in the development of new tools, particularly diagnostics, to effectively control zoonoses [ 4 , 10 ]. Zoonotic disease spill over events (when zoonotic pathogens are transmitted to humans) occur frequently in Africa, including direct zoonotic events, from animals to humans, such as Lassa fever across West Africa (2016–2018) [ 11 ] and Monkeypox in West and Central Africa [ 12 ], as well as secondary, human-to-human epidemiological cycles such as Ebola in West Africa (2013–2016) and the Democratic Republic of Congo (2019) [ 13 , 14 ]. There are also more localised, but common, outbreaks of diseases such as salmonellosis and Yellow Fever [ 15 ]. Some, such as Rift Valley Fever, can occur locally, such as in South Africa [ 16 ] or be more widespread, such as across Eastern Africa, as well as the Middle East [ 17 ]. As a first step in preventing emerging infectious disease outbreaks of animal origin in humans, it is important for SSA countries to know what level of support systems are already in place. The scientific literature mainly focuses on the research being done, rather than the infrastructure that supports it, but reports from global organisations such as the Africa Centre for disease Control (Africa CDC), US Centers for Disease Control (US CDC), WHO, World Organisation for Animal Health (OIE) and the Food and Agriculture Organisation (FAO) suggest that much more could be done to combat zoonoses [ 4 , 18 – 20 ]. The WHO reports that, due to multiple factors such as poor funding, a lack of veterinary and clinical cooperation and disease misdiagnosis, the real burden of zoonoses is often missed [ 4 ]. This creates a lack of reliable evidence for governments and policy makers to utilise when implementing legislation at both local and national levels [ 21 ]. This WHO report also suggests that control measures are often undertaken in isolation, which is therefore likely to make them less effective than when part of a suite of control measures [ 4 ]. Strengthening country-specific preparedness for potential public health risks, including zoonotic disease, has increasingly become the focus for governments and health organisations. The 58th World Health Assembly in 2005 resulted in the formation of the International Health Regulations, which developed the Joint External Evaluations (JEEs): voluntary, collaborative procedures that assess a country's capacity to 'prevent, detect and rapidly respond to public health risks' [ 22 ]. The zoonotic disease preparedness for countries is measured by the presence of a 'functional multisectoral, multidisciplinary mechanisms, policies, systems and practices to minimize the transmission of zoonotic diseases from animals to human populations' [ 23 ]. In March 2020, 86% (44/50) of the SSA countries (according to WHO definition), had completed a JEE [ 23 , 24 ]. JEE reports cover prevention, detection, response and IHR related hazards and points of entry areas, which are classified into 19 sub-areas. 'Zoonotic Disease' is one of these sub-areas and is divided into three categories. Technical questions and indicators scores are applied to evaluate the preparedness of a country to a specific zoonotic disease. The indicator scores range from 1 to 5, indicating if a country has respectively: no, limited, developed, demonstrated or sustainable capacity to address a particular sub-area. Whilst there are many areas of zoonotic disease to focus on, it would be beneficial to step back and look at the wider picture. A comparison of the zoonotic disease facilities and infrastructure across SSA has yet to be made, and using JEE's provides a robust, standardised set of indicators to detect stronger and weaker areas across SSA. This provides an outline of how SSA is tackling zoonotic diseases, and how well zoonotic disease preparedness is implemented compared to other public health concerns, such as antimicrobial resistance. When collated into African regions (West, Central, East and Southern), these strengths can be further focussed and adapted by countries who may be finding it more challenging. The data acquired from this study can inform zoonotic disease policy in the future. This study aimed to analyse the JEE data to understand zoonotic disease preparedness across SSA countries. Methods The completed JEE reports from 44 SSA countries (as described by the WHO regions), were accessed between 6th November 2019 and 22nd March 2020 and analysed [ 25 ] and the sub-areas scores were compared [ 24 ]. The analysis of these data is described in Elton et al. [ 24 ]. The 'Zoonotic disease' sub-area was compared to the other sub-areas (such as immunization and antimicrobial resistance preparedness) and a ranking table was compiled [ 24 ]. The percentage of countries in each score category was determined, as well as the percentage of the countries that reported having zoonotic disease prevention and control structures in position, which was identified from the accompanying technical question written reports, as described in Table 1 . In the first edition of the JEE tool [ 26 ], there were three categories within 'Zoonotic Disease', whereas by 2018, the second edition [ 22 ] had removed the category 'Veterinary or animal health workforce'. The majority (86%) of countries had completed JEEs before this change, thus have all three categories (Central African Republic, Republic of Congo, Malawi, Gabon, Guinea Bissau, São Tomé and Príncipe and Comoros were completed later and have two categories). These scores were weighted into regions (15 West, 7 Central, 17 East and 5 Southern African countries, as defined by the United Nations) to pinpoint any patterns of zoonotic disease preparedness strengths [ 27 ]. The guidelines for how scores are ascribed are outlined in the JEE Tool [ 22 ]. Data was analysed as described in [ 24 ]. Table 1 Zoonotic disease preparedness categories and the indicators assessed. Indicators are taken from the scoring table (see the JEE tool, second edition [ 25 ]) or a technical question answer from the written report Category Indicator Source Surveillance systems in place for priority zoonotic diseases/pathogens (in this paper referred to as 'Surveillance Systems') (this section is included in the JEE tool first and second editions) Is there a national surveillance system in place for up to 5 priority zoonotic diseases? Scoring table within the JEE tools document Has a multisectoral agreement on 5 priority zoonotic pathogens been made? Technical questions Is there a national surveillance system in place for more than 5 zoonotic diseases? Technical questions Do clinical and veterinary laboratories communicate data/results with each other? Technical questions Veterinary or animal health workforce (in this paper referred to as 'Veterinary Workforce') (this section is in the JEE tool first edition only) Is there an animal health workforce capacity within the national public health system and at all sub-national levels? Scoring table within the JEE tools document Are there training opportunities for veterinary workers in zoonotic disease and transmission? Technical questions Mechanisms for responding to infectious and potential zoonotic diseases are established and functional (in this paper referred to as 'Response Mechanisms') (this section is included in the JEE tool first and second editions) Is an established multisectoral operational action plan for coordinated response to outbreaks of zoonotic diseases in place? Scoring table within the JEE tools document Do you have an established inter-agency zoonotic disease response team? Technical questions Do you have the capacity to respond to 80% of zoonotic events in a timely manner? Technical questions Findings When the SSA mean score for the sub-area 'Zoonotic Disease' was compared to the other sub-areas, it scored 2.35, placing it 5th out of 19 (range 1.33–3.38; p  5 diseases 13 (30%) 4 (27%) 2 (29%) 4 (24%) 3 (60%) Clinical and veterinary communication 11 (25%) 4 (27%) 1 (14%) 5 (29%) 1 (20%) Veterinary Workforce National veterinary workforce capacity 13 (30%) 3 (20%) 1 (14%) 6 (35%) 3 (60%) Veterinary worker training 31 (70%) 11 (73%) 3 (43%) 12 (71%) 5 (100%) Response Mechanisms Established multisectoral action plan 6 (14%) 0 (0%) 2 (29%) 3 (18%) 1 (20%) Inter-agency zoonotic response team 15 (34%) 6 (40%) 2 (29%) 6 (35%) 1 (20%) Capacity to respond to 80% of events 4 (9%) 0 (0%) 1 (14%) 3 (18%) 0 (0%) When category scores were weighted by region, it showed that Southern Africa had the highest mean score across all 'Zoonotic Disease' categories (range: 1.71–2.87). Table 4 shows the regional mean scores for each of the 'Zoonotic Disease' categories, as well as mean 'Zoonotic Disease' category and overall JEE mean scores. The regions showed significant differences in all categories except 'Response Mechanisms'. Table 4 SSA mean zoonotic preparedness category scores by region. Southern Africa had the highest mean 'Zoonotic Disease' score. All categories except 'Response Mechanisms' showed significant differences in scores between regions Total countries (in SSA) Surveillance Veterinary Workforce Response Mechanisms 'Zoonotic Disease' category mean Overall JEE mean West 15 2.00 2.20 1.87 2.02 2.09 Central 7 1.57 1.57 2.00 1.71 1.75 East 17 2.29 2.59 2.00 2.29 2.38 Southern 5 2.40 4.00 2.20 2.87 2.40 p value p = 0.0017 p = 0.0036 P = 0.3947 p = 0.0006 p = 0.0074 When assessing the preparedness indicators, Southern Africa had the highest percentage of countries with an approved list of priority pathogens (60%), surveillance in place for > 5 zoonotic diseases (60%), sufficient veterinarians in all districts (60%) and public health training for veterinary and animal health practitioners (100%). Central Africa had the highest percentage of countries reporting surveillance for between 1 and 5 zoonotic diseases (100%) and having multisectoral National Action Plans for zoonotic diseases (33%). West Africa had the highest percentage of countries reporting having an established inter-agency zoonotic disease response team (33%) and East Africa had the highest percentage of countries reporting that clinical and veterinary laboratories/services communicated with each other (29%) and having the capacity to respond to 80% of zoonotic events on time (17%). Nineteen countries indicated that they had a multisector-approved list of priority zoonotic pathogens and had listed some or all of those diseases. The zoonotic diseases that appeared most in reported country priority lists were rabies and Highly Pathogenic Avian Influenza (HPAI) (both 89%), anthrax (83%), and brucellosis (78%) (Fig. 2 ). Whilst some of these are defined as 'true' zoonotic diseases, others such as Dengue and SARS are defined as emerging infectious diseases [ 28 ]. Twenty three pathogens were reported to be on SSA country priority pathogen lists and a further seven were identified as important zoonotic pathogens (either in countries without an approved priority list, or as non-priority but still worthy of surveillance). Fig. 2 Histogram to show zoonotic diseases by the percentage of countries citing them on their priority pathogen lists. Rabies, Highly Pathogenic Avian Influenza (HPAI), anthrax, brucellosis and bovine tuberculosis were most frequently reported on SSA country priority pathogen lists. MERS-CoV stands for Middle East respiratory syndrome and BSE stands for bovine spongiform encephalopathy. The diseases included in the term 'Haemorrhagic fevers' varies between countries Findings When the SSA mean score for the sub-area 'Zoonotic Disease' was compared to the other sub-areas, it scored 2.35, placing it 5th out of 19 (range 1.33–3.38; p  5 diseases 13 (30%) 4 (27%) 2 (29%) 4 (24%) 3 (60%) Clinical and veterinary communication 11 (25%) 4 (27%) 1 (14%) 5 (29%) 1 (20%) Veterinary Workforce National veterinary workforce capacity 13 (30%) 3 (20%) 1 (14%) 6 (35%) 3 (60%) Veterinary worker training 31 (70%) 11 (73%) 3 (43%) 12 (71%) 5 (100%) Response Mechanisms Established multisectoral action plan 6 (14%) 0 (0%) 2 (29%) 3 (18%) 1 (20%) Inter-agency zoonotic response team 15 (34%) 6 (40%) 2 (29%) 6 (35%) 1 (20%) Capacity to respond to 80% of events 4 (9%) 0 (0%) 1 (14%) 3 (18%) 0 (0%) When category scores were weighted by region, it showed that Southern Africa had the highest mean score across all 'Zoonotic Disease' categories (range: 1.71–2.87). Table 4 shows the regional mean scores for each of the 'Zoonotic Disease' categories, as well as mean 'Zoonotic Disease' category and overall JEE mean scores. The regions showed significant differences in all categories except 'Response Mechanisms'. Table 4 SSA mean zoonotic preparedness category scores by region. Southern Africa had the highest mean 'Zoonotic Disease' score. All categories except 'Response Mechanisms' showed significant differences in scores between regions Total countries (in SSA) Surveillance Veterinary Workforce Response Mechanisms 'Zoonotic Disease' category mean Overall JEE mean West 15 2.00 2.20 1.87 2.02 2.09 Central 7 1.57 1.57 2.00 1.71 1.75 East 17 2.29 2.59 2.00 2.29 2.38 Southern 5 2.40 4.00 2.20 2.87 2.40 p value p = 0.0017 p = 0.0036 P = 0.3947 p = 0.0006 p = 0.0074 When assessing the preparedness indicators, Southern Africa had the highest percentage of countries with an approved list of priority pathogens (60%), surveillance in place for > 5 zoonotic diseases (60%), sufficient veterinarians in all districts (60%) and public health training for veterinary and animal health practitioners (100%). Central Africa had the highest percentage of countries reporting surveillance for between 1 and 5 zoonotic diseases (100%) and having multisectoral National Action Plans for zoonotic diseases (33%). West Africa had the highest percentage of countries reporting having an established inter-agency zoonotic disease response team (33%) and East Africa had the highest percentage of countries reporting that clinical and veterinary laboratories/services communicated with each other (29%) and having the capacity to respond to 80% of zoonotic events on time (17%). Nineteen countries indicated that they had a multisector-approved list of priority zoonotic pathogens and had listed some or all of those diseases. The zoonotic diseases that appeared most in reported country priority lists were rabies and Highly Pathogenic Avian Influenza (HPAI) (both 89%), anthrax (83%), and brucellosis (78%) (Fig. 2 ). Whilst some of these are defined as 'true' zoonotic diseases, others such as Dengue and SARS are defined as emerging infectious diseases [ 28 ]. Twenty three pathogens were reported to be on SSA country priority pathogen lists and a further seven were identified as important zoonotic pathogens (either in countries without an approved priority list, or as non-priority but still worthy of surveillance). Fig. 2 Histogram to show zoonotic diseases by the percentage of countries citing them on their priority pathogen lists. Rabies, Highly Pathogenic Avian Influenza (HPAI), anthrax, brucellosis and bovine tuberculosis were most frequently reported on SSA country priority pathogen lists. MERS-CoV stands for Middle East respiratory syndrome and BSE stands for bovine spongiform encephalopathy. The diseases included in the term 'Haemorrhagic fevers' varies between countries Discussion JEEs are an effective way to highlight a country's ability to address global health issues, including zoonotic diseases and outbreaks. That the mean 'Zoonotic Disease' score for SSA countries is higher than the overall JEE score, and that 'Zoonotic Diseases' ranks 5th in the JEE sub-areas, suggests that zoonoses and outbreaks are perceived to be of importance by most African countries. However, as there was a larger range in the 'Zoonotic Disease' scores compared to the overall JEE score, this suggests that zoonoses are a greater priority in some countries than others. This may reflect the geography, land use, and prevalence and usage of animal species in communities, as well as the donor driven initiatives, which might dictate how often humans come into contact with these zoonotic diseases and therefore how often cases or spill over may occur. 'Veterinary Workforce' scored highest of the three 'Zoonotic Diseases' categories. This is potentially a very positive aspect of the scores, as having active national disease surveillance and well-trained staff is an excellent basis upon which to build an effective zoonotic disease preparedness system. The category 'Response Mechanisms' scored lowest and therefore concentrating on building up these factors, and disseminating to the appropriate veterinary or animal health staff within the national public health system, may quickly improve country scores. The written reports accompanying the scores gave a deeper insight into the facilities and measures currently in place and yet to be achieved for each SSA country, although this depended on the depth of the response. For the category 'Surveillance Systems', whilst three quarters reported that they undertake surveillance for 1–5 zoonotic diseases, under half reported having a list of priority pathogens agreed by all sectors. This suggests that although zoonotic diseases are certainly on the radar of many SSA countries, many do not yet have the regulatory and administrative capacity in place to fully manage the problem, although the process of prioritising has been published by a number of countries, enabling others to follow suit [ 29 , 30 ]. That only one quarter of countries reported that their clinical and veterinary laboratories communicated with each other suggests that real gains could be made by bringing together the different sectors to create One Health guidelines. The OIE have created a version of the JEEs, evaluating veterinary services, although so far data from these surveys are available for only 36 countries [ 31 ]. If the two were to be combined, this would create a much more cohesive response to global health issues. This is especially true when the fact that, in the 'Veterinary Workforce' category, almost three quarters of SSA countries reported having public health training for their veterinary and animal health staff within the national public health workforce, who are capable of following these guidelines. A list of One Health training programmes and resources has been compiled by Rwego et al. (2016) [ 34 ]. By increasing the amount of trained veterinary and animal health professionals, and therefore increasing the regional coverage, the lower scoring aspects can be developed and successfully rolled out across a country. In the 'Response Mechanisms' category, one third of countries reported having an established inter-agency response team in place, although only six countries reported having a multisectoral National Action Plan in place. This suggests that whilst many countries are engaging their different departments to tackle zoonotic diseases, more work needs to be done to improve legislation, which should enable a more successful response. The adoption of a resolution with an emphasis on the One Health approach for the successful control of 17 neglected tropical diseases, including zoonoses, at the World Health Assembly in 2013 suggests that there is an international urge to move this forward [ 32 ]. When the scores were weighted into regions, countries in the Southern African region had the highest mean 'Zoonotic Disease' preparedness score, whilst countries in the Central African region had the lowest. The score for countries in the Southern African region was bolstered by a particularly high 'Veterinary Workforce' score compared to the other regions, which suggests that lessons could be learned from this region in terms of the training of veterinary staff in the animal and public health sectors. That all but one of the top six most reported priority pathogens are well established diseases with a long history of human-animal transmission suggests that SSA has so far struggled to control these zoonoses adequately. Cost effective control measures already exist for several neglected zoonotic diseases such as rabies and brucellosis [ 4 ], so the fact that they are on around three quarters of SSA country's priority pathogen lists suggests that more effective systems may need to be implemented, utilising some of this study's highlighted points, such as greater communication between veterinary and clinical services. Other commonly cited zoonoses were salmonella, Ebola, and haemorrhagic diseases (for some countries, this collectively included Ebola). The latter two are classed as emerging infectious diseases, which have jumped from being zoonoses to having human to human transmission. Some countries have to contend not only with well established, 'neglected' zoonoses, but also with novel ones, which, in the cases of SARS-CoV-2 and Ebola, can put extreme pressure on already often stretched healthcare and infrastructure systems, especially when a country does not have the capacity to respond effectively. Countries reporting the dual problem of established and emerging zoonoses are the ones who need a robust response system the most and are more likely to require assistance and guidance when it comes to zoonotic disease preparedness. Control measures, not just for zoonotic diseases but for many other diseases, have historically been undertaken individually or in isolation. Stakeholders must embrace the multifactorial, and multisectoral approach to gain the maximum benefit out of these improvements. Zoonoses and their outbreaks need to be recognised as a One Health problem and, as such, greater cooperation must occur between departments to ensure the problems are approached from all sides. Whilst costs for zoonotic disease interventions may seem costly on top of public health strategies, the long term benefit and overall cost effectiveness is likely to be greater [ 33 ]. Conclusions As zoonotic outbreaks do not respect political borders, and having identified regions with more developed control strategies, this is an excellent opportunity for countries with more advanced zoonotic disease programmes (such as those in the Southern African region) to help other nations improve, by sharing protocols, strategies and training, thus providing better coverage across the continent, and globally. Groups from the region, e.g. SACIDS ( http://www.sacids.org/ ) should be called upon to tailor their knowledge and share it with the rest of the continent. SSA countries need to fully utilise public health, veterinary and environmental government departments, as well as the advice and fundamental research of both African and global organisations, including One Health consortia, such as the Pan-African network PANDORA-ID-Net ( www.pandora-id.net ), if they are to build a robust One Health zoonotic disease preparedness response. Study highlights 'Veterinary Workforce' was the strongest category across SSA 'Response Mechanisms' was the weakest category across SSA, and that improving communication between clinical, veterinary and environmental sectors could improve zoonotic disease responses Southern Africa had the highest mean score for all 'Zoonotic Disease' categories, suggesting that other countries could adapt their strengths to their own situations 30% of SSA countries had veterinarians in all districts The 5 most-cited zoonoses on SSA priority pathogen lists are rabies, Highly pathogenic avian influenza (HPAI), anthrax, brucellosis and bovine tuberculosis, all of which are 'neglected' diseases Authors' information (optional) Not applicable.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9875040/

Editorial: Nanomaterials based electrochemical sensors for hazardous pollutants detection

Author contributions DS prepared the draft and finalized the manuscript. AK and HZ edited and corrected the draft. All authors contributed to the article and approved the submitted version. Conflict of interest The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Publisher's note All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5734740/

The coordination of plural logics of action and its consequences: Evidence from plural medical systems

Drawing on the theory of social action in organizational and institutional sociology, this paper examines the behavioral consequences of plural logics of action. It addresses the question based on the empirical case of plural medical systems that are composed of both biomedicine and alternative medicine. Applying mixed methods of a cross-national panel data analysis and a content analysis of medical journal articles, it finds that plural systems affect health outcomes negatively when tensions between biomedicine and alternative medicine are unaddressed. In contrast, plural systems produce tangible health benefits when biomedicine and alternative medicine are coordinated through government policies or by health care organizations/professionals. This paper proposes plurality coordination as an important mechanism that modifies the behavioral consequences of plural logics. This proposition contributes to providing theoretical answers to the sociological puzzle that plural logics of action produce inconsistent behavioral consequences. Introduction Limits of the sociology of plural action logics Logics guide behavior in several forms, such as values, habits, organizations, institutions, skills, and cultures [ 1 – 6 ]. Behavior is increasingly shaped by the plurality of these logics, as by multiple organizational logics [ 7 – 9 ], diverse institutions [ 10 , 11 ], heterogeneous cultures [ 12 – 17 ], hybrid habits and habitus/habiti [ 18 – 20 ], or multiple identities [ 21 – 24 ]. These plural action logics are not always compatible with one another. They are often conflicting and contradictory. Thus, how social actors organize their behavior under plural logics and what behavioral consequences these logics produce are substantial sociological questions in prominent subfields [ 18 , 25 – 29 ]. Two competing accounts exist. When plural logics are compatible with one another, more logics add up to one another and produce more resources for action, promising favorable consequences in health [ 22 , 23 ], government welfare benefits [ 21 ], school lives [ 30 ], and movement mobilization [ 31 ]. When plural logics are conflicting, more logics produce strains on action, making behavior under plural logics uneasy and unsuccessful as in adolescent sexual behavior [ 12 ] and national economic policies [ 24 , 32 ]. Neither account, however, explains these contradictory consequences at once. In particular, neither of these simplistic accounts can accommodate the inconsistent consequences from plural medical systems that are composed of both biomedicine and "alternative medicine," such as acupuncture, herbal medicine, ethnic medical traditions, and spiritual healing [ 33 – 35 ]. The medical literature employs a variety of terms to refer to these medical practices that are in tensions with conventional biomedicine, such as complementary and alternative medicine (CAM), complementary and integrative medicine (CIM), and traditional, complementary and alternative medicine (TCAM). Highlighted in these terms are different normative views on the relationship between conventional biomedicine and unconventional medical practices. As a way to remain neutral and thus to be ultimately able to examine the relationship in practice, this paper uses the term "alternative medicine" which may or may not be integrated into biomedicine as complementary medicine. (For the rationale to use the binary representation of biomedicine and alternative medicine, refer to the multi-dimensional tensions between biomedicine and alternative medicine elaborated right after the Introduction; for the operational conceptualization of alternative medicine to delineate the empirical data, refer to the Data sections of panel data analysis and content analysis.) The account of additive resourcefulness in the sociological literature seems to be consistent only with the findings that plural medical systems composed of biomedicine and alternative medicine provide diverse medical resources which result in occasional treatment synergies between biomedicine and alternative medicine [ 36 , 37 ] and the emotional support and empowerment of patients with alternative medical practices [ 19 , 38 , 39 ]. However, this additivity account does not simultaneously explain tensions between biomedicine and alternative medicine and subsequent adverse consequences, such as conflicts between treatment regimens in biomedicine and those in alternative medicine [ 40 ], confrontations between professionals of biomedicine and those of alternative medicine [ 41 ], insufficient communication between medical doctors and minority patients on alternative medicine [ 42 ], premature deaths of HIV/AIDS patients selecting alternative medicine over the biomedical drug treatment [ 19 ], and adverse interactions between medications and herbal/dietary supplements [ 43 ]. The alternative sociological account of strains on action seems to be more suitable for accommodating these findings. However, this strains account then fails to incorporate the aforementioned beneficial consequences. The critical puzzle that plural medical systems lead to inconsistent healthcare outcomes still remains to be answered. Limits of medical studies Medical studies, for their part, attempt to comprehend these inconsistent outcomes with scientific medical trials [ 44 ]. Upon realizing that significant discrepancies persist even between these trials, they then use a meta-analytic approach which recalibrates the true efficacy of medical interventions by relating inconsistent outcomes to the variations in trial design, such as the characteristics of trial subjects, the medical technologies used, and the outcome measures [ 45 , 46 ]. While resolving some of the inconsistent outcomes, the meta-analytic approach is still perplexed by the remaining inconsistencies between trials of an identical design. These remaining inconsistencies seem to be related to the geographical location of trials [ 47 ] or the ways in which alternative medicine interventions are arranged with the surrounding biomedical interventions [ 48 ]. However, medical trials and meta-analyses rarely investigate how users and practitioners behave–and are indeed conditioned to behave–at the intersection of the two different logics of medical practices: biomedicine and alternative medicine. The wisdom that the treatment effect of biomedical interventions is significantly modified by indigenous medical cultures [ 14 , 49 ] has not spawned the notion that the effect of alternative medicine can be conditioned by biomedical treatment environments. This paper fills in this deficiency and further develops this notion by embedding it to the sociology of organizations and institutions that goes beyond the two preceding simple accounts of plural action logics. Re-engaging the sociology of plural action logics The sociological literature shows that even conflicting plural logics are organized to co-exist with one another. First, they are organized through the individual wisdom and skill of actors, as in skillful workmen with a toolkit of diverse cultural models during settled lives [ 13 ], artful individuals and organizations managing institutional conflicts in work–family relations [ 50 ] and democracy–bureaucracy relations [ 10 ], innovative government actors conjuring novel institutions by patching institutions of diverse national origins [ 51 ], creative LGBT movement organizers anchoring unrelated and divisive movement motives in the most common social movement ideas [ 7 ], life scientists constructing a hybrid exchange logic between the conflicting logics of science and commerce by redefining the meaning and practices of patenting [ 11 ], museum professionals progressively upholding the conflicting goals of conservation and exhibition [ 8 ], and lay people aptly appropriating foreign medical elements into the indigenous medical system [ 52 ]. Second, supra-individual forces guide conflicting logics to be organized and co-practiced, such as the historical-structural conditions surrounding individuals with unwieldy tools during unsettled periods [ 13 ]. The medical habitus of South Africans who co-practice the incongruent elements of biomedicine and traditional African medicine is guided structurally by the forces of late global capitalism [ 19 ]. In addition, the hybrid medical habitus of U.S. immigrants, who often rely on ethnic/folk medicine along with biomedicine, is directed by collective forces, such as ethnicity and social ties [ 18 ]. While informative, these studies still fall short of a theory of the behavioral consequences of plural logics. They do not examine the consequences of plural logics and, instead, only demonstrate the organizability of conflicting logics. It remains unexamined whether plural logics produce any difference in behavioral outcomes when they are organized compared to when they are not. For example, it is unclear how the hybrid logic of exchange developed by skilled scientists [ 11 ] changes their practice of science, compared to when there is no such hybrid logic. Likewise, it is unexamined what kinds of differences the "thin" [ 15 ] and "limited" [ 16 ] anchoring of divergent movement ideas in the most general causes (e.g. community-building and equality) brings to the LGBT movement [ 7 ], compared to when there is no such anchoring. Art museums' and museum professionals' performance during the brief co-existence of the conflicting goals of conservation and exhibition is not examined against their performance under the single goal of exhibition [ 8 ]. Meanwhile, some of these studies provide intriguingly contradictory observations on the consequences of organized plural logics. For example, the plural health behavior of HIV/AIDS patients in Africa produces prolonged lives in some countries [ 38 , 53 ] and premature deaths in others [ 19 ], all of which are situated in the same structural condition (late global capitalism). While the medical habitus utilizing both biomedicine and alternative medicine commonly produces psychological relief and empowerment [ 19 , 39 ], one case leads to premature deaths [ 19 ] while the other shows no clear impact on lives [ 39 ]. However, these substantial contrasts, such as the winning-vs- losing hybrid habitus, the winning-vs-losing organization of plural logics, and the quick/easy–vs–slow/difficult organization of logics, have not been theorized within a single conceptual frame yet. A closer examination even reveals that conflicting logics are in fact organized only among some actors and not others. For example, the appropriation of the meaning of patenting is observed unevenly among life scientists. Thus, the subsequent organization of conflicting logics of science and commerce is found to be relatively easier among "senior" scientists than "juniors" [ 11 ]; the organization is easier for scientists who often interact with other professionals in law and administration than those who do not [ 11 ]. Different logics of medical practices, such as biomedicine and traditional African medicine, are organized with the help of medical care professionals only in some countries [ 38 , 53 ] and not in others [ 19 ]. The organizability and the locus of organizing agency during "settled" periods differ from those during "unsettled" periods [ 13 ]. Museum professionals' progressive upholding of the conflicting goals of conservation and exhibition seems to be tenable when there is sufficient budgetary support, whereas this is easily threatened by budgetary constraints [ 8 ]. Thus, this paper purports to illuminate that some conflicting plural logics are coordinated and produce favorable consequences while others are hardly coordinated and result in undesired consequences. The paper argues that conflicting plural logics produce favorable consequences when they are coordinated within supportive social contexts. Otherwise, it is difficult for individuals to organize conflicting logics and utilize them to their benefits. The paper bases this argument on the varying and even contradictory health outcomes under the plural medical systems of biomedicine and alternative medicine. A cross-national panel data analysis tests two hypotheses about the varying health outcomes of plural medical systems. Then, a comparative analysis of the U.S. and Japan demonstrates how results from the cross-national panel analysis are triangulated by medical treatment episodes in medical journals. Plural medical systems and varying health consequences Medical practices are found to be pluralist for economic, cultural, and medical reasons in the developed as well as developing world [ 33 , 54 , 55 ]. A majority of countries have official policies on popular alternative medicine practices, such as Ayurveda, Chinese medicine, homeopathy, osteopathy, chiropractic, bone-setting, herbal medicine, and spiritual/religious therapies [ 33 ]. About half of these countries have government financial provision for alternative medicine. Alternative medicine is utilized widely to treat various medical conditions, such as infectious diseases (e.g. HIV/AIDS) and chronic health problems (e.g. cancers, cardiovascular conditions, and musculoskeletal problems). Therefore, its impact can be consequential for national health outcomes across the world in a positive or negative way. The additivity account of plural logics posits that plural logics linearly add up to the resourcefulness of strategies for action. Thus, a plural medical system provides more resources to manage health and promises better health outcomes than a system of biomedicine alone. This resourceful additivity account finds any adverse health outcomes to be unexpected and attributable only to the invalidity of alternative medicine for health management [ 44 ]. On the other hand, the strains account of plural logics suggests that a plural medical system is laden with conflicts between biomedicine and alternative medicine. Adverse health outcomes from the conflict–loaded plural medical system are inevitable. These adverse outcomes are not prevented by simply improving the safety and efficacy of alternative medicine. On the contrary, any desired positive health outcomes are viewed as unexpected consequences. Challenging is not only the conceptual bifurcation of these two accounts but also the discrepancy between each account and empirical evidence. To address these deficiencies, this paper formulates an alternative account in which tensions between biomedicine and alternative medicine are not overlooked nor deemed insurmountable. Plural medical systems are indeed laden with tensions between biomedicine and alternative medicine and, for this reason, produce difficulties for medical practices among lay users and professional practitioners. First, these tensions are medical-theoretical. In terms of etiology, alternative medicine views a human being in its totality within a wide and remote ecology, and attributes ill health to the disequilibrium of this total ecological system vis-à -vis biomedicine's emphasis on proximate causative agents and their pathogenic evolution within the boundaries of a body [ 56 , 57 ]. From a social scientific approach that views medicine as a social systemic process of allocating accountability for individual and collective failures/successes, alternative medicine features a holistic and spiritual allocation of accountability, whereas biomedicine relatively centers on a partial and physical allocation [ 58 ]. Reflecting this view, the system of MeSH terms underlying the indexing in the PubMed database specifies a subset of MeSH terms that group together a variety of medical practices as complementary and alternative medicine (for detailed strategies, refer to http://www.nlm.nih.gov/bsd/pubmed_subsets/comp_med_strategy.html ) [ 59 ]. Second, the tensions are political and cultural as well. In the social constructionist view of medicine and medicine-as-profession [ 60 , 61 ], any brand of medicine employs the process of political and cultural negotiation. Biomedicine is historically negotiated as science and progressive modernity, whereas alternative medicine is portrayed as non-science, magic, and a backward tradition [ 62 – 64 ]. In colonial and postcolonial contexts, biomedicine is aligned with global domination projects, whereas alternative medicine is often identified with local autonomy [ 19 , 65 ]. These medical-theoretical and political-cultural tensions are evident in the behavior of medical service users and professional practitioners. Cancer patients who use both biomedicine and alternative medicine are often caught up between two conflicting theories, or biomedicine's depersonalized approach and alternative medicine's individualized approach [ 66 ]. When patients reclaim their control over medical decision-making through the use of alternative medicine, they are not always confident about their knowledge base and are often dependent on the advice of biomedical professionals [ 67 ]. Patients' orientation toward spiritual healing comes into conflict with physicians' treatment regimens [ 68 , 69 ]. Users of alternative medicine do not readily discuss their use of alternative medicine with their physicians, since physicians are often found to ignore alternative medicine [ 70 , 71 ]. These tensions between biomedicine and alternative medicine, however, do not create the same difficulties in different plural medical systems. Health behaviors and outcomes vary widely, depending on how these tensions are articulated in the system. For instance, national political contexts heighten or mitigate these tensions, producing different consequences. Post-Apartheid nationalist politics in South Africa, that portray alternative medicine as an indigenous African tradition and biomedicine as "Western White" science, is accused of strengthening inter-professional barriers in HIV treatment between medical doctors who provide drug therapy and traditional healers who practice herbal and spiritual healing [ 19 ]. It is also blamed for creating unnecessary social criticism on HIV patients that prevents them from taking an effective biomedical drug therapy (ART). In other African countries (e.g. Uganda, Kenya, and Tanzania) whose political environments promote alternative medicine education for biomedical practitioners and the cross-referral of patients between biomedicine and alternative medicine practitioners, HIV patients take the drug therapy for controlling the viral loads and, simultaneously, use herbs and spiritual healing for their immune systems and emotional well-being [ 38 , 53 ]. Micro-level therapeutic environments matter, as well. For instance, Steve Jobs spent the first nine months of his diagnosis with pancreatic cancer trying alternative medicine, such as diets, fruit juices, acupuncture, herbal remedies, and spiritual practices [ 72 ]. Disagreement over his behavior aside, his personal account exemplifies that alternative medicine users outside institutionalized settings often go through difficult responses from families, friends, and even doctors who are "infuriated" and "distressed" by their reliance on alternative medicine. Thus, patients often have to "search the Internet" to get information about alternative medicine. The use of alternative medicine becomes a lonely and secretive process. In another context as in integrative cancer care centers, however, plural medical practices are open and communicative between patients, families, and practitioners [ 73 , 74 ]. Hypotheses and two complementary analyses Therefore, this paper hypothesizes that tensions between biomedicine and alternative medicine and their management in institutional settings are consequential for health behavior [ 75 , 76 ] and outcomes [ 59 ]. It conducts two complementary analyses: one at the macro level with cross-national panel data, testing two hypotheses on the changing relationships between plural medical systems and national health outcomes; the other at the micro level with data extracted from medical journal articles in the U.S. and Japan, elaborating how macro-level patterns are related to the treatment-level evidence. The macro-level analysis takes two steps. First, it examines how the plurality of a national medical system affects health outcomes. It is hypothesized that plurality will affect health outcomes negatively as tensions between biomedicine and alternative medicine produce strains on the behaviors of users and practitioners ( Hypothesis 1 ). Strains on health behaviors are identifiable in the delayed utilization of relevant medical resources, whether they are biomedicine or alternative medicine [ 77 , 78 ], the non-utilization of these relevant resources [ 58 , 79 ], and the ill-informed/ill-sequenced combinations of biomedicine and alternative medicine that lead to adverse interactions between the interventions [ 80 ]. Delayed utilization, non-utilization, and ill-informed utilization of medical resources, in turn, affect health care outcomes adversely. Second, this paper examines how measures that ameliorate tensions between biomedicine and alternative medicine change the relationship between plurality and health outcomes. It hypothesizes that measures of tension reduction will weaken the negative effect of plurality and convert it into a positive one, since reduced tensions help resolve the strains of plural logics and generate their resourcefulness through treatment synergies between biomedicine and alternative medicine ( Hypothesis 2 ). In particular, the paper examines government funding for alternative medicine and medical knowledge production as two promising measures of coordination mechanism that reduces the tensions. Government funding for alternative medicine refers to the existence of public financial support for alternative medicine through government insurance coverage or free provisions of alternative medicine in public health facilities. Either through direct service provisions or rules for insurance reimbursement, governments provide specific ways to practice alternative medicine along with biomedicine, thus reducing biomedicine–alternative medicine tensions and adverse outcomes. Medical knowledge production refers to the annual number of medical journal articles produced by each country. It measures the extent to which information on medical interventions is circulated within the medical community. In countries producing more papers, the communication between biomedicine and alternative medicine is greater and helps inform medical practitioners on both sides, thus lowering professional barriers between the two sides. The second-step micro-level analysis conducts a content analysis to elaborate different ways to coordinate plural medical systems and the extent to which plurality coordination moderates the effects of plural medical systems [ 81 ] in the U.S. and Japan. This analysis reveals the medical treatment-level evidence on biomedicine–alternative medicine tensions, their institutional environments, and treatment outcomes. Subsequent multivariate regression models of the results of this content analysis ultimately demonstrate that cross-national differences in the treatment outcomes are explained by cross-national differences in biomedicine–alternative medicine tensions and their coordination. This analysis examines the U.S. and Japan as a comparative pair, since they feature relatively pluralist medical systems [ 33 , 56 ] and provide a clear contrast in the institutionalization of the plural systems [ 59 ]. The U.S. features a market-based decentralized and informal coordination of biomedicine and alternative medicine [ 82 , 83 ], whereas Japan represents a government-driven centralized formal coordination [ 84 , 85 ]. This contrast makes it promising to examine how different institutional settings are aligned with treatment outcomes. Limits of the sociology of plural action logics Logics guide behavior in several forms, such as values, habits, organizations, institutions, skills, and cultures [ 1 – 6 ]. Behavior is increasingly shaped by the plurality of these logics, as by multiple organizational logics [ 7 – 9 ], diverse institutions [ 10 , 11 ], heterogeneous cultures [ 12 – 17 ], hybrid habits and habitus/habiti [ 18 – 20 ], or multiple identities [ 21 – 24 ]. These plural action logics are not always compatible with one another. They are often conflicting and contradictory. Thus, how social actors organize their behavior under plural logics and what behavioral consequences these logics produce are substantial sociological questions in prominent subfields [ 18 , 25 – 29 ]. Two competing accounts exist. When plural logics are compatible with one another, more logics add up to one another and produce more resources for action, promising favorable consequences in health [ 22 , 23 ], government welfare benefits [ 21 ], school lives [ 30 ], and movement mobilization [ 31 ]. When plural logics are conflicting, more logics produce strains on action, making behavior under plural logics uneasy and unsuccessful as in adolescent sexual behavior [ 12 ] and national economic policies [ 24 , 32 ]. Neither account, however, explains these contradictory consequences at once. In particular, neither of these simplistic accounts can accommodate the inconsistent consequences from plural medical systems that are composed of both biomedicine and "alternative medicine," such as acupuncture, herbal medicine, ethnic medical traditions, and spiritual healing [ 33 – 35 ]. The medical literature employs a variety of terms to refer to these medical practices that are in tensions with conventional biomedicine, such as complementary and alternative medicine (CAM), complementary and integrative medicine (CIM), and traditional, complementary and alternative medicine (TCAM). Highlighted in these terms are different normative views on the relationship between conventional biomedicine and unconventional medical practices. As a way to remain neutral and thus to be ultimately able to examine the relationship in practice, this paper uses the term "alternative medicine" which may or may not be integrated into biomedicine as complementary medicine. (For the rationale to use the binary representation of biomedicine and alternative medicine, refer to the multi-dimensional tensions between biomedicine and alternative medicine elaborated right after the Introduction; for the operational conceptualization of alternative medicine to delineate the empirical data, refer to the Data sections of panel data analysis and content analysis.) The account of additive resourcefulness in the sociological literature seems to be consistent only with the findings that plural medical systems composed of biomedicine and alternative medicine provide diverse medical resources which result in occasional treatment synergies between biomedicine and alternative medicine [ 36 , 37 ] and the emotional support and empowerment of patients with alternative medical practices [ 19 , 38 , 39 ]. However, this additivity account does not simultaneously explain tensions between biomedicine and alternative medicine and subsequent adverse consequences, such as conflicts between treatment regimens in biomedicine and those in alternative medicine [ 40 ], confrontations between professionals of biomedicine and those of alternative medicine [ 41 ], insufficient communication between medical doctors and minority patients on alternative medicine [ 42 ], premature deaths of HIV/AIDS patients selecting alternative medicine over the biomedical drug treatment [ 19 ], and adverse interactions between medications and herbal/dietary supplements [ 43 ]. The alternative sociological account of strains on action seems to be more suitable for accommodating these findings. However, this strains account then fails to incorporate the aforementioned beneficial consequences. The critical puzzle that plural medical systems lead to inconsistent healthcare outcomes still remains to be answered. Limits of medical studies Medical studies, for their part, attempt to comprehend these inconsistent outcomes with scientific medical trials [ 44 ]. Upon realizing that significant discrepancies persist even between these trials, they then use a meta-analytic approach which recalibrates the true efficacy of medical interventions by relating inconsistent outcomes to the variations in trial design, such as the characteristics of trial subjects, the medical technologies used, and the outcome measures [ 45 , 46 ]. While resolving some of the inconsistent outcomes, the meta-analytic approach is still perplexed by the remaining inconsistencies between trials of an identical design. These remaining inconsistencies seem to be related to the geographical location of trials [ 47 ] or the ways in which alternative medicine interventions are arranged with the surrounding biomedical interventions [ 48 ]. However, medical trials and meta-analyses rarely investigate how users and practitioners behave–and are indeed conditioned to behave–at the intersection of the two different logics of medical practices: biomedicine and alternative medicine. The wisdom that the treatment effect of biomedical interventions is significantly modified by indigenous medical cultures [ 14 , 49 ] has not spawned the notion that the effect of alternative medicine can be conditioned by biomedical treatment environments. This paper fills in this deficiency and further develops this notion by embedding it to the sociology of organizations and institutions that goes beyond the two preceding simple accounts of plural action logics. Re-engaging the sociology of plural action logics The sociological literature shows that even conflicting plural logics are organized to co-exist with one another. First, they are organized through the individual wisdom and skill of actors, as in skillful workmen with a toolkit of diverse cultural models during settled lives [ 13 ], artful individuals and organizations managing institutional conflicts in work–family relations [ 50 ] and democracy–bureaucracy relations [ 10 ], innovative government actors conjuring novel institutions by patching institutions of diverse national origins [ 51 ], creative LGBT movement organizers anchoring unrelated and divisive movement motives in the most common social movement ideas [ 7 ], life scientists constructing a hybrid exchange logic between the conflicting logics of science and commerce by redefining the meaning and practices of patenting [ 11 ], museum professionals progressively upholding the conflicting goals of conservation and exhibition [ 8 ], and lay people aptly appropriating foreign medical elements into the indigenous medical system [ 52 ]. Second, supra-individual forces guide conflicting logics to be organized and co-practiced, such as the historical-structural conditions surrounding individuals with unwieldy tools during unsettled periods [ 13 ]. The medical habitus of South Africans who co-practice the incongruent elements of biomedicine and traditional African medicine is guided structurally by the forces of late global capitalism [ 19 ]. In addition, the hybrid medical habitus of U.S. immigrants, who often rely on ethnic/folk medicine along with biomedicine, is directed by collective forces, such as ethnicity and social ties [ 18 ]. While informative, these studies still fall short of a theory of the behavioral consequences of plural logics. They do not examine the consequences of plural logics and, instead, only demonstrate the organizability of conflicting logics. It remains unexamined whether plural logics produce any difference in behavioral outcomes when they are organized compared to when they are not. For example, it is unclear how the hybrid logic of exchange developed by skilled scientists [ 11 ] changes their practice of science, compared to when there is no such hybrid logic. Likewise, it is unexamined what kinds of differences the "thin" [ 15 ] and "limited" [ 16 ] anchoring of divergent movement ideas in the most general causes (e.g. community-building and equality) brings to the LGBT movement [ 7 ], compared to when there is no such anchoring. Art museums' and museum professionals' performance during the brief co-existence of the conflicting goals of conservation and exhibition is not examined against their performance under the single goal of exhibition [ 8 ]. Meanwhile, some of these studies provide intriguingly contradictory observations on the consequences of organized plural logics. For example, the plural health behavior of HIV/AIDS patients in Africa produces prolonged lives in some countries [ 38 , 53 ] and premature deaths in others [ 19 ], all of which are situated in the same structural condition (late global capitalism). While the medical habitus utilizing both biomedicine and alternative medicine commonly produces psychological relief and empowerment [ 19 , 39 ], one case leads to premature deaths [ 19 ] while the other shows no clear impact on lives [ 39 ]. However, these substantial contrasts, such as the winning-vs- losing hybrid habitus, the winning-vs-losing organization of plural logics, and the quick/easy–vs–slow/difficult organization of logics, have not been theorized within a single conceptual frame yet. A closer examination even reveals that conflicting logics are in fact organized only among some actors and not others. For example, the appropriation of the meaning of patenting is observed unevenly among life scientists. Thus, the subsequent organization of conflicting logics of science and commerce is found to be relatively easier among "senior" scientists than "juniors" [ 11 ]; the organization is easier for scientists who often interact with other professionals in law and administration than those who do not [ 11 ]. Different logics of medical practices, such as biomedicine and traditional African medicine, are organized with the help of medical care professionals only in some countries [ 38 , 53 ] and not in others [ 19 ]. The organizability and the locus of organizing agency during "settled" periods differ from those during "unsettled" periods [ 13 ]. Museum professionals' progressive upholding of the conflicting goals of conservation and exhibition seems to be tenable when there is sufficient budgetary support, whereas this is easily threatened by budgetary constraints [ 8 ]. Thus, this paper purports to illuminate that some conflicting plural logics are coordinated and produce favorable consequences while others are hardly coordinated and result in undesired consequences. The paper argues that conflicting plural logics produce favorable consequences when they are coordinated within supportive social contexts. Otherwise, it is difficult for individuals to organize conflicting logics and utilize them to their benefits. The paper bases this argument on the varying and even contradictory health outcomes under the plural medical systems of biomedicine and alternative medicine. A cross-national panel data analysis tests two hypotheses about the varying health outcomes of plural medical systems. Then, a comparative analysis of the U.S. and Japan demonstrates how results from the cross-national panel analysis are triangulated by medical treatment episodes in medical journals. Plural medical systems and varying health consequences Medical practices are found to be pluralist for economic, cultural, and medical reasons in the developed as well as developing world [ 33 , 54 , 55 ]. A majority of countries have official policies on popular alternative medicine practices, such as Ayurveda, Chinese medicine, homeopathy, osteopathy, chiropractic, bone-setting, herbal medicine, and spiritual/religious therapies [ 33 ]. About half of these countries have government financial provision for alternative medicine. Alternative medicine is utilized widely to treat various medical conditions, such as infectious diseases (e.g. HIV/AIDS) and chronic health problems (e.g. cancers, cardiovascular conditions, and musculoskeletal problems). Therefore, its impact can be consequential for national health outcomes across the world in a positive or negative way. The additivity account of plural logics posits that plural logics linearly add up to the resourcefulness of strategies for action. Thus, a plural medical system provides more resources to manage health and promises better health outcomes than a system of biomedicine alone. This resourceful additivity account finds any adverse health outcomes to be unexpected and attributable only to the invalidity of alternative medicine for health management [ 44 ]. On the other hand, the strains account of plural logics suggests that a plural medical system is laden with conflicts between biomedicine and alternative medicine. Adverse health outcomes from the conflict–loaded plural medical system are inevitable. These adverse outcomes are not prevented by simply improving the safety and efficacy of alternative medicine. On the contrary, any desired positive health outcomes are viewed as unexpected consequences. Challenging is not only the conceptual bifurcation of these two accounts but also the discrepancy between each account and empirical evidence. To address these deficiencies, this paper formulates an alternative account in which tensions between biomedicine and alternative medicine are not overlooked nor deemed insurmountable. Plural medical systems are indeed laden with tensions between biomedicine and alternative medicine and, for this reason, produce difficulties for medical practices among lay users and professional practitioners. First, these tensions are medical-theoretical. In terms of etiology, alternative medicine views a human being in its totality within a wide and remote ecology, and attributes ill health to the disequilibrium of this total ecological system vis-à -vis biomedicine's emphasis on proximate causative agents and their pathogenic evolution within the boundaries of a body [ 56 , 57 ]. From a social scientific approach that views medicine as a social systemic process of allocating accountability for individual and collective failures/successes, alternative medicine features a holistic and spiritual allocation of accountability, whereas biomedicine relatively centers on a partial and physical allocation [ 58 ]. Reflecting this view, the system of MeSH terms underlying the indexing in the PubMed database specifies a subset of MeSH terms that group together a variety of medical practices as complementary and alternative medicine (for detailed strategies, refer to http://www.nlm.nih.gov/bsd/pubmed_subsets/comp_med_strategy.html ) [ 59 ]. Second, the tensions are political and cultural as well. In the social constructionist view of medicine and medicine-as-profession [ 60 , 61 ], any brand of medicine employs the process of political and cultural negotiation. Biomedicine is historically negotiated as science and progressive modernity, whereas alternative medicine is portrayed as non-science, magic, and a backward tradition [ 62 – 64 ]. In colonial and postcolonial contexts, biomedicine is aligned with global domination projects, whereas alternative medicine is often identified with local autonomy [ 19 , 65 ]. These medical-theoretical and political-cultural tensions are evident in the behavior of medical service users and professional practitioners. Cancer patients who use both biomedicine and alternative medicine are often caught up between two conflicting theories, or biomedicine's depersonalized approach and alternative medicine's individualized approach [ 66 ]. When patients reclaim their control over medical decision-making through the use of alternative medicine, they are not always confident about their knowledge base and are often dependent on the advice of biomedical professionals [ 67 ]. Patients' orientation toward spiritual healing comes into conflict with physicians' treatment regimens [ 68 , 69 ]. Users of alternative medicine do not readily discuss their use of alternative medicine with their physicians, since physicians are often found to ignore alternative medicine [ 70 , 71 ]. These tensions between biomedicine and alternative medicine, however, do not create the same difficulties in different plural medical systems. Health behaviors and outcomes vary widely, depending on how these tensions are articulated in the system. For instance, national political contexts heighten or mitigate these tensions, producing different consequences. Post-Apartheid nationalist politics in South Africa, that portray alternative medicine as an indigenous African tradition and biomedicine as "Western White" science, is accused of strengthening inter-professional barriers in HIV treatment between medical doctors who provide drug therapy and traditional healers who practice herbal and spiritual healing [ 19 ]. It is also blamed for creating unnecessary social criticism on HIV patients that prevents them from taking an effective biomedical drug therapy (ART). In other African countries (e.g. Uganda, Kenya, and Tanzania) whose political environments promote alternative medicine education for biomedical practitioners and the cross-referral of patients between biomedicine and alternative medicine practitioners, HIV patients take the drug therapy for controlling the viral loads and, simultaneously, use herbs and spiritual healing for their immune systems and emotional well-being [ 38 , 53 ]. Micro-level therapeutic environments matter, as well. For instance, Steve Jobs spent the first nine months of his diagnosis with pancreatic cancer trying alternative medicine, such as diets, fruit juices, acupuncture, herbal remedies, and spiritual practices [ 72 ]. Disagreement over his behavior aside, his personal account exemplifies that alternative medicine users outside institutionalized settings often go through difficult responses from families, friends, and even doctors who are "infuriated" and "distressed" by their reliance on alternative medicine. Thus, patients often have to "search the Internet" to get information about alternative medicine. The use of alternative medicine becomes a lonely and secretive process. In another context as in integrative cancer care centers, however, plural medical practices are open and communicative between patients, families, and practitioners [ 73 , 74 ]. Hypotheses and two complementary analyses Therefore, this paper hypothesizes that tensions between biomedicine and alternative medicine and their management in institutional settings are consequential for health behavior [ 75 , 76 ] and outcomes [ 59 ]. It conducts two complementary analyses: one at the macro level with cross-national panel data, testing two hypotheses on the changing relationships between plural medical systems and national health outcomes; the other at the micro level with data extracted from medical journal articles in the U.S. and Japan, elaborating how macro-level patterns are related to the treatment-level evidence. The macro-level analysis takes two steps. First, it examines how the plurality of a national medical system affects health outcomes. It is hypothesized that plurality will affect health outcomes negatively as tensions between biomedicine and alternative medicine produce strains on the behaviors of users and practitioners ( Hypothesis 1 ). Strains on health behaviors are identifiable in the delayed utilization of relevant medical resources, whether they are biomedicine or alternative medicine [ 77 , 78 ], the non-utilization of these relevant resources [ 58 , 79 ], and the ill-informed/ill-sequenced combinations of biomedicine and alternative medicine that lead to adverse interactions between the interventions [ 80 ]. Delayed utilization, non-utilization, and ill-informed utilization of medical resources, in turn, affect health care outcomes adversely. Second, this paper examines how measures that ameliorate tensions between biomedicine and alternative medicine change the relationship between plurality and health outcomes. It hypothesizes that measures of tension reduction will weaken the negative effect of plurality and convert it into a positive one, since reduced tensions help resolve the strains of plural logics and generate their resourcefulness through treatment synergies between biomedicine and alternative medicine ( Hypothesis 2 ). In particular, the paper examines government funding for alternative medicine and medical knowledge production as two promising measures of coordination mechanism that reduces the tensions. Government funding for alternative medicine refers to the existence of public financial support for alternative medicine through government insurance coverage or free provisions of alternative medicine in public health facilities. Either through direct service provisions or rules for insurance reimbursement, governments provide specific ways to practice alternative medicine along with biomedicine, thus reducing biomedicine–alternative medicine tensions and adverse outcomes. Medical knowledge production refers to the annual number of medical journal articles produced by each country. It measures the extent to which information on medical interventions is circulated within the medical community. In countries producing more papers, the communication between biomedicine and alternative medicine is greater and helps inform medical practitioners on both sides, thus lowering professional barriers between the two sides. The second-step micro-level analysis conducts a content analysis to elaborate different ways to coordinate plural medical systems and the extent to which plurality coordination moderates the effects of plural medical systems [ 81 ] in the U.S. and Japan. This analysis reveals the medical treatment-level evidence on biomedicine–alternative medicine tensions, their institutional environments, and treatment outcomes. Subsequent multivariate regression models of the results of this content analysis ultimately demonstrate that cross-national differences in the treatment outcomes are explained by cross-national differences in biomedicine–alternative medicine tensions and their coordination. This analysis examines the U.S. and Japan as a comparative pair, since they feature relatively pluralist medical systems [ 33 , 56 ] and provide a clear contrast in the institutionalization of the plural systems [ 59 ]. The U.S. features a market-based decentralized and informal coordination of biomedicine and alternative medicine [ 82 , 83 ], whereas Japan represents a government-driven centralized formal coordination [ 84 , 85 ]. This contrast makes it promising to examine how different institutional settings are aligned with treatment outcomes. Analysis 1: cross-national panel data Data and measures This paper uses an unbalanced panel dataset of 246 observations for 97 countries for three time points ( S1 File ). The dependent variable, national health outcomes, is measured by life expectancy at birth in 1995, 2000, and 2005 [ 86 ]. Independent variables, including the key variable of medical plurality, are measured five years earlier for these years, in order to allow the delay in time when changes in independent variables are reflected in the dependent variable. Medical plurality is defined as the extent to which a national medical system is composed of two different elements: biomedicine and alternative medicine. To quantify this quality, this paper adopts the idea commonly applied to measuring racial/ethnic diversity [ 24 , 87 ]. The literature quantifies societal diversity with the following formula: 1 − ∑ n i ( n i − 1 ) N ( N − 1 ) 1 − 1 K ( w h e r e N : t o t a l p o p u l a t i o n , K : t h e n u m b e r o f s u b g r o u p s , n i : p o p u l a t i o n i n s u b g r o u p i ) It refers to the probability that two individuals chosen randomly and independently from the total population do not belong to the same racial/ethnic subgroup. The minimum value 0 indicates that no pair of individuals belongs to different subgroups; the maximum value 1 indicates that every random pair is composed of individuals from two different subgroups. This paper defines a medical plurality index with the following formula. Since the number of subgroups (K) is 2, the preceding formula becomes simplified and refers to the probability that a randomly chosen pair of medical resources belong to two different subgroups: biomedicine (n 1 ) and alternative medicine (n 2 ). [ 1 − ∑ n i ( n i − 1 ) N ( N − 1 ) ] × 2 ( w h e r e i = 1 , 2 ) The amounts of biomedical resources (n 1 ), alternative medicine resources (n 2 ), and total medical resources (N = n 1 + n 2 ) are approximated from the medicine section of the Yearbook of International Organizations [ 88 ]. This yearbook lists international professional medical organizations with their member states and organizational aims/activities. Thus, it is possible to compute for each country the number of memberships in biomedicine organizations (n 1 ), the number of memberships in alternative medicine organizations (n 2 ), and their sum (N) (for further detail, see Text 1 in S2 File and Table A in S2 File ). The mechanism of plurality coordination that reduces tensions between biomedicine and alternative medicine is first measured by government funding for alternative medicine [ 33 , 89 ]. It is an indicator variable coded 1 if there is a government financial provision in any of the following three manners: (1) government insurance coverage of alternative medicine therapies; (2) free-of-charge provisions of the therapies in public health facilities; (3) other public financial support. The second measure of plurality coordination is examined in terms of medical knowledge production . It measures the total number of medical journal articles that each country produces per year. To get this measure, this paper links medical journal articles indexed in the U.S. National Library of Medicine's MEDLINE database to authors' institutional affiliation data in Thomson Reuters' Web of Science [ 90 ]. By tagging the geographical information of author institutions to papers, it assigns each of these papers to one or more countries. In order to specify the effect of the medical plurality index, net of the impacts of the amounts of different medical organizational memberships, this paper controls for alternative medicine organizational membership rate , biomedicine organizational membership rate , and the total organizational membership rate . These are computed by the counts of organizational memberships divided by national population (per million people). With these control variables, this paper distinguishes the effect of medical plurality from that of the amount of medical resources approximated in the numbers of organizational memberships in biomedicine, alternative medicine, and both. This paper controls for additional variables, such as economic development [ 91 ], education [ 92 ], and income inequality [ 93 ]. Economic development is measured by GDP per capita , education by the average total years of schooling for people aged 25 or more [ 94 ], and income inequality by the standardized Gini coefficients of income [ 95 ]. Models Under the panel data structure, the random effects model (REM) and the fixed effects model (FEM) are used [ 96 ]. In order to adjust for the unmeasured temporal changes, the models incorporate two dummy variables for 1995 and 2000 against 1990. Hypothesis 1 is tested by: L i f e E x p e c t a n c y i ( t + 5 ) = Y e a r D u m m i e s + M e d i c a l P l u r a l i t y I n d e x i t + A m o u n t s o f M e d i c a l R e s o u r c e s i t + S o c i o e c o n o m i c C o n d i t i o n s i t + a i + ε i ( t + 5 ) (Eq 1) where i indicates each country; t indicates year; a i indicates a time-constant unit-specific effect for country i ; ɛ i(t+5) is a random error term for country i in year t+5 . This equation aims to specify regression coefficients for Medical Plurality Index it . To test Hypothesis 2 , this paper adds interaction terms between the medical plurality index and the measures of plurality coordination whose regression coefficients are the analytical focus: L i f e E x p e c t a n c y i ( t + 5 ) = Y e a r D u m m i e s + M e d i c a l P l u r a l i t y I n d e x i t + A m o u n t s o f M e d i c a l R e s o u r c e s i t + P l u r a l i t y C o o r d i n a t i o n i t + M e d i c a l P l u r a l i t y I n d e x i t × P l u r a l i t y C o o r d i n a t i o n i t + S o c i o e c o n o m i c C o n d i t i o n s i t + a i + ε i ( t + 5 ) (Eq 2) FEM estimates are reported (for REM estimates, see Table B in S2 File and Table C in S2 File ). Statistically, the Hausman test reports significant differences between FEM estimates and REM estimates. This is reasonable because estimates in FEM and REM tend to be different in a dataset with a few time points like the dataset of this paper, whereas they tend to become similar with more time points. FEM estimates are preferred for several reasons. Since my units of analysis–i.e. countries–are hardly a random sample drawn from a larger population, it makes sense theoretically to interpret "each as a separate intercept" for each country, rather than a random variable [ 96 ]. In addition, FEM estimates do not rely on REM's stricter assumption of the independence between independent variables and unobservable country-specific effects. Results Fig 1 graphically represents how life expectancy at year t+5 changes as the medical plurality index at year t (= 1990, 1995, and 2000) changes. Moving from bottom to top, each line refers to a country's trajectory in the past five years. Positive slopes mean that life expectancy increases as medical plurality index increases. Negative slopes refer to the opposite development. It is certain in this historical representation that the relationships between medical plurality and life expectancy are not unidirectional. The first model specification ( Eq 1 ) aims to find a dominant relationship, positive or negative, whereas the second ( Eq 2 ) is geared to identify conditions in which the dominant relationship changes into the other direction. 10.1371/journal.pone.0189841.g001 Fig 1 Historical scatterplot of life expectancy at birth over medical plurality index across years with lines connecting observations of a country: 246 observations for 97 countries across 1990, 1995, and 2000 (dotted lines for 18 OECD countries; solid lines for the others). The FEM estimates for Eq 1 ( Table 1 ) show that medical plurality has a negative effect on life expectancy (Model 1). Against the suspicion that this negative effect is simply confounded by the negative effect of the allegedly invalid alternative medicine and the positive effect of the allegedly valid biomedicine, the independent effects of biomedicine and alternative medicine are further controlled in Models 2 to 4. The negative effect of medical plurality still remains significant in these models. It remains so with even more controls in Models 5 to 8. 10.1371/journal.pone.0189841.t001 Table 1 Unstandardized coefficients from the fixed effects models of life expectancy regressed on medical plurality index and control variables. Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8 Year Dummies (Reference = 1990) 1995 1.29 ** 1.23 ** 1.19 ** 1.22 ** 1.35 ** 1.37 ** 1.32 ** 1.36 ** (0.18) (0.19) (0.20) (0.19) (0.27) (0.28) (0.28) (0.28) 2000 2.38 ** 2.24 ** 2.18 ** 2.22 ** 2.39 ** 2.43 ** 2.34 ** 2.42 ** (0.21) (0.29) (0.30) (0.29) (0.45) (0.49) (0.50) (0.49) Medical Plurality Index (MPI) -5.75 * -5.58 * -6.31 ** -5.65 * -5.19 + -5.25 + -5.41 + -5.20 + (2.35) (2.35) (2.40) (2.33) (2.71) (2.71) (2.90) (2.71) Control Variables: Level of Organizational Memberships Biomedicine Organizational Memberships 0.05 -0.02 (per million people) (0.04) (0.04) A.M. Organizational Memberships 0.45 0.15 (per million people) (0.27) (0.28) Total Organizational Memberships 0.05 -0.01 (per million people) (0.03) (0.04) Control Variables: Socio-economic Conditions GDP per capita (in hundreds) 0.01 * 0.01 + 0.01 + 0.01 + (0.01) (0.01) (0.01) (0.01) Standardized Gini Coefficient of Income 0.004 0.004 0.003 0.004 (0.06) (0.06) (0.06) (0.06) Years of Education -0.46 -0.48 -0.44 -0.47 (0.82) (0.83) (0.83) (0.83) Constant 71.3 ** 70.8 ** 71.0 ** 70.8 ** 72.4 ** 72.5 ** 72.4 ** 72.5 ** (1.29) (1.30) (1.22) (1.28) (5.17) (5.28) (5.20) (5.26) Observations 246 246 246 246 246 246 246 246 Number of Countries 97 97 97 97 97 97 97 97 R-squared 0.40 0.40 0.41 0.40 0.42 0.42 0.42 0.42 Note : Robust standard errors in parentheses + significant at 10% * significant at 5% ** significant at 1% (two-tailed tests). This result supports the notion in Hypothesis 1 that actors are not presumably skilled to handle conflicting logics, such that plural logics produce strains on health behaviors and result in the ill-informed or limited utilization of available plural medical resources. Through users' ill-informed utilization or non-utilization of available plural resources, plurality affects health outcomes negatively. In Model 8, one standard deviation increase in medical plurality index (0.15) leads to a decrease of 0.75 years in life expectancy. This identification of the dominant pattern needs to be complemented with an additional account for positive lines in Fig 1 . This is achieved by Eq 2 ( Table 2 ) that specifies conditions in which the dominant negative effect of plurality changes into a positive one. 10.1371/journal.pone.0189841.t002 Table 2 Unstandardized coefficients from the fixed effects models of life expectancy regressed on interaction variables with medical plurality index and control variables. Model 1 Model 2 Model 3 Model 4 Model 5 b ) Model 6 b ) Year Dummies (Reference = 1990) 1995 1.203 ** 1.323 ** 1.236 ** 1.412 ** 1.254 ** 1.427 ** (0.201) (0.294) (0.203) (0.294) (0.213) (0.352) 2000 2.223 ** 2.383 ** 2.233 ** 2.495 ** 2.375 ** 2.594 ** (0.298) (0.506) (0.312) (0.513) (0.309) (0.595) Medical Plurality Index (MPI) -6.045 * -5.647 + -5.716 * -5.323 + -9.903 ** -8.862 + (2.470) (2.899) (2.387) (2.761) (3.702) (4.664) Control Variables: Level of Organizational Memberships Total Organizational Memberships 0.054 -0.017 0.055 -0.034 0.040 -0.016 (per million people) (0.035) (0.038) (0.037) (0.046) (0.033) (0.034) OECD18 Dummy a ) (1 for OECD18 countries; 0 for else) MPI × OECD18 Dummy 6.236 * 6.253 + (3.045) (3.489) Number of Medical Journal Papers (in thousands) -0.050 -0.095 (0.059) (0.078) MPI × Number of Medical Papers 0.079 0.117 (0.092) (0.120) Government Funding for A.M. a ) (1 for funding; 0 for else) MPI × Funding for A.M. 10.174 * 8.895 + (4.243) (5.033) Control Variables: Socio-economic Conditions GDP per capita (in hundreds) 0.019 * 0.023 * 0.014 (0.010) (0.011) (0.010) Standardized Gini Coefficient of Income 0.004 0.005 -0.018 (0.065) (0.066) (0.082) Years of Education -0.436 -0.529 -0.424 (0.850) (0.850) (0.973) Constant 70.290 ** 71.699 ** 70.885 ** 72.753 ** 71.374 ** 73.902 ** (1.081) (5.301) (1.313) (5.312) (1.033) (6.764) Observations 246 246 246 246 203 203 Number of Countries 97 97 97 97 76 76 R-squared 0.41 0.43 0.40 0.42 0.52 0.54 Note : Robust standard errors in parentheses + significant at 10% * significant at 5% ** significant at 1% (two-tailed tests) a) Time-invariant indicator variables for countries are automatically dropped out of the FEM models b) It is based on a subset of 203 observations with data on government funding for alternative medicine (A.M.). When I specify all models in Tables 1 and 2 among this subset of 203 observations, the findings agree with those reported here. Results for the subset are reported in Table D in S2 File and Table E in S2 File ). In Fig 1 , positive slopes exist more among countries with high life expectancies than those with low life expectancies. Most cases with high life expectancies are affluent countries. Indeed, the estimate of the interaction term between medical plurality and OECD 18 dummy in Model 1 indicates that the effect of medical plurality on life expectancy is positive among the 18 OECD countries and negative among the non-OECD countries. Interestingly, this positive OECD effect remains the same even when socio-economic conditions are controlled (Model 2). As hypothesized, factors other than these socio-economic control variables explain the difference in the remaining four models in Table 2 . Models 3 and 4 reveal a positive interaction effect between medical knowledge production and medical plurality. As countries produce more medical research papers, the negative effect of medical plurality on life expectancy decreases and, finally, converts into a positive one. It seems that medical knowledge production plays the role of plurality coordination. Since the interaction effect is not statistically significant, however, Hypothesis 2 cannot be univocally supported based solely on this measure. Models 5 and 6, on the other hand, support the coordination hypothesis at a statistically significant level. The models report a significant positive interaction effect between government funding for alternative medicine and medical plurality. Government funding for alternative medicine is an important coordination mechanism through which the negative effect of medical plurality on life expectancy turns positive. In Model 6, one standard deviation increase in medical plurality index (0.15) leads to a 0.005 year gain in life expectancy among countries with government funding. On the contrary, the same increase in medical plurality leads to a 1.32 year loss in life expectancy among countries without such funding. The policy contexts in which medical plurality is practiced make such a striking difference. Data and measures This paper uses an unbalanced panel dataset of 246 observations for 97 countries for three time points ( S1 File ). The dependent variable, national health outcomes, is measured by life expectancy at birth in 1995, 2000, and 2005 [ 86 ]. Independent variables, including the key variable of medical plurality, are measured five years earlier for these years, in order to allow the delay in time when changes in independent variables are reflected in the dependent variable. Medical plurality is defined as the extent to which a national medical system is composed of two different elements: biomedicine and alternative medicine. To quantify this quality, this paper adopts the idea commonly applied to measuring racial/ethnic diversity [ 24 , 87 ]. The literature quantifies societal diversity with the following formula: 1 − ∑ n i ( n i − 1 ) N ( N − 1 ) 1 − 1 K ( w h e r e N : t o t a l p o p u l a t i o n , K : t h e n u m b e r o f s u b g r o u p s , n i : p o p u l a t i o n i n s u b g r o u p i ) It refers to the probability that two individuals chosen randomly and independently from the total population do not belong to the same racial/ethnic subgroup. The minimum value 0 indicates that no pair of individuals belongs to different subgroups; the maximum value 1 indicates that every random pair is composed of individuals from two different subgroups. This paper defines a medical plurality index with the following formula. Since the number of subgroups (K) is 2, the preceding formula becomes simplified and refers to the probability that a randomly chosen pair of medical resources belong to two different subgroups: biomedicine (n 1 ) and alternative medicine (n 2 ). [ 1 − ∑ n i ( n i − 1 ) N ( N − 1 ) ] × 2 ( w h e r e i = 1 , 2 ) The amounts of biomedical resources (n 1 ), alternative medicine resources (n 2 ), and total medical resources (N = n 1 + n 2 ) are approximated from the medicine section of the Yearbook of International Organizations [ 88 ]. This yearbook lists international professional medical organizations with their member states and organizational aims/activities. Thus, it is possible to compute for each country the number of memberships in biomedicine organizations (n 1 ), the number of memberships in alternative medicine organizations (n 2 ), and their sum (N) (for further detail, see Text 1 in S2 File and Table A in S2 File ). The mechanism of plurality coordination that reduces tensions between biomedicine and alternative medicine is first measured by government funding for alternative medicine [ 33 , 89 ]. It is an indicator variable coded 1 if there is a government financial provision in any of the following three manners: (1) government insurance coverage of alternative medicine therapies; (2) free-of-charge provisions of the therapies in public health facilities; (3) other public financial support. The second measure of plurality coordination is examined in terms of medical knowledge production . It measures the total number of medical journal articles that each country produces per year. To get this measure, this paper links medical journal articles indexed in the U.S. National Library of Medicine's MEDLINE database to authors' institutional affiliation data in Thomson Reuters' Web of Science [ 90 ]. By tagging the geographical information of author institutions to papers, it assigns each of these papers to one or more countries. In order to specify the effect of the medical plurality index, net of the impacts of the amounts of different medical organizational memberships, this paper controls for alternative medicine organizational membership rate , biomedicine organizational membership rate , and the total organizational membership rate . These are computed by the counts of organizational memberships divided by national population (per million people). With these control variables, this paper distinguishes the effect of medical plurality from that of the amount of medical resources approximated in the numbers of organizational memberships in biomedicine, alternative medicine, and both. This paper controls for additional variables, such as economic development [ 91 ], education [ 92 ], and income inequality [ 93 ]. Economic development is measured by GDP per capita , education by the average total years of schooling for people aged 25 or more [ 94 ], and income inequality by the standardized Gini coefficients of income [ 95 ]. Models Under the panel data structure, the random effects model (REM) and the fixed effects model (FEM) are used [ 96 ]. In order to adjust for the unmeasured temporal changes, the models incorporate two dummy variables for 1995 and 2000 against 1990. Hypothesis 1 is tested by: L i f e E x p e c t a n c y i ( t + 5 ) = Y e a r D u m m i e s + M e d i c a l P l u r a l i t y I n d e x i t + A m o u n t s o f M e d i c a l R e s o u r c e s i t + S o c i o e c o n o m i c C o n d i t i o n s i t + a i + ε i ( t + 5 ) (Eq 1) where i indicates each country; t indicates year; a i indicates a time-constant unit-specific effect for country i ; ɛ i(t+5) is a random error term for country i in year t+5 . This equation aims to specify regression coefficients for Medical Plurality Index it . To test Hypothesis 2 , this paper adds interaction terms between the medical plurality index and the measures of plurality coordination whose regression coefficients are the analytical focus: L i f e E x p e c t a n c y i ( t + 5 ) = Y e a r D u m m i e s + M e d i c a l P l u r a l i t y I n d e x i t + A m o u n t s o f M e d i c a l R e s o u r c e s i t + P l u r a l i t y C o o r d i n a t i o n i t + M e d i c a l P l u r a l i t y I n d e x i t × P l u r a l i t y C o o r d i n a t i o n i t + S o c i o e c o n o m i c C o n d i t i o n s i t + a i + ε i ( t + 5 ) (Eq 2) FEM estimates are reported (for REM estimates, see Table B in S2 File and Table C in S2 File ). Statistically, the Hausman test reports significant differences between FEM estimates and REM estimates. This is reasonable because estimates in FEM and REM tend to be different in a dataset with a few time points like the dataset of this paper, whereas they tend to become similar with more time points. FEM estimates are preferred for several reasons. Since my units of analysis–i.e. countries–are hardly a random sample drawn from a larger population, it makes sense theoretically to interpret "each as a separate intercept" for each country, rather than a random variable [ 96 ]. In addition, FEM estimates do not rely on REM's stricter assumption of the independence between independent variables and unobservable country-specific effects. Results Fig 1 graphically represents how life expectancy at year t+5 changes as the medical plurality index at year t (= 1990, 1995, and 2000) changes. Moving from bottom to top, each line refers to a country's trajectory in the past five years. Positive slopes mean that life expectancy increases as medical plurality index increases. Negative slopes refer to the opposite development. It is certain in this historical representation that the relationships between medical plurality and life expectancy are not unidirectional. The first model specification ( Eq 1 ) aims to find a dominant relationship, positive or negative, whereas the second ( Eq 2 ) is geared to identify conditions in which the dominant relationship changes into the other direction. 10.1371/journal.pone.0189841.g001 Fig 1 Historical scatterplot of life expectancy at birth over medical plurality index across years with lines connecting observations of a country: 246 observations for 97 countries across 1990, 1995, and 2000 (dotted lines for 18 OECD countries; solid lines for the others). The FEM estimates for Eq 1 ( Table 1 ) show that medical plurality has a negative effect on life expectancy (Model 1). Against the suspicion that this negative effect is simply confounded by the negative effect of the allegedly invalid alternative medicine and the positive effect of the allegedly valid biomedicine, the independent effects of biomedicine and alternative medicine are further controlled in Models 2 to 4. The negative effect of medical plurality still remains significant in these models. It remains so with even more controls in Models 5 to 8. 10.1371/journal.pone.0189841.t001 Table 1 Unstandardized coefficients from the fixed effects models of life expectancy regressed on medical plurality index and control variables. Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8 Year Dummies (Reference = 1990) 1995 1.29 ** 1.23 ** 1.19 ** 1.22 ** 1.35 ** 1.37 ** 1.32 ** 1.36 ** (0.18) (0.19) (0.20) (0.19) (0.27) (0.28) (0.28) (0.28) 2000 2.38 ** 2.24 ** 2.18 ** 2.22 ** 2.39 ** 2.43 ** 2.34 ** 2.42 ** (0.21) (0.29) (0.30) (0.29) (0.45) (0.49) (0.50) (0.49) Medical Plurality Index (MPI) -5.75 * -5.58 * -6.31 ** -5.65 * -5.19 + -5.25 + -5.41 + -5.20 + (2.35) (2.35) (2.40) (2.33) (2.71) (2.71) (2.90) (2.71) Control Variables: Level of Organizational Memberships Biomedicine Organizational Memberships 0.05 -0.02 (per million people) (0.04) (0.04) A.M. Organizational Memberships 0.45 0.15 (per million people) (0.27) (0.28) Total Organizational Memberships 0.05 -0.01 (per million people) (0.03) (0.04) Control Variables: Socio-economic Conditions GDP per capita (in hundreds) 0.01 * 0.01 + 0.01 + 0.01 + (0.01) (0.01) (0.01) (0.01) Standardized Gini Coefficient of Income 0.004 0.004 0.003 0.004 (0.06) (0.06) (0.06) (0.06) Years of Education -0.46 -0.48 -0.44 -0.47 (0.82) (0.83) (0.83) (0.83) Constant 71.3 ** 70.8 ** 71.0 ** 70.8 ** 72.4 ** 72.5 ** 72.4 ** 72.5 ** (1.29) (1.30) (1.22) (1.28) (5.17) (5.28) (5.20) (5.26) Observations 246 246 246 246 246 246 246 246 Number of Countries 97 97 97 97 97 97 97 97 R-squared 0.40 0.40 0.41 0.40 0.42 0.42 0.42 0.42 Note : Robust standard errors in parentheses + significant at 10% * significant at 5% ** significant at 1% (two-tailed tests). This result supports the notion in Hypothesis 1 that actors are not presumably skilled to handle conflicting logics, such that plural logics produce strains on health behaviors and result in the ill-informed or limited utilization of available plural medical resources. Through users' ill-informed utilization or non-utilization of available plural resources, plurality affects health outcomes negatively. In Model 8, one standard deviation increase in medical plurality index (0.15) leads to a decrease of 0.75 years in life expectancy. This identification of the dominant pattern needs to be complemented with an additional account for positive lines in Fig 1 . This is achieved by Eq 2 ( Table 2 ) that specifies conditions in which the dominant negative effect of plurality changes into a positive one. 10.1371/journal.pone.0189841.t002 Table 2 Unstandardized coefficients from the fixed effects models of life expectancy regressed on interaction variables with medical plurality index and control variables. Model 1 Model 2 Model 3 Model 4 Model 5 b ) Model 6 b ) Year Dummies (Reference = 1990) 1995 1.203 ** 1.323 ** 1.236 ** 1.412 ** 1.254 ** 1.427 ** (0.201) (0.294) (0.203) (0.294) (0.213) (0.352) 2000 2.223 ** 2.383 ** 2.233 ** 2.495 ** 2.375 ** 2.594 ** (0.298) (0.506) (0.312) (0.513) (0.309) (0.595) Medical Plurality Index (MPI) -6.045 * -5.647 + -5.716 * -5.323 + -9.903 ** -8.862 + (2.470) (2.899) (2.387) (2.761) (3.702) (4.664) Control Variables: Level of Organizational Memberships Total Organizational Memberships 0.054 -0.017 0.055 -0.034 0.040 -0.016 (per million people) (0.035) (0.038) (0.037) (0.046) (0.033) (0.034) OECD18 Dummy a ) (1 for OECD18 countries; 0 for else) MPI × OECD18 Dummy 6.236 * 6.253 + (3.045) (3.489) Number of Medical Journal Papers (in thousands) -0.050 -0.095 (0.059) (0.078) MPI × Number of Medical Papers 0.079 0.117 (0.092) (0.120) Government Funding for A.M. a ) (1 for funding; 0 for else) MPI × Funding for A.M. 10.174 * 8.895 + (4.243) (5.033) Control Variables: Socio-economic Conditions GDP per capita (in hundreds) 0.019 * 0.023 * 0.014 (0.010) (0.011) (0.010) Standardized Gini Coefficient of Income 0.004 0.005 -0.018 (0.065) (0.066) (0.082) Years of Education -0.436 -0.529 -0.424 (0.850) (0.850) (0.973) Constant 70.290 ** 71.699 ** 70.885 ** 72.753 ** 71.374 ** 73.902 ** (1.081) (5.301) (1.313) (5.312) (1.033) (6.764) Observations 246 246 246 246 203 203 Number of Countries 97 97 97 97 76 76 R-squared 0.41 0.43 0.40 0.42 0.52 0.54 Note : Robust standard errors in parentheses + significant at 10% * significant at 5% ** significant at 1% (two-tailed tests) a) Time-invariant indicator variables for countries are automatically dropped out of the FEM models b) It is based on a subset of 203 observations with data on government funding for alternative medicine (A.M.). When I specify all models in Tables 1 and 2 among this subset of 203 observations, the findings agree with those reported here. Results for the subset are reported in Table D in S2 File and Table E in S2 File ). In Fig 1 , positive slopes exist more among countries with high life expectancies than those with low life expectancies. Most cases with high life expectancies are affluent countries. Indeed, the estimate of the interaction term between medical plurality and OECD 18 dummy in Model 1 indicates that the effect of medical plurality on life expectancy is positive among the 18 OECD countries and negative among the non-OECD countries. Interestingly, this positive OECD effect remains the same even when socio-economic conditions are controlled (Model 2). As hypothesized, factors other than these socio-economic control variables explain the difference in the remaining four models in Table 2 . Models 3 and 4 reveal a positive interaction effect between medical knowledge production and medical plurality. As countries produce more medical research papers, the negative effect of medical plurality on life expectancy decreases and, finally, converts into a positive one. It seems that medical knowledge production plays the role of plurality coordination. Since the interaction effect is not statistically significant, however, Hypothesis 2 cannot be univocally supported based solely on this measure. Models 5 and 6, on the other hand, support the coordination hypothesis at a statistically significant level. The models report a significant positive interaction effect between government funding for alternative medicine and medical plurality. Government funding for alternative medicine is an important coordination mechanism through which the negative effect of medical plurality on life expectancy turns positive. In Model 6, one standard deviation increase in medical plurality index (0.15) leads to a 0.005 year gain in life expectancy among countries with government funding. On the contrary, the same increase in medical plurality leads to a 1.32 year loss in life expectancy among countries without such funding. The policy contexts in which medical plurality is practiced make such a striking difference. Analysis 2: content analysis comparing the U.S. and Japan Data and measures Content analysis purports to elaborate tensions between biomedicine and alternative medicine, plurality coordination, and their impacts on health care outcomes. It uses a prominent medical journal database MEDLINE [ 97 ] to collect treatment-level evidence of alternative medicine in different institutional settings in the U.S. and Japan. Using search terms in the Medical Subject Headings (MeSH), such as "the United States" and "Japan," this paper retrieves 30,588 papers based on trials in the U.S. and 2,993 papers based in Japan from MEDLINE's subset of alternative medicine ( http://www.nlm.nih.gov/bsd/pubmed_subsets/comp_med_strategy.html ). Then, it randomly samples 341 papers in the U.S. (1.1% of the total) and 279 papers in Japan (9.3% of the total) for content analysis. 29.6% of the sample in the U.S. (101 papers) and 28.7% of the sample in Japan (80 papers) are subsequently excluded because they are actually unrelated to alternative medicine. 56 more papers in the U.S. and 22 more papers in Japan are excluded, because they employ multiple outcome measures and report inconsistent results. This has led to a final sample of 184 papers from the U.S. and 177 from Japan ( Fig 2 ). 10.1371/journal.pone.0189841.g002 Fig 2 Sample selection process for the content analysis of medical journal papers. Using a paper as the unit of analysis, each paper is coded regarding what kind of alternative medicine intervention is used, for what medical conditions, and whether the intervention is found effective in treating the medical conditions. It is also coded whether there are biomedicine–alternative medicine tensions in the reported behaviors of users or practitioners, coordination efforts to reduce these tensions, and adverse-vs-synergic treatment interactions between biomedicine and alternative medicine interventions ( S3 File ). Results Compared to Japan, alternative medicine in the U.S. is found to be less coordinated with biomedicine. Correspondingly, alternative medicine in the U.S. is found less effective, more in tension with biomedicine, and less synergic with biomedicine. In detail, the percentage of papers identifying a deficiency of plurality coordination is much higher in the U.S. (17.9% vs. 2.8%; p < 0.01). On the contrary, the percentage of papers that verify the treatment effectiveness of alternative medicine is lower in the U.S. than Japan (42.4% vs. 60.5%; p < 0.01). Tensions between biomedicine and alternative medicine are reported more often in the U.S. (6.5% vs. 0.6%; p < 0.01). So are adverse interactions between biomedicine and alternative medicine (12.0% vs. 0.6%; p < 0.01). Synergic interactions are found less often in the U.S. (3.3% vs. 5.7%; p = 0.27). Across the U.S. and Japan, papers reporting that alternative medicine is not coordinated with biomedicine are less likely to report effective treatment outcomes from alternative medicine. The multivariate analysis of these results ( Table 3 ) also verifies this relationship between the deficiency of plurality coordination and the treatment outcomes of alternative medicine. In particular, models in Table 3 specify 1) the U.S.–Japan difference in the effectiveness of alternative medicine treatments and 2) the extent to which this difference is explained by the difference in plurality coordination. 10.1371/journal.pone.0189841.t003 Table 3 Log odds ratios from the Logit models of the effectiveness of alternative medicine a ) regressed on the deficiency of plurality coordination and control variables. Model 1 Model 2 Model 3 Location of Study b ) 0.731 ** 0.567 * 0.544 + (1 = Japan, 0 = US) (0.214) (0.221) (0.293) Plurality Coordination Deficiency c ) -1.301 ** -2.117 ** (1 = Reported, 0 = Not reported) (0.422) (0.511) Control Variables Journal of Publication d ) 0.914 * (1 = Alternative medicine journal, 0 = General medical journal) (0.368) Study Design e ) 2.490 ** (1 = RCT, 0 = Otherwise) (0.792) Modality f ) (Reference = Mind-Body Therapies) Alternative Whole Medical Systems 1.574 ** (0.525) Biologically Based Modalities 2.636 ** (0.397) Manipulation-based Modalities 3.649 ** (0.741) Other 3.074 ** (0.410) Constant -0.307 * -0.106 -2.568 ** (0.149) (0.161) (0.368) Observations (N of Papers) 361 361 361 Note : Standard errors in parentheses. + significant at 10% * significant at 5% ** significant at 1%. a) A binary variable coded 1 if alternative medical treatment is found effective in the paper and 0 otherwise. b) And indicator coded 1 for the paper based on trials in Japan and 0 for the U.S. c) An indicator coded 1 if an occasion of coordination deficiency is reported in the paper. d) An indicator coded 1 if the paper's publication outlet is a specialty journal of alternative medicine classified by MEDLINE. e) An indicator coded 1 if the paper is based on randomized controlled trials (RCT). f) A set of dummy variables indicating one of the NCCAM-generated five major modalities of alternative medicine [ 57 ] to which the specific treatment that the paper is testing belongs. Five modalities are "alternative whole medical system" involving acupuncture, Ayurveda, chelation therapy, traditional healers, alternative medical belief system, naturopathy, and homeopathy; "biologically based modalities" involving herbs, dietary supplements, vitamins/minerals, and special diets; "manipulation-based modalities" involving chiropractic, osteopathy, massage, anma, shiatsu, and movement therapies; "mind-body modalities" involving biofeedback, energy healing, hypnosis, yoga/taichi/qigong, relaxation, and psychotherapy; "other" involving spiritual/religious healing, self-care, and others. Papers reporting practices in Japan are 2.07 times more likely than those in the U.S. to report the effectiveness of alternative medicine ( e 0.731 = 2.07 in Model 1). As hypothesized, a substantial part of this association is then explained by the deficiency of plurality coordination (Model 2). As soon as the U.S.–Japan difference in plurality coordination is taken into account, the relatively higher odds of Japanese studies finding alternative medicine to be effective reduce to 1.76 (= e 0.567 ). This result remains unchanged even when Model 3 additionally controls for publication outlet, study design, and the modality of alternative medicine (see the Note in Table 3 for details of the variables). On the other hand, plurality coordination has a significant effect on treatment outcomes persistently across Models 2 and 3. When an alternative medicine treatment is found to be uncoordinated with biomedicine, it is less likely to produce effective treatment outcomes (odds ratio 0.12 = e -2.117 ). Content analysis further provides qualitative evidence underlying this multivariate pattern. Many papers in the sample explicitly attribute ineffective and adverse treatment outcomes to coordination deficiency. First, unregulated and defective products of alternative medicine are held accountable. One article reports two patient cases in the U.S. that developed liver injuries by unwittingly taking bacteria-contaminated herbal supplements [ 98 ]. Practitioner commentaries argue that herbal medicine presents challenges to the medical community because the processes of cultivation, harvest, and manufacturing are not properly regulated for quality control [ 99 , 100 ]. The 1994 U.S. Dietary Supplement and Health Education Act (DSHEA), which categorizes herbs and other botanicals as dietary supplements (and not drugs), was mostly criticized since it placed herbs outside of the regulation of the Food and Drug Acts [ 98 , 101 – 104 ]. In the Japanese sample, early whirlpool bath-tubs are reported to have caused a significant number of people to drown while bathing, due to problems with the safety devices [ 105 ]. Second, the reports hold insufficient practitioner knowledge of alternative medicine responsible. A systematic review of 124 patients suffering adverse events from acupuncture in 89 medical reports finds that 85% of these adverse events resulted from the negligence, ignorance, or malpractice of acupuncturists who lacked proper education and training [ 106 ]. In the U.S., MDs and DOs practice acupuncture only with 220 hours of training and become a full member of the American Academy of Medical Acupuncture (AAMA) without board examinations [ 107 ]. In addition, clinical practice guidelines (CPGs) have mostly been found inadequate to guide physicians on how to cope with the use of alternative medicine in clinical settings [ 108 ]. A case report of a Cambodian immigrant woman in the U.S. shows that physicians' unfamiliarity with alternative folk medicine that is often used among immigrant communities can result in an adverse treatment outcome [ 109 ]. The 73-year old Cambodian immigrant took a common blood thinner Warfarin after cardiac surgery for artificial heart valves. Later, she unexpectedly developed excessive bleedings recurrently for several years. Her doctor assumed that she did not comply with her medication or did not follow instructions to avoid a list of vitamin K foods that reportedly produce adverse interactions with Warfarin. As it turned out, however, the doctor was wrong. When the doctor had to visit the patient's home because the patient declined to come to the hospital, he unexpectedly found that the patient was taking bitter melon which is high in vitamin K and was not on the doctor's list of vitamin K foods. The patient was taking it from her own vegetable garden which is a common practice in her culture, concurrently with her prescribed medication. As soon as she stopped taking bitter melon, the bleeding problem was solved. The case report concludes with a note about the importance of plurality coordination through professional practitioners: "Our clinical dietary advice had been based on our own expectations of food availability and consumption. What is natural to others was clearly not readily apparent to us. Probing for cultural or dietary practices rather than simply prescribing may have uncovered the secret of the bitter melon much earlier" (highlights added). Third, the personal, informal, and commercial circulation of treatment information produces consequences as well. One article finds that, although the U.S. Dietary Supplement and Health Education Act regulated claims of "disease prevention, treatment, or cure" as unlawful, such claims were prevalent on commercial Internet sites of herbal supplements for cancers [ 110 ]. In this environment, consumers experienced difficulties in judging the quality of dietary products and services to treat obesity [ 111 ]. Another article stresses the importance of trained homeopathic practitioners, rather than misinformed self-care, in applying homeopathic drugs effectively [ 112 ]. Another article also finds that most of the Echinacea consumed in the U.S. is misused, based on its public misrepresentation as an effective cold and flu remedy, against its proven efficacy for treating infections [ 113 ]. At the same time, however, papers do report desirable synergic outcomes from the utilization of alternative medicine when it is coordinated with biomedical treatments. Treatment synergies have emerged from the text largely in three types. First, alternative medicine takes the place of conventional biomedical interventions that are less optimal, like acupuncture that was used for surgical analgesia that replaced drug analgesia producing perioperative side effects [ 114 , 115 ]. Second, some modalities of alternative medicine complement biomedical interventions to the extent that their absence would make biomedical interventions impossible, such as herbs and acupuncture for bodily and sexual energy during chemotherapy and hormone therapies among cancer patients [ 37 , 116 ] and religious/spiritual interventions to help patients to actively accept life even during extended biomedical treatments [ 41 ]. Third, alternative medicine is found to be simply additive to existing interventions, producing better outcomes. Several studies in Japan report the superior effectiveness of alternative medicine added to biomedical treatments, compared to the effectiveness of biomedical treatments alone, such as a Japanese herbal medicine ( Kampo ) added to Tamiflu for type A influenza [ 117 ] and acupuncture added to a conventional medication for chronic respiratory diseases (COPD) [ 36 ]. Data and measures Content analysis purports to elaborate tensions between biomedicine and alternative medicine, plurality coordination, and their impacts on health care outcomes. It uses a prominent medical journal database MEDLINE [ 97 ] to collect treatment-level evidence of alternative medicine in different institutional settings in the U.S. and Japan. Using search terms in the Medical Subject Headings (MeSH), such as "the United States" and "Japan," this paper retrieves 30,588 papers based on trials in the U.S. and 2,993 papers based in Japan from MEDLINE's subset of alternative medicine ( http://www.nlm.nih.gov/bsd/pubmed_subsets/comp_med_strategy.html ). Then, it randomly samples 341 papers in the U.S. (1.1% of the total) and 279 papers in Japan (9.3% of the total) for content analysis. 29.6% of the sample in the U.S. (101 papers) and 28.7% of the sample in Japan (80 papers) are subsequently excluded because they are actually unrelated to alternative medicine. 56 more papers in the U.S. and 22 more papers in Japan are excluded, because they employ multiple outcome measures and report inconsistent results. This has led to a final sample of 184 papers from the U.S. and 177 from Japan ( Fig 2 ). 10.1371/journal.pone.0189841.g002 Fig 2 Sample selection process for the content analysis of medical journal papers. Using a paper as the unit of analysis, each paper is coded regarding what kind of alternative medicine intervention is used, for what medical conditions, and whether the intervention is found effective in treating the medical conditions. It is also coded whether there are biomedicine–alternative medicine tensions in the reported behaviors of users or practitioners, coordination efforts to reduce these tensions, and adverse-vs-synergic treatment interactions between biomedicine and alternative medicine interventions ( S3 File ). Results Compared to Japan, alternative medicine in the U.S. is found to be less coordinated with biomedicine. Correspondingly, alternative medicine in the U.S. is found less effective, more in tension with biomedicine, and less synergic with biomedicine. In detail, the percentage of papers identifying a deficiency of plurality coordination is much higher in the U.S. (17.9% vs. 2.8%; p < 0.01). On the contrary, the percentage of papers that verify the treatment effectiveness of alternative medicine is lower in the U.S. than Japan (42.4% vs. 60.5%; p < 0.01). Tensions between biomedicine and alternative medicine are reported more often in the U.S. (6.5% vs. 0.6%; p < 0.01). So are adverse interactions between biomedicine and alternative medicine (12.0% vs. 0.6%; p < 0.01). Synergic interactions are found less often in the U.S. (3.3% vs. 5.7%; p = 0.27). Across the U.S. and Japan, papers reporting that alternative medicine is not coordinated with biomedicine are less likely to report effective treatment outcomes from alternative medicine. The multivariate analysis of these results ( Table 3 ) also verifies this relationship between the deficiency of plurality coordination and the treatment outcomes of alternative medicine. In particular, models in Table 3 specify 1) the U.S.–Japan difference in the effectiveness of alternative medicine treatments and 2) the extent to which this difference is explained by the difference in plurality coordination. 10.1371/journal.pone.0189841.t003 Table 3 Log odds ratios from the Logit models of the effectiveness of alternative medicine a ) regressed on the deficiency of plurality coordination and control variables. Model 1 Model 2 Model 3 Location of Study b ) 0.731 ** 0.567 * 0.544 + (1 = Japan, 0 = US) (0.214) (0.221) (0.293) Plurality Coordination Deficiency c ) -1.301 ** -2.117 ** (1 = Reported, 0 = Not reported) (0.422) (0.511) Control Variables Journal of Publication d ) 0.914 * (1 = Alternative medicine journal, 0 = General medical journal) (0.368) Study Design e ) 2.490 ** (1 = RCT, 0 = Otherwise) (0.792) Modality f ) (Reference = Mind-Body Therapies) Alternative Whole Medical Systems 1.574 ** (0.525) Biologically Based Modalities 2.636 ** (0.397) Manipulation-based Modalities 3.649 ** (0.741) Other 3.074 ** (0.410) Constant -0.307 * -0.106 -2.568 ** (0.149) (0.161) (0.368) Observations (N of Papers) 361 361 361 Note : Standard errors in parentheses. + significant at 10% * significant at 5% ** significant at 1%. a) A binary variable coded 1 if alternative medical treatment is found effective in the paper and 0 otherwise. b) And indicator coded 1 for the paper based on trials in Japan and 0 for the U.S. c) An indicator coded 1 if an occasion of coordination deficiency is reported in the paper. d) An indicator coded 1 if the paper's publication outlet is a specialty journal of alternative medicine classified by MEDLINE. e) An indicator coded 1 if the paper is based on randomized controlled trials (RCT). f) A set of dummy variables indicating one of the NCCAM-generated five major modalities of alternative medicine [ 57 ] to which the specific treatment that the paper is testing belongs. Five modalities are "alternative whole medical system" involving acupuncture, Ayurveda, chelation therapy, traditional healers, alternative medical belief system, naturopathy, and homeopathy; "biologically based modalities" involving herbs, dietary supplements, vitamins/minerals, and special diets; "manipulation-based modalities" involving chiropractic, osteopathy, massage, anma, shiatsu, and movement therapies; "mind-body modalities" involving biofeedback, energy healing, hypnosis, yoga/taichi/qigong, relaxation, and psychotherapy; "other" involving spiritual/religious healing, self-care, and others. Papers reporting practices in Japan are 2.07 times more likely than those in the U.S. to report the effectiveness of alternative medicine ( e 0.731 = 2.07 in Model 1). As hypothesized, a substantial part of this association is then explained by the deficiency of plurality coordination (Model 2). As soon as the U.S.–Japan difference in plurality coordination is taken into account, the relatively higher odds of Japanese studies finding alternative medicine to be effective reduce to 1.76 (= e 0.567 ). This result remains unchanged even when Model 3 additionally controls for publication outlet, study design, and the modality of alternative medicine (see the Note in Table 3 for details of the variables). On the other hand, plurality coordination has a significant effect on treatment outcomes persistently across Models 2 and 3. When an alternative medicine treatment is found to be uncoordinated with biomedicine, it is less likely to produce effective treatment outcomes (odds ratio 0.12 = e -2.117 ). Content analysis further provides qualitative evidence underlying this multivariate pattern. Many papers in the sample explicitly attribute ineffective and adverse treatment outcomes to coordination deficiency. First, unregulated and defective products of alternative medicine are held accountable. One article reports two patient cases in the U.S. that developed liver injuries by unwittingly taking bacteria-contaminated herbal supplements [ 98 ]. Practitioner commentaries argue that herbal medicine presents challenges to the medical community because the processes of cultivation, harvest, and manufacturing are not properly regulated for quality control [ 99 , 100 ]. The 1994 U.S. Dietary Supplement and Health Education Act (DSHEA), which categorizes herbs and other botanicals as dietary supplements (and not drugs), was mostly criticized since it placed herbs outside of the regulation of the Food and Drug Acts [ 98 , 101 – 104 ]. In the Japanese sample, early whirlpool bath-tubs are reported to have caused a significant number of people to drown while bathing, due to problems with the safety devices [ 105 ]. Second, the reports hold insufficient practitioner knowledge of alternative medicine responsible. A systematic review of 124 patients suffering adverse events from acupuncture in 89 medical reports finds that 85% of these adverse events resulted from the negligence, ignorance, or malpractice of acupuncturists who lacked proper education and training [ 106 ]. In the U.S., MDs and DOs practice acupuncture only with 220 hours of training and become a full member of the American Academy of Medical Acupuncture (AAMA) without board examinations [ 107 ]. In addition, clinical practice guidelines (CPGs) have mostly been found inadequate to guide physicians on how to cope with the use of alternative medicine in clinical settings [ 108 ]. A case report of a Cambodian immigrant woman in the U.S. shows that physicians' unfamiliarity with alternative folk medicine that is often used among immigrant communities can result in an adverse treatment outcome [ 109 ]. The 73-year old Cambodian immigrant took a common blood thinner Warfarin after cardiac surgery for artificial heart valves. Later, she unexpectedly developed excessive bleedings recurrently for several years. Her doctor assumed that she did not comply with her medication or did not follow instructions to avoid a list of vitamin K foods that reportedly produce adverse interactions with Warfarin. As it turned out, however, the doctor was wrong. When the doctor had to visit the patient's home because the patient declined to come to the hospital, he unexpectedly found that the patient was taking bitter melon which is high in vitamin K and was not on the doctor's list of vitamin K foods. The patient was taking it from her own vegetable garden which is a common practice in her culture, concurrently with her prescribed medication. As soon as she stopped taking bitter melon, the bleeding problem was solved. The case report concludes with a note about the importance of plurality coordination through professional practitioners: "Our clinical dietary advice had been based on our own expectations of food availability and consumption. What is natural to others was clearly not readily apparent to us. Probing for cultural or dietary practices rather than simply prescribing may have uncovered the secret of the bitter melon much earlier" (highlights added). Third, the personal, informal, and commercial circulation of treatment information produces consequences as well. One article finds that, although the U.S. Dietary Supplement and Health Education Act regulated claims of "disease prevention, treatment, or cure" as unlawful, such claims were prevalent on commercial Internet sites of herbal supplements for cancers [ 110 ]. In this environment, consumers experienced difficulties in judging the quality of dietary products and services to treat obesity [ 111 ]. Another article stresses the importance of trained homeopathic practitioners, rather than misinformed self-care, in applying homeopathic drugs effectively [ 112 ]. Another article also finds that most of the Echinacea consumed in the U.S. is misused, based on its public misrepresentation as an effective cold and flu remedy, against its proven efficacy for treating infections [ 113 ]. At the same time, however, papers do report desirable synergic outcomes from the utilization of alternative medicine when it is coordinated with biomedical treatments. Treatment synergies have emerged from the text largely in three types. First, alternative medicine takes the place of conventional biomedical interventions that are less optimal, like acupuncture that was used for surgical analgesia that replaced drug analgesia producing perioperative side effects [ 114 , 115 ]. Second, some modalities of alternative medicine complement biomedical interventions to the extent that their absence would make biomedical interventions impossible, such as herbs and acupuncture for bodily and sexual energy during chemotherapy and hormone therapies among cancer patients [ 37 , 116 ] and religious/spiritual interventions to help patients to actively accept life even during extended biomedical treatments [ 41 ]. Third, alternative medicine is found to be simply additive to existing interventions, producing better outcomes. Several studies in Japan report the superior effectiveness of alternative medicine added to biomedical treatments, compared to the effectiveness of biomedical treatments alone, such as a Japanese herbal medicine ( Kampo ) added to Tamiflu for type A influenza [ 117 ] and acupuncture added to a conventional medication for chronic respiratory diseases (COPD) [ 36 ]. Conclusions Findings from the two complementary analyses have demonstrated that plural logics in medicine are consequential for health behavior and health care outcomes. When coordinated, the two different medical logics of biomedicine and alternative medicine have together generated resourcefulness for medical service users and practitioners, leading to desirable health care outcomes. When uncoordinated, on the contrary, the plural logics have generated strains for health behavior and led to undesirable outcomes. This result invites a reconsideration of existing models of plural action logics in medical sociology in particular, and those in organizational and institutional sociology in general. The ethnographic notion about plural health behavior [ 52 ] that actors can somehow organize different elements of medical traditions within cultural orders should not overlook the fact that plural elements do not always co-exist easily. Even when these plural elements have come to an easy co-existence, this co-existence may most likely have come only after much difficulty and trouble, which is worth a close examination. This paper finds support for this argument from two recent efforts to re-conceptualize plural health behavior in terms of hybrid medical habitus, one among immigrant minorities in the U.S. [ 18 ] and the other among indigenous people in South Africa [ 19 ]. They both point to difficulty and trouble in plural health behavior, emphasizing the multiplicity and the incongruity of logics in plural medical systems. While resonating with these two studies, this paper newly adds that there are significant variations in the behavioral consequences of multiple and incongruent plural logics, depending on how these uneasy and conflicting logics are managed. In this respect, this paper draws on a comparative cross-national insight that individuals subscribing to plural medical logics tend to have varying expectations of alternative medicine, depending on the institutional settings into which alternative medicine is practiced [ 81 ]. In this regard, this paper further demonstrates that it is not only the subjective user expectations but also the objective health care outcomes that are influenced by institutional settings. Regarding the organizational and institutional sociology of plural action logics, this paper stresses that plural logics can have tensions with one another and actors are more or less skilled in dealing with these logics, depending on the social contexts in which plural logics and actors are situated. When it comes to the behavioral consequences of plural logics, therefore, this paper informs the simple resourcefulness account and the strains account of plural logics that it is problematic to monotonously assume either positive or negative consequences from plural logics. Instead, supra-individual social conditions need to be given analytical attention at multiple levels, such as government policies, insurance schemes, medical service organizations, and practitioners' medical expertise and communication skills. Contingent upon these conditions, plural logics are open to generating both resourcefulness and strains for action. To capture the significance of these conditions, this paper proposes plurality coordination as an important mechanism that generates the different behavioral consequences of plural logics. This proposition contributes to the existing literature in two ways. It provides a novel frame in which the divergent developments of plural logics between "settled" and "unsettled" times [ 13 ], between "senior" and "junior" scientists [ 11 ], between "resourceful" and "restrictive" museum environments [ 8 ], or between adolescents in "rich" and "poor" urban neighborhoods [ 12 ] that are discernible only to close readers in descriptive terms, can now be explicitly recognized and formally conceptualized. In addition, it also provides a meta-analytic perspective that can incorporate puzzling disagreements between empirical studies, for instance, why a similar set of plural logics are found to produce desirable outcomes in some study locales [ 38 , 53 ] and not in others [ 19 ]. Lastly, this paper provides implications for the professional practice of medicine which in not a few countries has to engage with medical service users who rely on both biomedicine and alternative medicine at the same time. The findings suggest that the behavior of service users will be shaped by social contexts. Studies indeed agree by demonstrating how divergent the user behavior is in different institutional settings of healthcare system. The degree to which medical service users utilize both biomedicine and alternative medicine varies significantly between different healthcare systems [ 76 ]. The extent to which the users of alternative medicine adhere to its remedies also varies in different healthcare systems [ 75 ]. While awaiting future research that specifies how these behavioral variations lead to healthcare outcomes, practitioners of biomedicine as well as practitioners of alternative medicine need to be aware of the ways in which their healthcare systems affect the behavior of medical service users. This paper has some limitations that future research needs to address. With regard to the cross-national panel data analysis, it uses proxy measures of medical resources for alternative medicine. This is inevitable given the scarcity of data in this field. Efforts need to be made in collecting more direct measures, such as the practitioner numbers of alternative medicine, equivalent to those of biomedicine. Regarding the content analysis of medical journal articles on alternative medicine, future research needs to expand the currently limited coverage of publication years (i.e. 1964 to 2009), in order to reflect developments in more recent medical trials and real-world practices. In a similar vein, efforts need to be made to collect live experiences of medical service users in plural medical systems that this paper is not able to apprehend in the cross-national panel or medical journal articles. Supporting information S1 File Dataset1_panel analysis. (DTA) Click here for additional data file. S2 File Supporting text and tables. (DOCX) Click here for additional data file. S3 File Dataset2_content analysis. (DTA) Click here for additional data file.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6511024/

Living with plague: Lessons from the Soviet Union’s antiplague system

Zoonoses, such as plague, are primarily animal diseases that spill over into human populations. While the goal of eradicating such diseases is enticing, historical experience validates abandoning eradication in favor of ecologically based control strategies (which reduce morbidity and mortality to a locally accepted risk level). During the 20th century, one of the most extensive plague-eradication efforts in recorded history was undertaken to enable large-scale changes in land use in the former Soviet Union (including vast areas of central Asia). Despite expending tremendous resources in its attempt to eradicate plague, the Soviet antiplague response gradually abandoned the goal of eradication in favor of plague control linked with developing basic knowledge of plague ecology. Drawing from this experience, we combine new gray-literature sources, historical and recent research, and fieldwork to outline best practices for the control of spillover from zoonoses while minimally disrupting wildlife ecosystems, and we briefly compare the Soviet case with that of endemic plague in the western United States. We argue for the allocation of sufficient resources to maintain ongoing local surveillance, education, and targeted control measures; to incorporate novel technologies selectively; and to use ecological research to inform developing landscape-based models for transmission interruption. We conclude that living with emergent and reemergent zoonotic diseases—switching to control—opens wider possibilities for interrupting spillover while preserving natural ecosystems, encouraging adaptation to local conditions, and using technological tools judiciously and in a cost-effective way. From Eradication to Control: Different Models Global eradication schemes have had two notable successes (smallpox and rinderpest, both viruses with well-defined transmission pathways and effective vaccines), but eradication has proven elusive for other diseases with more complex ecologies ( 45 ). Global public health experts have classified antidisease schemes into categories, from "disease extinction" and "eradication" to "elimination of infection," "elimination of disease," and "control" ( 46 ). Control differs from the other categories because it reduces morbidity and mortality to a "locally accepted level," rather than aiming to reduce disease incidence and/or human infections to zero (whether locally or globally) ( 47 ). Control has been less attractive because human and animal cases may still occur, continued intervention efforts seem endless and costly, and it does not seize politicians' and funding agencies' attention. However, a brief exposition of an eradication campaign that enjoyed optimal resources and conditions, yet failed, argues forcefully in favor of control over extinction, eradication, or elimination. Between 1917 and 1991, one of the most extensive endemic plague-eradication efforts in recorded history was undertaken in the Union of Soviet Socialist Republics (USSR) to enable large-scale changes in land use (such as mining and agricultural development). Eradication meant zero human plague cases and eliminating endemic plague by destroying the ecological systems that enabled Y. pestis to survive and circulate in vast regions. These centrally planned eradication campaigns, coordinated by the APIs, began in the 1920s with ecological and serological surveys to identify infected and susceptible rodent host populations and predators ( 48 – 50 ), key species of the flea vectors ( 51 ), and the roles of burrow microhabitats and soil ( 52 , 53 ) in maintaining endemic plague. Public health workers and scientists attempted to "sanitize" or "liquidate"—the terms used at the time—all wild rodent and flea populations by engaging tens of thousands of local people to place poisons (such as chloropicrin, zinc phosphide, or carbon disulfide) manually into each burrow entrance and by using airplanes to spray vast regions with dichlorodiphenyltrichloroethane (DDT) during the 1950s and 1960s ( 34 , 48 , 54 ). Animals resistant to poison baits (such as marmots) were gassed, snared, or shot ( 55 ). The economic expense of these efforts is unknown but must have been astronomical; the chemical methods alone were "costly," especially when carried out over such vast areas ( 56 ). The goal was "complete purges of animal-infested territories" through poisoning, burning all vegetation, and finally plowing up wild rodent colonies ( 57 ), to "keep districts free from plague…by decreasing the number of sources of future epizootics" ( 58 ). This goal dovetailed well with Soviet plans to make the steppes productive, including Nikita Khrushchev's "Virgin Lands" initiative (1953–1964) in which workers plowed up over 30 million hectares of grassland and dug massive irrigation works ( 59 ). Remaking this landscape on a vast scale required removing primary plague foci. In comparison, no such centralized eradication scheme was carried out in the United States, where locally determined land use predominated. Plague arrived in North America circa 1900 in Los Angeles, San Francisco, and other port cities. Local and state public health officials, collaborating with the US Public Health Service and university-based scientists, recognized the importance of preventing Y. pestis from spreading beyond ports of entry. The local eradication-and-containment strategy succeeded in places like New Orleans, Louisiana and Galveston, Texas ( 60 ). [The same was true in Australia, where a combination of strict quarantine, mandatory fumigation, rat killing, and concrete walls managed to contain plague ( 61 ).] Despite equivalent efforts in California, containment failed due to the presence of almost-ideal ecological factors: high urban and ex-urban rodent contact rates, variably susceptible wild rodent and flea species, and a congenial climate. By 1935, containment had shifted to county-by-county control (facilitated by sporadic state and federal investment, never on the Soviet scale) ( 4 , 60 ). By the 1960s plague had become established in secondary endemic foci throughout the western United States, newly entrenched in wild burrowing rodents in several states. In contrast to its long-standing presence in central Asia, plague's geographic expansion into the interior United States demonstrated an epidemiological pattern similar to an exotic invasive plant or animal: early success in port cities, then a quiescent period of geographic spread, and then, in ecologically appropriate areas, consistent endemic/enzootic cycles with sporadic human cases in rural areas ( 44 ). North American secondary plague foci now function similarly to primary foci in central Asia, where the history of eradication and control campaigns provides some object lessons on how to minimize human cases in populations that must live or work in endemic plague areas. In the Soviet hinterlands, plague-eradication campaigns looked successful at first, but within periods of 5 to 20 y ecological surveys of plague foci again found burrows containing rodents, fleas, and Y. pestis , demonstrating the resilience of the endemic plague ecosystem ( 62 ). For example, scientists cited plague-eradication campaigns begun in the 1930s in southwest Russia as models of success. However, by the 1950s and early 1960s, rodents and plague had returned on the eastern side of the Volga River and in the Kalmyk autonomous area ( 63 – 65 ). Plague's resilience meant that even full-out interventions failed to reduce infected fleas and wild animals to zero and failed to prevent future recurrences. Fenyuk ( 66 ) cautioned that "single campaigns" were not sufficient to "liquidate the foci"; the "disinfection" of endemic plague foci had to be repeated for several years. In the Kyzyl Kum area, Stogov ( 67 ) estimated that endemic plague could not be eradicated unless >90% of all gerbils were killed. Iakolev ( 68 ) noted that gerbil population density in this area nonetheless recovered from eradication campaigns in about 3 to 4 y under favorable climatic and food conditions. Moreover, scientists increasingly worried about the use of carcinogenic, toxic, and long-lasting chemicals such as DDT and the exposure of people and nontarget animals ( 57 , 69 ). Human plague cases did significantly decrease by the 1950s, but not to zero. In 1959, B.N. Pastukhov notified the WHO that "since about 1928, when plague control measures began to be developed on a large scale, there have been no cases of plague in human beings in the USSR" ( 70 ). Pastukhov's assertion hid the truth. As seen in Fig. 2 , plague nonetheless persisted in Kazakhstan and other areas of the USSR. Soviet plague researchers could not publish information about epizootics until 1956 and were forbidden to discuss human cases at any time ( 71 ). Once reported to Moscow, plague cases and outbreaks (in animals and humans) did not appear in media or WHO reports because, as one former high-ranking official remembered, "the totalitarian state was concerned about its respectable image" ( 72 ). Therefore, Union-level data about plague (and other "sensitive" diseases) in both animal and human populations were systematically underreported, underscoring the manner in which science was subverted to political ends. By the 1970s, although eradication of plague was still the ostensible goal for the Soviet APIs, practices had shifted toward management and control of plague ( 73 ). Fig. 3 shows a generalized timeline of changes in strategies, especially in Kazakhstan. The shift from eradication to control included increased emphasis on managing flea densities. This made epidemiological sense: By reducing vector densities in the ecosystem, tangible reductions in the transmission of plague, frequency of epizootics, and the risk of human spillover were achieved, even without complete rodent eradication. The availability of chlorinated hydrocarbons (such as DDT) after the Second World War made vector control achievable. As DDT became less effective and its dangers more apparent, however, scientists shifted to using other compounds by the 1980s. Draconian interventions, such as burning vegetation and plowing up rodent colonies, were largely abandoned. By 1990, eradication practices such as widespread rodent extermination had also been abandoned in favor of a preventive regime built around predictive modeling, surveillance, education of local people, and vector control ( 34 ). Fig. 3. Timeline of shift from plague eradication (blue) to control (red), USSR. Central Asian antiplague activities are highlighted. Demonstrates timing of shifts from rodent eradication to vector control, and from DDT to nonchlorinated hydrocarbon residuals, and increasing use of surveillance data in predictive models. The history of the antiplague campaigns in the former USSR also warns us to be critical of eradication claims and the official data that were used to support them. Union-level data during Soviet days were often subject to political pressures and thus biased to some degree. However, this does not generally apply to regionally curated data: The KSCQZD and Kazakh antiplague stations hold decades of continuous plague surveillance data collected by trained and dedicated scientists and specialists ( 74 , 75 ). Using the Soviet-era data most productively today requires that they also be subject to scholarly historical analysis ( 76 ), including documentary investigation into the external social, cultural, and political factors influencing scientific work in a given place and time; semistructured or open interviews of participants (or their successors); full integration of both ecological/biological and human social data for a given dataset; notation of problems with datasets; and close collaboration with scientists using these datasets. Historical analysis is key to analyzing change over time in disease patterns and responses. Accepting Plague Persistence The shift from eradication to control was possible because of enhanced ecological understanding of the complete plague system and acceptance of plague as a disease of endemic wildlife. Spillover to humans is more likely with epizootics, which occur only at particular rodent and flea species' density thresholds ( 77 , 78 ). Humans are exposed through flea bites and direct contact (eating, skinning, or butchering infected animals). Urban areas near endemic foci may experience transmission to commensal rodents and the spillover to humans may escalate through direct (pneumonic) and possibly indirect (human ectoparasites) transmission ( 16 , 17 ). The historical Soviet data strongly suggest that for successful control we should focus on four attributes of plague ecology: flea vector species' distributions and densities, understanding the role of the burrow in plague persistence, the interplay between multiple rodent and flea species and their resistance to Y. pestis , and the role of landscape-level features in plague epizootics. This is validated by recent work using niche modeling, which found that plague cases occurred only where host animal ranges overlapped with "plague niches" whose locations depended on nonhost factors such as vector (flea species) distribution ( 79 ). Flea Distribution and Density. Flea species differ in their Y. pestis transmission capability, efficiency, and temporality. In southeastern Kazakhstan, X skrjabini, X. hirtipes , and X. gerbilli minax become quickly blocked by Y. pestis biofilm aggregates in their guts, leading to intensive feeding behaviors on multiple hosts that make them rapid transmitters of plague (mediated by individual host-vector interactions) ( 80 ). By contrast, Coptopsylla spp. (becoming active in the autumn as Xenopsylla goes dormant) are slower transmitters ( 8 ). Fleas may exhibit different transmission potentials even at the subspecies level: In the United States, Oropsylla tuberculate cynomuris has been found to transmit Y. pestis three times more efficiently than does the closely related Oropsylla hirsuta ( 81 ). The relative abundance of these different flea subspecies may contribute to epizootic potential in specific areas. Plague persistence in foci and spillover transmission can be interrupted by reducing flea abundance, especially if locally present flea species maintain Y. pestis without vectoring (some infected flea species live much longer than acutely infected hosts) ( 15 , 82 ). Soviet scientists also cataloged flea species, such as X. gerbilli spp., that feed on multiple hosts ( R. opimus , Meriones spp., and rats); the distribution and density of such flea species may determine Y. pestis circulation among hosts ( 83 , 84 ). Certain flea species may also facilitate cross-species mammalian transmission: One study found Y. pestis -infected X. skrjabini on great gerbils, foxes, weasels, and steppe polecats in a single test area, for example ( 85 ). In the United States, Pulex simulans functions as a cosmopolitan feeder (and probably spreader) ( 86 ). Reducing flea densities became a cornerstone of Soviet policy in 1959, when Pastukhov outlined the new official antiplague program that combined flea and rodent eradication to "give prospects of a more rapid sanitation of natural foci and the final eradication of the plague epizootics" ( 70 ). That goal proved elusive, however. Between the 1940s and 1970s, scientists reported increasing insect resistance, mammalian toxic effects, and environmental persistence of chlorinated hydrocarbon insecticides ( 87 , 88 ). These problems stimulated a search for alternative insecticides. In Kazakhstan, for example, antiplague scientists shifted first to pyrethrins, then more recently to fipronil (a phenylpyrazole that disrupts insects' central nervous systems without affecting mammals) ( 89 ). Recent attempts to deploy fipronil in treated grain baits have shown effective reductions of flea vectors on individual mammalian hosts ( 90 ). With recent studies highlighting the role of climate on flea abundance ( 86 , 91 ), ongoing surveillance of species-specific vector densities and distributions is a crucial component of endemic plague control. Burrow Ecology. A focus on flea control rather than rodent eradication also led to an interest in how the burrow environment influenced plague persistence. Fleas can become relatively dormant if conditions are poor ( 92 ); others, most notably Xenopsylla , overwinter in the burrows. Some flea species ( Citellophilus tesquorum , for example) can harbor Y. pestis for up to 18 mo ( 93 , 94 ), making them potential key sources of plague persistence. Soviet scientists have investigated the roles of burrow microhabitats and soil in maintaining plague at least since the 1960s ( 52 , 95 ), finding, for example, that Y. pestis persisted in the detritus of deeper burrows even after host abandonment ( 93 ). Great gerbil burrows vary in depth up to 3 m, and the depth is critical to flea survival, with shallow burrows less favorable for fleas ( 96 , 97 ). Burrows in sandy soil and those surrounded by vegetation may be more likely to contain plague-bearing fleas ( 9 ). Using PCR assays ( 98 ), Y. pestis can be more sensitively detected in burrow-collected fleas, and flea species more likely to contribute to plague persistence locally can be identified. Work on flea behavior in the context of rodent ecology, including selection of microhabitats within burrows, would be a potentially fruitful area for future research. Host Diversity and Behavior. How does the local composition and social behavior of rodent species influence plague control? In the former USSR, 15 rodent species are considered main plague host reservoirs, but many other rodent populations coexist with these main hosts ( 10 ). While it is unclear what traits make a rodent population a suitable main plague host—the rodent species itself is not a sufficient determinant ( 43 )—one hypothesis is that main plague host populations are characterized by a heterogeneous susceptibility to the disease and a long duration of infection in some ill individuals ( 65 ). Another is social behavior (most main plague hosts are social or facultatively social species). Social species facilitate transmission more effectively than solitary ones through direct contacts and colony-mediated spatial distributions ( 96 ). Together these traits facilitate both the survival of the plague host species and of the disease and assessing them are important host surveillance goals. Cost-effective detection tools used to identify human cases, such as lateral flow strip-testing, may be adaptable for identifying exposed surviving primary host animals ( 99 – 101 ). Secondary plague host species, some of which are highly susceptible, may have an amplifying effect on the spread of the disease during epizootics, or others may enable plague persistence ( 102 ). Furthermore, secondary species' habitat ranges can serve as a bridge between plague reservoir hosts and peridomestic rodents, thus putting humans at risk. Secondary plague hosts are characterized by a low resistance to plague (with little heterogeneity within their populations) ( 8 , 103 ). Control efforts must include assessment of the temporal population dynamics and habitat range of both primary and secondary hosts to accurately estimate epizootic risk in a particular area. Landscape/Population Ecology. Landscape-level features are crucial in promoting or inhibiting epizootics. As noted above, many of the rodents harboring plague live in burrow colonies, often with complex spatial and population dynamics that can only be appreciated at a broad geographic scale ( 7 , 13 , 104 ). Burrows can be drivers of disease in several respects ( 14 , 75 ). They often contain multiple individuals, which allows longer-term interactions and potentially greater contact rates. Species using complex burrow systems also may exhibit greater philopatry, which in turn alters the spread of infection from one burrow system to another. In R. opimus , dispersal between colonies is common; one study in Uzbekistan found that 42.8% of female great gerbils and 100% of males switched colonies at least once during a year ( 105 ), potentially spreading Y. pestis . An enhanced potential for plasticity of social organization in complex burrow systems also yields different potential for dispersal and hence transmission of disease. Soviet researchers recognized the contributions of spatial structure to plague, noting that female kinship, shared burrows, and male-biased dispersal would contribute to high contact rates and plague persistence at low host abundance ( 57 , 106 – 108 ). More recently, Wilschut et al. ( 96 , 97 ) conducted a survey in Kazakhstan of great gerbil colonies (whose burrow systems include multiple branching tunnels that can occupy an area of 20 to 60 m 2 ). They then created landscape objects from satellite images which were linked to plague data from 1949 to 1995. Burrow distribution was nonrandom, associated with greener areas, and the direction of plague outbreaks were aligned with burrows and their connections. This method of tracking great gerbil dispersion with freely available data promises to guide efficient on-ground surveillance. These ecological approaches to plague persistence, combined with modeling and statistical inference of outbreak potential, can be used to help predict shifts in targeted enzootic foci by increasing surveyable surface area without increasing resource costs. Indeed, efforts to create "risk maps" from annual surveillance data have been tested in the Pre-Balkash plague focus, although problems with the model used included high rates of false-positive predictions ( 109 , 110 ). Incorporating the ecological data we highlighted earlier, such as flea density, will maximize the usefulness of this model to field surveillance teams by reducing false positive risk predictions. Adding a thorough understanding of the ecology of plague and the sensitivity of its actors to climate fluctuations are the keys to integrating modeling tools with the ground-level work currently done by plague control centers. Accepting plague persistence also means being attentive to conservation to maintain healthy ecosystems. Plague-reservoir species, such as R. opimus , may promote ecosystem resilience and should not be eradicated. Resilient ecosystems include communities of organisms whose interactions sustain one another in the system and diminish disruptions to it; endemic diseases can shape populations and communities ( 111 ). In our case study, healthy burrowing rodent populations function as "ecosystem engineers," enhancing steppe soils, influencing the nitrogen content, and creating a "fertile island" effect ( 112 , 113 ). Annihilating rodent populations not only destroys this effect but also temporarily encourages starving fleas to seek new hosts (including humans and domesticated animals), facilitating the transmission of the disease. Thus, pathogens and attempts to regulate them can threaten biodiversity ( 114 ), loss of biodiversity may increase risk of disease transmission ( 115 ), and climate change may amplify that risk ( 116 ). Flea Distribution and Density. Flea species differ in their Y. pestis transmission capability, efficiency, and temporality. In southeastern Kazakhstan, X skrjabini, X. hirtipes , and X. gerbilli minax become quickly blocked by Y. pestis biofilm aggregates in their guts, leading to intensive feeding behaviors on multiple hosts that make them rapid transmitters of plague (mediated by individual host-vector interactions) ( 80 ). By contrast, Coptopsylla spp. (becoming active in the autumn as Xenopsylla goes dormant) are slower transmitters ( 8 ). Fleas may exhibit different transmission potentials even at the subspecies level: In the United States, Oropsylla tuberculate cynomuris has been found to transmit Y. pestis three times more efficiently than does the closely related Oropsylla hirsuta ( 81 ). The relative abundance of these different flea subspecies may contribute to epizootic potential in specific areas. Plague persistence in foci and spillover transmission can be interrupted by reducing flea abundance, especially if locally present flea species maintain Y. pestis without vectoring (some infected flea species live much longer than acutely infected hosts) ( 15 , 82 ). Soviet scientists also cataloged flea species, such as X. gerbilli spp., that feed on multiple hosts ( R. opimus , Meriones spp., and rats); the distribution and density of such flea species may determine Y. pestis circulation among hosts ( 83 , 84 ). Certain flea species may also facilitate cross-species mammalian transmission: One study found Y. pestis -infected X. skrjabini on great gerbils, foxes, weasels, and steppe polecats in a single test area, for example ( 85 ). In the United States, Pulex simulans functions as a cosmopolitan feeder (and probably spreader) ( 86 ). Reducing flea densities became a cornerstone of Soviet policy in 1959, when Pastukhov outlined the new official antiplague program that combined flea and rodent eradication to "give prospects of a more rapid sanitation of natural foci and the final eradication of the plague epizootics" ( 70 ). That goal proved elusive, however. Between the 1940s and 1970s, scientists reported increasing insect resistance, mammalian toxic effects, and environmental persistence of chlorinated hydrocarbon insecticides ( 87 , 88 ). These problems stimulated a search for alternative insecticides. In Kazakhstan, for example, antiplague scientists shifted first to pyrethrins, then more recently to fipronil (a phenylpyrazole that disrupts insects' central nervous systems without affecting mammals) ( 89 ). Recent attempts to deploy fipronil in treated grain baits have shown effective reductions of flea vectors on individual mammalian hosts ( 90 ). With recent studies highlighting the role of climate on flea abundance ( 86 , 91 ), ongoing surveillance of species-specific vector densities and distributions is a crucial component of endemic plague control. Burrow Ecology. A focus on flea control rather than rodent eradication also led to an interest in how the burrow environment influenced plague persistence. Fleas can become relatively dormant if conditions are poor ( 92 ); others, most notably Xenopsylla , overwinter in the burrows. Some flea species ( Citellophilus tesquorum , for example) can harbor Y. pestis for up to 18 mo ( 93 , 94 ), making them potential key sources of plague persistence. Soviet scientists have investigated the roles of burrow microhabitats and soil in maintaining plague at least since the 1960s ( 52 , 95 ), finding, for example, that Y. pestis persisted in the detritus of deeper burrows even after host abandonment ( 93 ). Great gerbil burrows vary in depth up to 3 m, and the depth is critical to flea survival, with shallow burrows less favorable for fleas ( 96 , 97 ). Burrows in sandy soil and those surrounded by vegetation may be more likely to contain plague-bearing fleas ( 9 ). Using PCR assays ( 98 ), Y. pestis can be more sensitively detected in burrow-collected fleas, and flea species more likely to contribute to plague persistence locally can be identified. Work on flea behavior in the context of rodent ecology, including selection of microhabitats within burrows, would be a potentially fruitful area for future research. Host Diversity and Behavior. How does the local composition and social behavior of rodent species influence plague control? In the former USSR, 15 rodent species are considered main plague host reservoirs, but many other rodent populations coexist with these main hosts ( 10 ). While it is unclear what traits make a rodent population a suitable main plague host—the rodent species itself is not a sufficient determinant ( 43 )—one hypothesis is that main plague host populations are characterized by a heterogeneous susceptibility to the disease and a long duration of infection in some ill individuals ( 65 ). Another is social behavior (most main plague hosts are social or facultatively social species). Social species facilitate transmission more effectively than solitary ones through direct contacts and colony-mediated spatial distributions ( 96 ). Together these traits facilitate both the survival of the plague host species and of the disease and assessing them are important host surveillance goals. Cost-effective detection tools used to identify human cases, such as lateral flow strip-testing, may be adaptable for identifying exposed surviving primary host animals ( 99 – 101 ). Secondary plague host species, some of which are highly susceptible, may have an amplifying effect on the spread of the disease during epizootics, or others may enable plague persistence ( 102 ). Furthermore, secondary species' habitat ranges can serve as a bridge between plague reservoir hosts and peridomestic rodents, thus putting humans at risk. Secondary plague hosts are characterized by a low resistance to plague (with little heterogeneity within their populations) ( 8 , 103 ). Control efforts must include assessment of the temporal population dynamics and habitat range of both primary and secondary hosts to accurately estimate epizootic risk in a particular area. Landscape/Population Ecology. Landscape-level features are crucial in promoting or inhibiting epizootics. As noted above, many of the rodents harboring plague live in burrow colonies, often with complex spatial and population dynamics that can only be appreciated at a broad geographic scale ( 7 , 13 , 104 ). Burrows can be drivers of disease in several respects ( 14 , 75 ). They often contain multiple individuals, which allows longer-term interactions and potentially greater contact rates. Species using complex burrow systems also may exhibit greater philopatry, which in turn alters the spread of infection from one burrow system to another. In R. opimus , dispersal between colonies is common; one study in Uzbekistan found that 42.8% of female great gerbils and 100% of males switched colonies at least once during a year ( 105 ), potentially spreading Y. pestis . An enhanced potential for plasticity of social organization in complex burrow systems also yields different potential for dispersal and hence transmission of disease. Soviet researchers recognized the contributions of spatial structure to plague, noting that female kinship, shared burrows, and male-biased dispersal would contribute to high contact rates and plague persistence at low host abundance ( 57 , 106 – 108 ). More recently, Wilschut et al. ( 96 , 97 ) conducted a survey in Kazakhstan of great gerbil colonies (whose burrow systems include multiple branching tunnels that can occupy an area of 20 to 60 m 2 ). They then created landscape objects from satellite images which were linked to plague data from 1949 to 1995. Burrow distribution was nonrandom, associated with greener areas, and the direction of plague outbreaks were aligned with burrows and their connections. This method of tracking great gerbil dispersion with freely available data promises to guide efficient on-ground surveillance. These ecological approaches to plague persistence, combined with modeling and statistical inference of outbreak potential, can be used to help predict shifts in targeted enzootic foci by increasing surveyable surface area without increasing resource costs. Indeed, efforts to create "risk maps" from annual surveillance data have been tested in the Pre-Balkash plague focus, although problems with the model used included high rates of false-positive predictions ( 109 , 110 ). Incorporating the ecological data we highlighted earlier, such as flea density, will maximize the usefulness of this model to field surveillance teams by reducing false positive risk predictions. Adding a thorough understanding of the ecology of plague and the sensitivity of its actors to climate fluctuations are the keys to integrating modeling tools with the ground-level work currently done by plague control centers. Accepting plague persistence also means being attentive to conservation to maintain healthy ecosystems. Plague-reservoir species, such as R. opimus , may promote ecosystem resilience and should not be eradicated. Resilient ecosystems include communities of organisms whose interactions sustain one another in the system and diminish disruptions to it; endemic diseases can shape populations and communities ( 111 ). In our case study, healthy burrowing rodent populations function as "ecosystem engineers," enhancing steppe soils, influencing the nitrogen content, and creating a "fertile island" effect ( 112 , 113 ). Annihilating rodent populations not only destroys this effect but also temporarily encourages starving fleas to seek new hosts (including humans and domesticated animals), facilitating the transmission of the disease. Thus, pathogens and attempts to regulate them can threaten biodiversity ( 114 ), loss of biodiversity may increase risk of disease transmission ( 115 ), and climate change may amplify that risk ( 116 ). Conclusion As the old saying goes, "those who do not understand history are doomed to repeat it." Ecologically complex endemic zoonoses such as plague resist eradication efforts, and several social and biological lessons follow from the experience of Kazakh and Soviet scientists in the ancient plague foci of central Asia. Over time, Soviet antiplague policies abandoned eradication in favor of control, with emphasis on developing basic knowledge of plague ecology in local areas. Switching to control opens wider possibilities for interrupting spillover while preserving natural ecosystems, encourages us to adapt to local conditions, and uses technological tools judiciously and in a cost-effective way. Finally, using history as a guide reminds us that political and social influences will always affect scientific work and public policy and must be factored into successful disease control programs. Once our focus shifts from eradication to control, we are better equipped to respond to existing, emerging, and reemerging diseases in wildlife systems that naturally spill over into humans. Interrupting transmission is essential when disease cannot be eliminated from the environment, either because we lack the necessary knowledge and tools to control a newly emergent disease or because some diseases (like plague) are highly resilient to eradication. Surveillance, vector control, and preventive-measures education are the cornerstones of a generic endemic disease control system. Local infrastructure in place to rapidly respond to new spillover events is essential, rather than waiting for outbreaks to occur and then developing a response. On a practical level, it is important to support less-glamorous but necessary needs such as vehicles and pumps that are often neglected in favor of alluring new technologies. Resource-intensive schemes such as ambitious sequencing efforts (for example) can come at the expense—literally—of practical results ( 23 ). We do not advocate eschewing novel technologies but instead argue for their selective incorporation in the context of basic ecological knowledge about endemic disease systems. The history of antiplague work in central Asia suggests several avenues for further research on plague in endemic foci (including established secondary foci such as those in the western United States). First, since fleas are the key to plague transmission, better understanding of variation in flea species' behavior in different types of burrows, on different host rodent species, and over seasonal and shorter-scale variation is essential ( 82 , 83 , 92 , 117 ). Second, more landscape-level models and other efforts to understand patterns of rodent and flea dispersal and its influence on disease outbreaks should prove productive and predictive for transmission interruption ( 75 , 104 ). Third, more attention should be paid to how resistance to Y. pestis varies among and within rodent species ( 8 , 118 ). This is of particular interest in the context of recent efforts to develop oral vaccines for rodents that carry plague ( 119 ), since species may differ in the degree to which such vaccines are effective. Practical decisions and policy should be based on solid scientific knowledge, historical and current, in social context. Today, most plague cases occur in Africa, where secondary endemic foci are in the process of becoming established due to close proximity between wildlife, domestic rodents, and people. Along with the factors advocated above, attention to urban and rural resources for protecting people from vector bites (such as improved housing) and containment of Y. pestis spread is essential. Changing demographics and development in central Asia due to the Chinese-sponsored Belt and Road Initiative and infrastructure improvements in Africa means that more travelers will be exposed to endemic disease foci and can quickly move infections and vectors into new territories. It is therefore essential that monitoring and control networks collaborate to quickly detect and act upon the opportunistic spread of endemic diseases. While we must accept the persistence of plague and other zoonotic wildlife diseases, living with them requires a coordinated response of ecologists, public health officials, and people who are cognizant of the lessons of history. In our opinion, such an approach will be much more likely to reduce the overall human burden from wildlife spillover of diseases than single-disease-focused eradication attempts.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3367348/

Australian Public and Smallpox

A national survey of 1,001 Australians found that most were concerned about a bioterrorist attack and were ill-informed about smallpox prevention and response. Since general practitioners were commonly identified as the initial point of care, they should become a focus of bioterrorism response planning in Australia. The Study A list of private telephone numbers was randomly selected for each of Australia's 8 states and territories that was proportional to their contribution to the adult population. Participants were recruited to provide a sample size of 1,000, which allowed a precision of 2%–3% when calculating a 95% confidence interval of a dichotomous variable with a base proportion ranging from 10% to 50%. Eight experienced telephone interviewers conducted the survey during July 2004. Repeat calls were conducted when persons indicated interest in participating but were unable to do so during the initial contact. The questionnaire was administered upon agreement to participate, after introducing the survey's purpose, providing a guarantee of confidentiality, and giving reassurance of freedom to withdraw consent. The questionnaire was pretested for length and comprehensibility in a pilot study during May 2004. The final instrument took 10–15 minutes to administer and contained 22 questions. Data were analyzed by using SPSS for Windows version 11 (SPSS Inc., Chicago, IL, USA). Ethical approval was granted by the human ethics subcommittee at James Cook University (Nr. H1745). A total of 1,001 Australian adults completed the survey. Two hundred thirteen were excluded (38 children, 91 adults with limited English ability, 9 incoherent adults, and 75 adults contacted at their workplace), and 582 refused to participate (response rate 63.2%). Respondents were geographically representative of the Australian population. Respondents' ages were normally distributed (mean 52.2 years, standard deviation 17 years) and 62.8% (629) were female. Most (58.6%, 587) lived in cities, which reflected the situation in Australia, where 66.3% of the population live in urban areas. The level of education of respondents reflected that of the Australian population. Concern about the risk of a bioterrorist attack in Australia was perceived as high by 182 (18.2%), medium by 392 (39.2%), low by 339 (33.9%), and nonexistent by 14 (1.4%); 72 (7.2%) did not know and 2 (0.2%) did not answer. Logistic regression modeling showed that age was the only demographic feature significantly associated with perceiving high risk of a bioterrorist attack (compared with low, medium, or none), with an odds ratio of 1.016 per year (p<0.001). Most respondents (60.6%, 606) believed that human smallpox cases had occurred in the past 5 years and that effective medical treatment existed for smallpox ( Table 1 ). The likelihood of contracting smallpox by working in close contact with someone with the disease (e.g., in the same office) was considered low by 157 (15.7%), medium by 163 (16.3%), and high by 419 (41.9%); 261 (26.1%) did not know and 1 (0.1%) did not answer. Table 1 Responses to questions on general knowledge of smallpox, Australia, 2004 In the past 5 years, do you think… Yes, n (%) No, n (%) Do not know, n (%) No answer, n (%) Human cases of smallpox have occurred in Australia? 127 (12.7) 538 (53.7) 336 (33.6) 0 Human cases of smallpox have occurred somewhere in the world? 606 (60.6) 233 (23.3) 161 (16.1) 1 (0.1) An effective medical treatment exists for smallpox? 524 (52.3) 114 (11.4) 363 (36.3) 0 A total of 583 (58.2%) respondents stated that they had been vaccinated against smallpox; 346 (34.6%) indicated no prior vaccination against smallpox, 71 (7.1%) did not know, and 1 (0.1%) did not answer. Among 61 respondents born since 1979, the year that smallpox was eradicated and worldwide childhood vaccination terminated, 32 (52.5%) indicated that they had not been vaccinated against smallpox, 20 (32.8%) reported that they had been vaccinated, and 9 (14.8%) did not know. Of 841 respondents born before 1980, 502 (59.8%) reported that they had been vaccinated against smallpox. The acceptance of vaccination against smallpox under specific hypothetical scenarios was explored. Vaccination would be accepted as an immediate precautionary measure by 41.7% of respondents, while 42.3%, 48.9%, and 56.3% would accept vaccination if cases were reported somewhere in the world, Australia, or their own community, respectively. Among respondents who did not report previous vaccination, 44.5% would accept vaccination as a precautionary measure ( Table 2 ). Table 2 Responses of 418 persons who did not report being vaccinated against smallpox who would accept vaccination under various hypothetical conditions, Australia, 2004 Would accept vaccination… Yes, n (%) No, n (%) Do not know, n (%) No answer, n (%) As a precautionary measure? 186 (44.5) 202 (48.3) 15 (3.6) 15 (3.6) If cases were reported in the world? 193 (46.2) 199 (47.6) 11 (2.6) 15 (3.6) If cases were reported in Australia? 259 (62.0) 132 (31.6) 12 (2.9) 15 (3.6) If cases were reported in your community? 332 (79.4) 62 (14.8) 9 (2.2) 15 (3.6) Modeling the readiness to accept vaccination showed that older persons were less likely to accept smallpox vaccination (odds ratio 0.977 per year, p<0.001). Respondents with more education were also less likely to accept vaccination under any scenario (odds ratio 0.845 per education category, p<0.01). When asked in an open-ended question where they would first seek diagnosis or care if they thought they had contracted smallpox, 591 (59.0%) respondents mentioned their general practitioner (family physician). Hospital emergency departments were indicated by 330 (33.0%), a public health department by 43 (4.3%), and other sources by 18 (1.8%); 16 (1.6%) did not know and 3 (0.3%) did not answer. Overall, 418 (41.8%) indicated a high level of confidence in their physicians' ability to recognize symptoms of smallpox, 291 (29.1%) a medium level of confidence, 177 (17.7%) a low level of confidence, and 42 (4.2%) no confidence; 68 (6.8%) did not know, and 5 (0.5%) did not answer. Conclusions Most Australian adults interviewed in this national survey reported medium-to-high concern about the risk of a bioterrorism attack in Australia (57.4%) and believed that human smallpox cases had occurred in the past 5 years (60.6%). This finding may explain the general willingness to accept vaccination as a precautionary measure in the absence of a bioterrorism event ( 7 ). This finding is similar to that of a US survey, which indicated a strong community desire for precautionary vaccination against smallpox ( 5 ). However, the general public is unlikely to be sufficiently informed to balance the risks of a bioterrorism event against the potential for harm from vaccination ( 8 ). Given that the currently available smallpox vaccine must produce a significant lesion to be considered effective and commonly results in other adverse events, some severe, mass vaccination as an antiterrorism strategy must be epidemiologically justified by a substantial risk ( 9 – 11 ). Accurate information on smallpox vaccine adverse effects must be made available to the Australian public, although this information may affect acceptance of vaccination, as was documented among potential medical first responders in the United States ( 12 ). Participants were unclear about their personal smallpox vaccination status. Although respondents born after smallpox vaccination was stopped in Australia were incorrect if they believed that they had been vaccinated against smallpox, 33% of this group falsely indicated that had been vaccinated. This belief may lead to a false sense of security in the event of an actual bioterrorist attack with smallpox virus. Despite the desire for precautionary vaccination, only 259 (62%) respondents who believed they were unvaccinated would accept smallpox vaccination if cases were reported in Australia. A false belief that effective medical treatment exists for smallpox, which was held by more than half of the respondents, may influence decisions to accept vaccination in response to locally occurring cases ( 13 ). Public health authorities have a clear mandate to improve the community's knowledge of smallpox and bioterrorism. These efforts must involve groups, particularly the elderly and those with more education, who appear more unwilling to accept indicated public health measures. General practitioners emerged as a pivotal group should a bioterrorism event occur in Australia; respondents identified these medical professionals as the preferred source of initial diagnosis and management and expressed a high level of confidence in their ability to correctly diagnose smallpox. This central role for general practitioners in optimizing biopreparedness in Australia has previously been hypothesized ( 14 ). Whether the community's belief in the ability and skills of general practitioners is justified is unknown, and this aspect clearly warrants investigation ( 15 ). Specific training courses for general practitioners that heighten their clinical index of suspicion, introduce public health containment and surveillance principles, and emphasize effective communication strategies should be developed in Australia and accredited for continuing professional development. Findings in this Australian survey are similar to those in a survey in the United States, even though Australia has not experienced a bioterrorism event. In the US study, a similar proportion of respondents (63%) believed that smallpox cases had occurred in the past 5 years, but a greater proportion would accept precautionary vaccination (61%) and a slightly lower proportion (52%) would go to their own physician for diagnosis and care ( 5 ). The participation rate of 63% for this survey was similar to that in the US study (65%). This national survey found that the Australian public holds many inaccurate beliefs about smallpox and smallpox vaccination, and this misinformation could negatively affect response to a bioterrorist event. General practitioners were identified as the primary point of care and should become an important focus of bioterrorism response planning in Australia.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9009689/

Biosecurity in an age of open science

The risk of accidental or deliberate misuse of biological research is increasing as biotechnology advances. As open science becomes widespread, we must consider its impact on those risks and develop solutions that ensure security while facilitating scientific progress. Here, we examine the interaction between open science practices and biosecurity and biosafety to identify risks and opportunities for risk mitigation. Increasing the availability of computational tools, datasets, and protocols could increase risks from research with misuse potential. For instance, in the context of viral engineering, open code, data, and materials may increase the risk of release of enhanced pathogens. For this dangerous subset of research, both open science and biosecurity goals may be achieved by using access-controlled repositories or application programming interfaces. While preprints accelerate dissemination of findings, their increased use could challenge strategies for risk mitigation at the publication stage. This highlights the importance of oversight earlier in the research lifecycle. Preregistration of research, a practice promoted by the open science community, provides an opportunity for achieving biosecurity risk assessment at the conception of research. Open science and biosecurity experts have an important role to play in enabling responsible research with maximal societal benefit. Introduction Open science aims to increase the reliability and efficiency of scientific research [ 1 , 2 ]. Despite the importance of work to improve scientific practice, increased openness may increase the chance of deliberate or accidental misuse of research. These concerns are particularly salient for biological research on pathogen synthesis and engineering. Here, we examine how certain open science practices interface with risks arising from the potential misuse of biological research. For the majority of biological research, increased openness is likely to improve our ability to deal with biological threats through improving the efficiency and reliability of science. However, the subset of research in which biological risks may be increased should not be ignored, and tools and systems to encourage and facilitate responsible access to scientific findings must be available. Collaboration between the open science community and the biosecurity and biosafety communities may be mutually beneficial and will allow more consideration of the potential risks and opportunities associated with open science. Addressing risks and opportunities raised in this paper should help to reduce threats that could undermine the significant progress made in open science to date and have catastrophic consequences for society. Risks from biological research Natural pandemics have posed major threats to human populations throughout history. In recent decades, research in immunology, virology, and other biomedical domains has greatly improved global pandemic preparedness. However, some life science research has the potential to be misused. Misuse risks can be classified into biosafety risks, concerning accidental exposure and release, and biosecurity risks, concerning deliberate misuse [ 3 ]. As biotechnology grows more powerful and accessible [ 4 ], risks from accidental or deliberate misuse of research may increase. There are numerous examples of high-profile biosafety incidents [ 5 ]. The 1977 H1N1 influenza epidemic likely resulted from vaccine trials in the Soviet Union, or accidental release from a laboratory, reintroducing a strain that was circulating in the 1950s [ 6 , 7 ]. In 2015, the United States realised that it had distributed live anthrax, as opposed to the intended inactivated anthrax, in 575 shipments to 8 countries over a decade [ 8 ]. Laboratory accidents involving dangerous pathogens happen frequently and, given the leading cause of such accidents is human error, are difficult to mitigate completely even in high-level biocontainment facilities [ 9 ]. For instance, in 2003 and 2004, incidents in 3 different labs researching or containing SARS-CoV-1 lead to a total of 6 lab-acquired infections, which could have sparked further epidemics [ 10 ]. Deliberate release of pathogens may take place in the context of biological warfare or terrorism. A historic precedent for the intentional misuse of biological agents is set by the Soviet Union's extensive biological weapons program that involved the synthesis and enhancement of plague, smallpox, and other agents [ 11 ]. There have also been terrorist incidents such as the 2001 US anthrax attacks, likely conducted by a single perpetrator with training and access to biological research materials, showing that greater access to materials and methods can have tragic consequences when abused [ 12 ]. Certain life sciences research may be misused and increase the risk from deliberate biological events. For example, though advances in viral engineering may be important in areas like vaccine design and cancer therapy, they could be applied to engineer pathogens with increased virulence or transmissibility. Deliberate release of such pathogens could result in a pandemic of unprecedented severity. Research with the greatest misuse potential has been labelled dual-use research of concern (DURC), defined by the National Institutes of Health in the US as "life sciences research that, based on current understanding, can be reasonably anticipated to provide knowledge, information, products or technologies that could be directly misapplied to pose a significant threat with broad potential consequences to public health and safety" [ 13 ]. According to the Global Health Security Index, only 1% of countries have appropriate oversight for potential dual-use life science research with especially dangerous pathogens [ 14 ]. The few existing frameworks are limited in scope; for instance, the US DURC policies only apply to research on 15 select agents and toxins at federally funded institutions [ 15 , 16 ]. That certain information may cause harm and should not be publicly accessible has long been accepted in nuclear physics. In 1946, the US Atomic Energy act turned all information on nuclear weapons into "restricted data" until formally declassified [ 17 ]. Today, information hazards in the life sciences, i.e., knowledge and insights that can cause harm, are frequently associated with greater dual-use potential than physical materials [ 18 , 19 ]. Advances in molecular biology, including DNA synthesis and gene editing, are democratising science and lowering the barrier to the synthesis and engineering of biological agents [ 20 ]. As biotechnology advances, we need to acknowledge that biological research, similarly to nuclear physics, may uncover information with security implications and consider the possibility that not all information should be made publicly available. Open science and risks from biological research "Open science" has been advanced [ 21 – 23 ] in part to address widespread problems identified across the sciences [ 24 ]; however, there is little consensus on what the term, or related terms such as "reproducibility", mean [ 25 – 27 ]. We consider open science to be a set of practices that aim to improve the reliability and efficiency of scientific research [ 23 , 28 ] that are generally characterised by increased transparency. We consider that open science achieves its aims through 3 instrumental mechanisms: accessibility, verification, and reuse. By accessibility, we mean making research outputs widely, and usually publicly, available. By verification, we mean the ability to review and critique aspects of the research to establish that what is described corresponds to what was done. An example of this is computational reproducibility, i.e., ensuring that the data and code achieve the claimed results. By reuse, we mean the ability to use research outputs for the same or another purpose, such as conducting a replication study, validation study, or secondary analysis. Accessibility in many cases facilitates efforts to verify and reuse research results. Open science may contribute to mitigating biosecurity and biosafety risks. Reliable and efficient science is important for effectively preventing and responding to pandemic risks: for example, in developing drugs, vaccines, and diagnostics, and implementing effective public health responses. Preprints may have played an important role in scientific and public engagement with COVID-19 research [ 29 ]. Research excellence and ethical research conduct, both encouraged by open science, are pillars of responsible life science research for global health security [ 30 ]. The move towards open science has involved a cultural shift related to conduct and sharing and research; a similar cultural shift may be required to encourage responsible research conduct and sharing to protect the life sciences from misuse. Open science may therefore represent a useful case study [ 31 ]. As we highlight, however, there are instances where open science may exacerbate biosecurity and biosafety risks. Addressing such risks will inevitably reduce the ability to verify, reuse, or access research. However, this need not necessarily reduce reliability and efficiency. When typical open science practices seem inadvisable, we provide tentative suggestions for how reliability and efficiency can still be improved. We focus on 3 practices that appear relevant to biological risks: open code, data, and materials; preprint publication; and preregistration ( Fig 1 ). These reflect, for example, the 3 options provided in the "Conduct your own open science" section on the Centre for Open Science home page: https://www.cos.io/ . 10.1371/journal.pbio.3001600.g001 Fig 1 Summary of risks arising from open science practices, strategies to mitigate these risks, and opportunities to improve biosecurity and biosafety. Open code, data, and materials: A challenge for mitigating misuse Sharing code and data allows the research community to reproduce and check analytical findings [ 32 , 33 ] and facilitates reuse. Code is typically shared via repositories such as the Open Science Framework, GitHub, Zenodo, or in supplementary files accompanying a published article. Data may be shared similarly, though discipline-specific repositories [ 34 ] are also common. By open materials, we mean detailed, recipe-like explanations (such as written instructions or videos) on how to do certain procedures (we limit discussion to digital rather than physical materials because physical materials are already more highly regulated and different considerations apply to them. However, physical materials are included in some definitions of open materials) [ 35 ]. Such materials can be shared in supplementary files, dedicated repositories like protocols.io or bio-protocol.org , general purpose repositories like Open Science Framework, or as stand-alone journal articles (e.g., [ 36 ]). Code could be used directly or adapted to achieve nefarious goals [ 37 , 38 ]. Machine learning–guided engineering of antibiotic resistance genes exemplifies this: A model for engineering Escherichia coli β-lactamase has been described and shared openly [ 39 ]. TEM-1 β-lactamase is a well-known antibiotic resistance gene that facilitates resistance of gram-negative bacteria to a range of key antibiotics, including penicillins and cephalosporins [ 40 ]. The model allows anyone with access to a computer, in theory, to generate "better-than-natural" [ 39 ] variants of TEM-1 β-lactamase. The authors claim that their work "demonstrates a generalisable and scalable paradigm for low N-protein engineering," where low-N protein engineering refers to the minimisation of the amount of laboratory work needed to apply the approach to other protein engineering tasks. If these claims are accurate, the machine learning model in question may be applied to the design of antibiotic resistance genes to make bacteria hyperresistant, including against last-resort antibiotics. Openly shared computational methods may therefore make pathogen engineering more accessible by reducing or even removing the need for laboratory expertise and equipment. Data may also be associated with misuse risks. The publication of experiments and genetic changes required to make avian influenza transmissible in mammals has previously raised concerns around the security risk of publishing experimental data [ 41 ]. Publicly available blueprints for particularly concerning pathogens, such as the genome of the 1918 pandemic influenza virus, feature growing potential for misuse given increasing access to viral synthesis capabilities. More powerful experimental methods mean that increasingly comprehensive datasets are generated with greater potential for misuse. For instance, in high-throughput experiments a virus may be mutated thousands of times and the effect on functions such as immune evasion or binding to human cell surface receptors recorded. Such work has been conducted on pathogens with pandemic potential including SARS-CoV-2 and influenza virus [ 42 , 43 ]. While these experiments are important for vaccine design, the publicly available datasets could be used by malicious actors to inform the enhancement of pandemic pathogens. Beyond the generation of datasets with greater potential for misuse, improved computational methods mean that data can be more effectively used for malicious bioengineering [ 38 ]. Publication of detailed methods, for example, for the synthesis and engineering of pandemic pathogens, may also increase the risk of accidents and misuse. Detailed protocols may lower the tacit knowledge required to perform certain procedures, making them more accessible to bad actors, inappropriately qualified personnel, or personnel working in inappropriate facilities [ 44 ]. A recent protocol detailing how to synthesise recombinant SARS-CoV-2 exemplifies this [ 45 ]. The described "reverse genetic system can be used to rapidly engineer viruses with desired mutations to study the virus in vitro and in vivo" and "enable researchers from different research backgrounds to master the use of the reverse genetic system." Given the misuse potential of research objects like code, datasets, and protocols, approaches for risk mitigation are needed. Across digital research objects, there appears to be a trend towards increased modularisation, i.e., sharing information in dedicated, purpose built repositories, in contrast to supplementary materials. This modularisation may allow differential access to research products according to the risk that they represent. Curated repositories with greater access control could be used that allow reuse and verification when full public disclosure of a research object is inadvisable. Such repositories are already critical for life sciences that deal with personally identifiable information. Peer reviewers could be given access during the article submission process, and subsequent access controls could be defined based on the perceived risk of the information. When information cannot be shared publicly, access controlled repositories would allow researchers to get credit for digital research object curation and creation through publication of persistent and citable metadata [ 46 ]. The Harvard Dataverse ( https://dataverse.harvard.edu/ ) is an example of an existing repository that allows archiving of data and code with customisable restrictions and searchable metadata to facilitate discoverability. While not dedicated to protocols, the Open Science Framework does allow controlled access to all research objects that it hosts. If functionality of servers dedicated to protocols is needed, private workspaces, such as those available at protocols.io, may be suitable with adaptation. Nonpublic clinical datasets may represent a useful model for concerning biological datasets. Participant level data often cannot be shared openly due to concerns with anonymity, so databases that regulate access are required. Clinical Practice Research Datalink (CPRD), for example, collects a range of primary healthcare data from general practices across the UK [ 47 ]. Access to data is provided through a research data governance process, which includes screening of applicants and review of a protocol. Access is only granted to bona fide researchers with trustworthy funders, and researchers must "have viable plans that maintain public and professional trust, ensure the research is of public benefit, and are methodologically robust" ( https://cprd.com/Data-access ). There is a thriving open science community with extensive code sharing related to analysis of such datasets, facilitated by use of standardised data formats [ 48 ]. OpenSAFELY ( https://www.opensafely.org/ ) is another example of a project enabling open science while protecting patient confidentiality. Unlike CPRD, one of the leading influenza and coronavirus sequence sharing platforms, GISAID ( www.gisaid.org ), requires agreement to a Database Access Agreement, but that agreement does not have any statement regarding the purpose of the research. More comprehensive agreements may be warranted when there are concerns over data misuse. "Code only access," where analytic code is run on datasets but researchers do not see the data directly [ 49 , 50 ], is another possibility. If widespread access to code or data is desirable, an application programming interface (API; the "mechanism by which users communicate with computers, code and databases in an automated way" [ 51 ]) could be used so that certain model functions or data can be freely accessed while use for nefarious purposes is prevented. In the context of datasets, APIs have been recommended, for example, for public health bioinformatics [ 51 ]. The risk-aware rollout of the OpenAI API platform for the GPT-3 language model provides precedent. Access was initially limited while risks were assessed, and, based on the assessed risks, the API continues to limit use and imposes safety standards on third-party applications [ 52 ]. Access controls and APIs might unnecessarily interfere with scientific research while not effectively preventing misuse. In the context of personal information, processes for controlled access to the database of Genotypes and Phenotypes (dbGaP) have been criticised for being unnecessarily difficult [ 53 ]. Of particular concern are the criteria used to determine access at a global level. For example, the use of institutional track records to establish trustworthiness may conflict with ensuring equitable access to scientific findings. Difficult trade-offs will inevitably need to be made. Interdisciplinary discussion drawing on social sciences, legal, biosecurity, and life science expertise may create practical guidance for addressing these trade-offs as well as measures for evaluating success. The use of APIs and access-controlled repositories could have benefits outside improved security. The "FAIR Guiding Principles" identify 4 principles for shared research objects: They should be findable, accessible, interoperable, and reusable for machines and people [ 54 ]. Though these principles have largely been applied to data, they are intended to apply to other digital research objects including research software [ 55 ]. APIs may increase reusability of computational models and aspects of accessibility (i.e., ability to actually use a model [ 55 ]). Code is often difficult to run due to dependencies, computing power requirements, or need for specialist skills; APIs can make it easy for anyone to use software. Access-controlled repositories may facilitate interoperability, reuse, and findability through enforcing or encouraging standards for metadata with common vocabularies and appropriate documentation; much data that is available openly is currently challenging to use and find. It is worth noting that the "A" of FAIR, accessibility, is often qualified: "as open as possible and as closed as necessary" [ 56 ]. Limiting access on the basis of security or safety concerns would not necessarily contradict this. Developing suitable repositories may therefore provide an opportunity to improve adherence to the FAIR principles and encourage their adoption across a wider range of research objects. In the absence of appropriate sharing methods, an immediately implementable recommendation for data, code, and materials repositories is to require a declaration that the submitted information does not have dual-use potential according to a list of criteria and to reject those that do unless mitigation is in place. When data cannot be shared at all, verification and assessment of reproducibility may still be possible. For example, simulated data, a subset of the data that represents less risk, or data that has been redacted to remove concerning information could be shared. Publication of a checksum with the analysis code is a further option [ 57 ]. Preprint publishing: A changing publication landscape offers challenges and opportunities The use of preprints—author-formatted articles publicly deposited in a repository—in biological and medical sciences has recently increased considerably [ 58 ]. Preprints offer a faster route to publishing research than traditional journals and increase the accessibility and ability to rapidly reuse research. There are proposals for funders to mandate preprint posting [ 59 ], and several journals now act as "overlay journals," collating and reviewing articles that have already been published as preprints [ 60 , 61 ]. A key difference compared to journal articles is that some preprint servers do not screen scientific articles before they are made publicly available: In one analysis, 68% provided some form of screening or moderation before the article was made public [ 62 ]. Three popular servers for biological research screen all submissions before posting, though the scope of that screening differs ( Box 1 ). Box 1. Preliminary investigation into preprint policies reveals potential gaps We investigated the policies for screening preprints at 3 major preprint servers—bioRxiv, medRxiv, and arXiv—based on publicly available information. We read submission guidelines, frequently asked questions, and moderation policies, and searched for the server names along with terms like "dual-use" and "dangerous" to identify other relevant information. medRxiv submissions are screened for "material that could potentially endanger the health of individual patients or the public," which "may include, but is not limited to, studies describing dual-use research" [ 63 ]. When launched in 2019, a cofounder of medRxiv stated that they would "almost certainly not post" studies of pathogens that could cause harm [ 64 ], and medRxiv have historically "declined work involving pathogens of pandemic potential" [ 65 ]. During the COVID-19 pandemic, however, medRxiv began accepting work on SARS-CoV-2 because "knowledge about viral variants gained from this work should be disseminated rapidly" [ 65 ]. bioRxiv submissions "undergo a basic screening process for … material that might pose a health or biosecurity risk" [ 66 ] and "dangerous" [ 67 ] content. We were unable to find further public information on what research would be considered a biosecurity risk or dangerous. However, we have identified several papers describing viral engineering approaches posted on bioRxiv, suggesting a fairly permissive standard [ 68 , 69 ]. arXiv submissions are moderated, though the explanation of the moderation process [ 70 ] does not mention dual-use, safety, security, or similar terms. Since the scope of arXiv includes "quantitative biology" [ 71 ], this may be an important shortcoming. Articles presenting data or models that were rejected from bioRxiv or medRxiv due to security concerns might be permitted at arXiv. We are aware of at least one example of a paper that was not accepted at bioRxiv due to concerns about public health later being posted on arXiv [ 72 ]. The existence of relevant wording for bioRxiv and medRxiv is promising, though the lack of mention of security or safety in arXiv may be an important gap. A more comprehensive study including more servers and details of policy implementation is needed. Preprints may therefore remove the "gatekeeper" role that journals could play in mitigating risks from the publication of research with potential for misuse. Authors may select preprint servers that do not screen research. Unlike publishing in particular journals, there is little incentive to post to a particular preprint server, so little reason not to select one that will immediately post the article. Whether this is important depends on the role that journals play in preventing or altering publication of research with potential for misuse. There are many routes to making information available publicly over which journals have no control, such as personal websites, news articles, or conference presentations. However, there are examples where journals and editors have been important in evaluating risks from publication: For example, in 2014, information was redacted from 2 manuscripts about gene sequences of a novel Clostridium botulinum toxin following consultation between editors, authors, and branches of the US government [ 73 , 74 ]. Some journals also request that reviewers help to identify dual-use risks [ 73 ]. Preprints may therefore increase the probability that dangerous methods or results are described publicly. Preprints challenge any model relying on review by journals at publication [ 18 ], emphasising the need for oversight at other stages in the research life cycle, such as during design and funding. The need to consider preprints in the context of research with dual-use potential has been expressed previously [ 75 ]. As recommended for code, data, and materials, preprint servers could require a declaration that the posted research does not present dual-use potential and reject posting of articles that do not provide this. There are few prominent preprint servers in comparison to the number of journals, which may represent an opportunity for coordination. Relatively few parties would need to agree on a policy for it to be implemented universally, or at least across all of the major servers (such as bioRxiv, medRxiv, arXiv, OSF preprints, and preprints.org ). If an article is flagged by at least one server as potentially concerning, other servers could agree not to post that article until it was appropriately peer reviewed. However, authors must be incentivised to post to those preprint servers with appropriate review processes, and funders, institutions, and possibly journals, rather than researchers, will likely need to encourage this. Further research into the role that preprints play in influencing biosafety and biosecurity risks, policies currently in place, and proposals to mitigate those risks, would be valuable. Box 1. Preliminary investigation into preprint policies reveals potential gaps We investigated the policies for screening preprints at 3 major preprint servers—bioRxiv, medRxiv, and arXiv—based on publicly available information. We read submission guidelines, frequently asked questions, and moderation policies, and searched for the server names along with terms like "dual-use" and "dangerous" to identify other relevant information. medRxiv submissions are screened for "material that could potentially endanger the health of individual patients or the public," which "may include, but is not limited to, studies describing dual-use research" [ 63 ]. When launched in 2019, a cofounder of medRxiv stated that they would "almost certainly not post" studies of pathogens that could cause harm [ 64 ], and medRxiv have historically "declined work involving pathogens of pandemic potential" [ 65 ]. During the COVID-19 pandemic, however, medRxiv began accepting work on SARS-CoV-2 because "knowledge about viral variants gained from this work should be disseminated rapidly" [ 65 ]. bioRxiv submissions "undergo a basic screening process for … material that might pose a health or biosecurity risk" [ 66 ] and "dangerous" [ 67 ] content. We were unable to find further public information on what research would be considered a biosecurity risk or dangerous. However, we have identified several papers describing viral engineering approaches posted on bioRxiv, suggesting a fairly permissive standard [ 68 , 69 ]. arXiv submissions are moderated, though the explanation of the moderation process [ 70 ] does not mention dual-use, safety, security, or similar terms. Since the scope of arXiv includes "quantitative biology" [ 71 ], this may be an important shortcoming. Articles presenting data or models that were rejected from bioRxiv or medRxiv due to security concerns might be permitted at arXiv. We are aware of at least one example of a paper that was not accepted at bioRxiv due to concerns about public health later being posted on arXiv [ 72 ]. The existence of relevant wording for bioRxiv and medRxiv is promising, though the lack of mention of security or safety in arXiv may be an important gap. A more comprehensive study including more servers and details of policy implementation is needed. Preregistration: An opportunity for dual-use oversight Preregistration means archiving a time-stamped protocol that can be referred back to once a project is completed. This protocol is generally made public. Study registries, such as clinicaltrials.gov for clinical trials or PROSPERO for systematic reviews, represent a minimal form of preregistration where details of the study design and study outcomes are provided. In a stronger form, preregistration involves providing a detailed plan for the conduct and analysis of a study, including possibly the analytic code. Such plans are often registered on the Open Science Framework or As Predicted. Preregistration aims to limit duplication and selective reporting through disclosure of research plans [ 76 ] and reduce the prevalence of questionable research practices by clearly distinguishing planned and exploratory analyses [ 77 ]. It seems likely that greater consideration of the research before it is started, as encouraged by preregistration, could help to mitigate misuse risks. Currently, biosecurity risk assessment and management is not consistently conducted at any stage throughout the research lifecycle; preregistration could encourage greater consideration of risks at an early stage. Submission platforms could ask researchers to reflect on the dual-use potential of their work. In certain high-risk fields, platforms could request that details of hazard assessment be provided, which could be incentivised by journals requesting evidence of such assessments on publication. A safety and security form is required as part of the International Genetically Engineered Machine (iGEM) competition [ 78 ], which may be a useful model. In cases where researchers are unsure or do have concerns, they could be directed to an expert or relevant resources. Registered reports are a type of preregistration and journal article that may present an opportunity for oversight [ 79 ]. In registered reports, the introduction and methods (the stage one submission) are peer-reviewed before the work begins. If the stage one submission is accepted, the authors are guaranteed publication of the complete article (the stage two submission) regardless of the results, provided that they follow the proposed methods and the interpretation of the results is reasonable. Peer review at an earlier stage in the research life cycle represents a clear opportunity. When stage one submissions are received that could present biosecurity or biosafety risks, editors could assign biosecurity experts to risk-assess the article, in parallel with the traditional scientific peer review. This is similar to how registered reports currently operate, with specialists in different areas relevant to the manuscript (e.g., statistics, particular methods) reviewing submissions. A biosecurity report could detail recommendations on how the research could be conducted and reported to minimise risk, or, if necessary, advise against conduct altogether. In cases where review identifies risks, it remains an open question whether review reports should be made openly available. Sharing may encourage responsible discourse, but some arguments against conducting research may themselves present risk by highlighting how research could be misused. The format of registered reports and preregistrations is currently focussed on hypothesis-testing, confirmatory research. In general, concerning biological research is more likely to be exploratory, involving discovery or development of tools, and without methods and aims that can be defined adequately in advance. However, there are proposals in the open science community to adapt preregistrations to be suitable for exploratory work [ 80 ]. If implemented, early collaboration with biosecurity experts would be advantageous in ensuring that risk of misuse is one of the criteria considered. Proactively encouraging early trials of any new format of advisory peer review in areas of perceived high risk, such as synthetic mammalian virology, as suggested elsewhere [ 20 ], could be beneficial. Addressing dual-use risks at the early stages of the research lifecycle may be more effective than suppressing the dissemination of dangerous insights after work is completed. Therefore, interventions aimed at encouraging review at the conception of research seem particularly promising. The way forward We highlight several opportunities for reducing risks from research with dual-use potential. First, increased modularisation of research may facilitate differential sharing of research outputs depending on the risk they represent. There is a need to evaluate the suitability of existing tools that are used for sharing in terms of usability and security. To encourage the maximal adoption of tools that facilitate restricted access when needed, they must be as simple as possible to use from the perspective of both the researchers depositing materials and later users of those materials. They must also be secure. If existing tools are not suitable, new platforms may need to be developed. In either case, encouraging and monitoring the adoption of security-friendly sharing practices will be essential. Second, preprints may remove any gatekeeper role that journals play, as evidenced by the lack of screening by some preprint servers, emphasising the need for oversight throughout the research lifecycle rather than solely at the publication stage. Finally, preregistration and registered reports may encourage greater consideration of dual-use potential early in the research process. Existing preregistration formats will require adaptation to be suitable for this purpose. There is a need for guidance and input from individuals or organisations with experience in assessing research with dual-use potential to guide and pilot those adaptations in relevant communities. The concerns and proposals discussed here may be relevant to a range of research areas. Aspects of vaccine research, gene therapy research, and cancer therapeutics development may be associated with risks [ 81 ]. Considerable discussion has taken place in the artificial intelligence community about the potential for misuse of published code, such as in the context of deepfake videos for blackmail [ 82 ]. Studies that model terrorist scenarios are available, which might assist bad actors in predicting the impact of attacks. It is difficult to identify and anticipate all concerning research areas and formal regulation cannot keep pace [ 78 ]. It is therefore important to consider the possibility of research misuse when developing and implementing new open science tools and initiatives for general use. Incentives for open science require careful consideration. Many have been proposed, including changing hiring practices to support open data, open materials and preregistration [ 83 – 85 ], open science leaderboards [ 86 ], journal scores based on transparency [ 87 ], badges acknowledging open science on published papers [ 88 ], and assessment of open science practices by funders [ 89 ]. These incentives must allow limited disclosure when it is justified on the basis of safety or security concerns. Open data badges, for example, are available "if sensitive, personal, data are available only from an approved third party" [ 90 ] but not explicitly when the data exhibits safety or security risk. Researchers must not be penalised for responsible disclosure or incentivised to disclose irresponsibly. Any proposal to allow researchers to reduce public sharing could be exploited by those unwilling to invest the effort that open science requires. For example, researchers who have inadequately documented or fabricated data might invent security concerns. Alternatives to public sharing must therefore include appropriate verification, perhaps through peer review, to ensure that they are available for reuse when appropriate. Since this may increase the burden on reviewers and editors, there may be a need for individuals or organisations with appropriate expertise who are willing to and capable of providing this service. Funders interested in biosecurity and biosafety could support this. Funders and institutions have an important role to play in improving biosecurity. Storing data in approved platforms, posting only to preprint servers with adequate review processes, and preregistration of research with high potential for misuse, for example, could be mandated or encouraged by funders or institutional oversight groups. While solutions are developed and implemented, clear policies should be in place for the communication of research outputs that involve safety or security risks. Common to much discussion in this paper is the need for input from experts in risks of biological research. As these risks appear to be relatively neglected, this may be a key bottleneck in developing and implementing changes. Greater investment in expertise related to biosecurity and biosafety will likely be important for realisation of any proposals involving peer review for risk assessment or mitigation purposes. More generally, consideration of downside risk of both open science and biological science appears to be neglected in comparison to its plausible magnitude. Education and outreach may help to increase awareness among relevant stakeholders. Conclusions Open science practices may in many cases contribute positively to our ability to deal with biological threats. However, given the many concerning examples of biosafety and biosecurity incidents, the potential threat posed by the increasing accessibility and usability of scientific research to all actors cannot be ignored. Increased sharing of code, data, and materials in particular are concerning in some cases. There is an urgent need to address the inadvertent risks associated with certain open science practices and encourage responsible sharing and access. For preprints, the lack of screening in some cases challenges strategies relying on assessment of dual-use potential at the publication stage, but interventions may be possible and should be explored. Preregistration is a useful model that could encourage risk assessment and advisory peer review of research with dual-use potential earlier in the research lifecycle. In general, there is a need for ongoing, critical evaluation of incumbent and changing scientific practices, and consideration of the risks that such practices represent.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7126145/

Approach to Adult Patients with Acute Dyspnea

Undifferentiated patients in respiratory distress require immediate attention in the emergency department. Using a thorough history and clinical examination, clinicians can determine the most likely causes of dyspnea. Understanding the pathophysiology of the most common diseases contributing to dyspnea guides rational testing and informed, expedited treatment decisions. Key points • The cause of dyspnea is often evident from a complete history and physical examination. • Rapid determination of the cause of dyspnea saves lives. • Shortness of breath is not always primarily a pulmonary problem. • Understanding the pathophysiology of each disease allows clinicians to make rational decisions about testing. Introduction Emergency Medical Services (EMS) calls en route with a 45-year-old woman who has a history of congestive heart failure, chronic bronchitis, morbid obesity, and diabetes. She is breathing 40 times per minute, maintaining oxygen saturations of 94%. She appears mildly confused. You have 5 minutes to consider algorithms before the patient arrives. Perhaps respiratory therapy is paged to supply a ventilator or a biphasic positive airway pressure (BIPAP) machine. Maybe you prepare airway equipment or ask your nursing staff to access medication in advance, or you might use the time to expand your differential and determine what brief information regarding the patient's history and initial physical examination will help you treat her. Acutely dyspneic patients present in various ways. Are the lungs full of fluid or pus? Did the throat swell shut or is the patient just anxious? Did the patient aspirate a foreign body or have a slow or rapid hemorrhage? Is the patient compensating for a severe metabolic acidosis or did the patient run out of beta agonists at home? This article provides helpful guidelines in the assessment and management of these diverse patients. Respiratory distress is responsible for nearly 4 million emergency department (ED) visits each year and is one of the most common presenting complaints in the elderly. 1 When a patient presents with dyspnea, the primary task of the emergency physician is to assess for and ensure stability of the patient's airway, breathing, and circulation. In the case of dyspnea, presentations may range from minor symptoms to extremis. Rapid assessment may necessitate the use of intubation, BIPAP, nebulizations, decompression, or other therapies in the immediate period following the patient's arrival, to treat dyspnea. Patient evaluation The American Thoracic Society suggests that "dyspnea results from a … mismatch between central respiratory motor activity and incoming afferent information from receptors in the airways, lungs and chest wall structures." 2 This dissociation can result from increased metabolic demand, decreased compliance, increased dead-space volume, or many other disorders that are discussed later. Each patient presenting short of breath uses a different set of phrases to describe the symptoms and examination reveals a different combination of disorders. The clinician's ability to interpret these varying constellations is necessary to provide appropriate treatment to these patients, who are often in serious distress. History Acute dyspnea, or shortness of breath, is one of the most common chief complaints in the ED. The differential diagnosis includes many disorders that can be divided based on obstructive, parenchymal, cardiac, and compensatory features. A careful history can begin to narrow this wide differential. In addition to common symptoms, consider risk factors such as past medical and family history, trauma, travel, medications, and exposures. Schwartzstein and Lewis 3 use the analogy of a machine to identify different causes of dyspnea based on pathophysiologic data. Dysfunctions of the respiratory system may be caused by faulty controllers, ventilatory pumps, or gas exchangers ( Table 1 ). This table makes it easier to understand the causes of shortness of breath related to respiratory causes. Table 1 A systemic approach to dyspnea by assessing the components of the respiratory process Part Description Manifestations Examples Controller Malfunction presents as abnormal respiratory rate or depth. Often related to abnormal feedback to brain from other parts of the system Air hunger, need to breathe Abnormal feedback to brain from other systems. Metabolic acidosis, anxiety Ventilatory pump Composed of muscles, nerves that signal the controller, chest wall, and pleura that create negative thoracic pressure, airways and alveoli allowing flow from atmosphere and gas exchange Increased work of breathing, low tidal volumes Neuromuscular problems (eg, Guillain-Barré), decreased chest wall compliance, pneumothorax, pneumonia, bronchospasm (COPD, asthma) Gas exchanger Oxygen and carbon dioxide cross the pulmonary capillaries in the alveoli. Membrane destruction or interruption of the interface between the gas and capillaries by fluid or inflammatory cells limit gas exchange Increased respiratory drive, hypoxemia, chronic hypercapnia Emphysema, pneumonia, pulmonary edema, pleural effusion, hemothorax Adapted from Schwartzstein RM, Lewis A. Dyspnea. In: Broaddus V, Mason RC, Ernst JD, et al, editors. Murray & Nadel's textbook of respiratory medicine. Elsevier health sciences. 6th edition. Philadelphia: Saunders/Elsevier; 2015. Cardiovascular disease manifests as dyspnea by causing disruptions of the system that pumps oxygenated blood to tissues and then transports the carbon dioxide back to the lung. Decreases in cardiac output or increases in resistance limit oxygen delivery. Similarly, decreased oxygen carrying capacity in anemia plays a role in its presentation with dyspnea. Physical Examination A detailed physical examination also provides important guidance ( Table 2 ). Respiratory rate and oxygen saturation are obtained with vital signs. The clinician should assess the patient's work of breathing, looking for any tripoding or retractions. Crepitance in the chest may indicate subcutaneous air and pneumothorax. Lung sounds such as wheezing, rales, and rhonchi further guide the differential. Decreased sounds, hyperresonance, or egophony may also provide additional clues. Table 2 Physical examination findings and correlating diagnoses Symptom Differential Diagnosis Wheeze COPD/emphysema, asthma, allergic reaction, CHF (cardiac wheeze) Cough Pneumonia, asthma, COPD/emphysema Pleuritic chest pain Pneumonia, pulmonary embolism, pneumothorax, COPD, asthma Orthopnea Acute heart failure Fever Pneumonia, bronchitis, TB, malignancy Hemoptysis Pneumonia, TB, pulmonary embolism, malignancy Edema Acute heart failure, pulmonary embolism (unilateral) pulmonary edema Acute and chronic heart failure, end stage renal and liver diseases, ARDS (sepsis) Tachypnea alone pulmonary embolism, acidosis (including aspirin toxicity), anxiety Abbreviations: ARDS, acute respiratory distress syndrome; CHF, congestive heart failure; COPD, chronic obstructive pulmonary disease; TB, tuberculosis. Jugular venous distension, S3 gallop, and peripheral edema indicate that a patient has fluid overload. Conjunctival pallor, capillary refill, and temperature of extremities can provide clues about blood volume and general circulation. Pulses must also be assessed. Testing Multiple tests are available to narrow the differential diagnosis of acute dyspnea. When using tests to augment clinical decision making, be sure to weigh the information they may provide with any risks involved in performing the tests ( Table 3 ). Table 3 Diagnostic testing for dyspneic patients Test General Information Pros Cons Chest radiograph Often primary imaging Low radiation, can assess consolidation, pleural fluid, hyperinflation, pneumothorax, and subcutaneous air. Heart size is apparent Low sensitivity in acute dyspnea. In one series only 8 of 26 pneumonias diagnosed on CT met CXR criteria 37 Ultrasonography Multiple protocols to assess acute dyspnea No radiation, fast, reproducible bedside test, can be done on unstable patients in department and in semirecumbent position Requires some skill to acquire and interpret bedside images. Patient factors such as subcutaneous air, body habitus, and so forth may limit images D-dimer Marker of fibrinolytic activity. When measured by ELISA or second-generation latex agglutination can be used to rule out PE in selected patients Serum test readily available Requires risk assessment and clear clinical question. Also increased in consumptive coagulopathy, infection, malignancy, trauma, dissection, preeclampsia, and other cardiovascular disorders Arterial blood gas Provides additional information about ventilation (Pa co 2 ) to patients with reliable pulse oximetry and bicarbonate level available on BMP May be faster than general laboratory tests. Useful in assessing anxiety-induced hyperventilation 36 Limited evidence for routine use in undifferentiated dyspnea Electrocardiogram Initial cardiac assessment for assessing dyspnea Fast and inexpensive. Easy to compare with prior examinations. Specific for dysrhythmias or ACS limiting perfusion May be nonspecific in findings such as right heart strain and P pulmonale Troponin Serum indicator of myocardial damage Serum test readily available Can narrow differential to cardiac causes. PE with right heart strain may have increased troponin levels; this finding predicts worse outcomes BNP and proBNP Useful in assessing for acute heart failure Serum test readily available Limited in obesity, mitral regurgitation, flash pulmonary edema, and renal insufficiency. Context is essential Complete blood count Provides information about oxygen carrying capacity based on hemoglobin and hematocrit. White blood cell count may indicate infection Serum test readily available Nonspecific CT scan Provides detailed imaging of cardiorespiratory system. Use is increasing, but practitioners should maintain clinical context and consider whether other modalities can answer the clinical question Offers sensitive and specific results Significant radiation exposure, contrast nephropathy, intravenous contrast dye reactions Ventilation/perfusion scan Radiolabeled aerosol and albumin aggregates are used to study ventilation and perfusion. Read as negative or low, medium or high probability for pulmonary embolism Low in radiation Limited by underlying pulmonary disease and availability of isotopes Abbreviations: ACS, acute coronary syndrome; BMP, basic metabolic panel; BNP, B-type natriuretic peptide; CT, computed tomography; CXR, chest radiograph; ELISA, enzyme-linked immunosorbent assay; proBNP, pro–B-type natriuretic peptide. Ultrasonography provides valuable information about the origin of symptoms, and, often, diagnosis in the initial assessment of an acutely dyspneic patient. These images may be obtained during or shortly after initial assessment, potentially guiding therapy faster than laboratory tests or other imaging studies would be available. The Bedside Lung Ultrasonography in Emergency (BLUE) protocol offers one approach to differentiate several causes of respiratory failure ( Table 4 and Fig. 1 , Fig. 2 , Fig. 3 , Fig. 4 , Fig. 5 , Fig. 6 , Fig. 7 , Fig. 8 , Fig. 9 ). Table 4 BLUE protocol Ultrasonography Finding Ultrasonography Approach Description Clinical Meaning Image Assess for artifacts: A lines Anterior Subpleural air causes repeated linear artifacts parallel to the pleural line (horizontal) Air in lung: either normal or pneumothorax Fig. 1 Assess for artifacts: B lines Anterior Seven features: hyperechoic, well-defined, hydroaeric comet-tail artifacts arising from the pleural line. They spread upwards indefinitely and obscure A lines. When lung sliding is present, they move with the lung Represents an interface of 2 widely different transmissions of ultrasound waves: in this case, air and fluid. When 3 or more B lines are in a single interspace, they are B+ lines (or pulmonary rockets), indicating interstitial syndrome Fig. 2 Assess for lung sliding Anterior Absence of lung sliding occurs with a disruption of the normal sliding of viscera on parietal pleura or separation of the two. In M mode, absence of lung slide is seen as the stratosphere sign (also known as bar-code sign) Absence of lung sliding in the presence of A lines necessitates search for pneumothorax. Lung point is the ultrasonography finding in which lung slide is seen in the same view with the abolished lung slide and A lines in the same location, indicating the tip of the lung Fig. 3 (stratosphere sign) Fig. 4 (normal lung) Fig. 5 (lung point) Assess for alveolar consolidation or pleural effusion (posterolateral alveolar and/or pleural syndrome) Lateral subposterior The classic anechoic, dependent pattern may be inconsistent. Other findings include (1) quad sign: pleural effusion on expiration noted between the pleural and regular, lower lung lines (viscera). (2) Shred sign: tissuelike pattern seen with alveolar consolidation, with the upper border of lung line (or pleural line when there is no effusion) with an irregular lower border. (3) Sinusoid sign: movement of the lung line toward the pleural line in inspiration Pleural effusion: sinusoid, plus may have quad sign. Alveolar consolidation: tissuelike appearance or shred sign, absent lung line, absent sinusoidal sign Figs. 6 and 7 (pleural effusion) Fig. 8 (tissuelike lung) Fig. 9 (sinusoidal) Deep venous thrombosis Femoral veins Visualization of thrombus in the lumen or lack of compressibility is positive test Consider pulmonary embolus if positive — Adapted from Lichtenstein DA, Mezière GA. Relevance of lung ultrasound in the diagnosis of acute respiratory failure: the BLUE protocol. Chest 2008;134(1):117–25; and Courtesy of Petra Duran, MD, RDMS, University of Florida College of Medicine-Jacksonville (FL). Fig. 1 A lines. Fig. 2 B lines. Fig. 3 Stratosphere sign. Fig. 4 Normal lung. Fig. 5 Lung point. Fig. 6 Pleural effusion. Fig. 7 Pleural effusion. Fig. 8 Tissuelike lung. Fig. 9 Sinusoidal sign. Other protocols include assessments to assess for other cardiac causes of dyspnea. 4 , 5 , 6 Focused evaluation of global left ventricular function, diastolic function, right ventricular size, and any pericardial effusion facilitates rapid assessment for massive myocardial infarction, cardiac tamponade, and massive pulmonary embolism at the bedside. In addition, inferior vena cava measurement can be used to assess for right-sided heart failure and to estimate central venous pressure. Computed tomography (CT) use to evaluate acute dyspnea has increased in the last decade. 7 Risks include contrast reactions and nephropathy as well as radiation-induced cancers. 8 Recent American College of Physicians recommendations advocate avoidance of CT as an initial test to evaluate patients at low risk for pulmonary embolism (PE). 9 Further, nearly one-fourth of patients undergoing CT for PE evaluation have clinically significant incidental findings. Although CT may provide vital diagnostic information, clinicians must not only consider the scan's necessity but also plan appropriate follow-up for any clinically important incidental findings. 10 Always consider whether CT is necessary or whether less risky modalities, such as chest radiograph or ultrasonography, will answer pertinent questions. History Acute dyspnea, or shortness of breath, is one of the most common chief complaints in the ED. The differential diagnosis includes many disorders that can be divided based on obstructive, parenchymal, cardiac, and compensatory features. A careful history can begin to narrow this wide differential. In addition to common symptoms, consider risk factors such as past medical and family history, trauma, travel, medications, and exposures. Schwartzstein and Lewis 3 use the analogy of a machine to identify different causes of dyspnea based on pathophysiologic data. Dysfunctions of the respiratory system may be caused by faulty controllers, ventilatory pumps, or gas exchangers ( Table 1 ). This table makes it easier to understand the causes of shortness of breath related to respiratory causes. Table 1 A systemic approach to dyspnea by assessing the components of the respiratory process Part Description Manifestations Examples Controller Malfunction presents as abnormal respiratory rate or depth. Often related to abnormal feedback to brain from other parts of the system Air hunger, need to breathe Abnormal feedback to brain from other systems. Metabolic acidosis, anxiety Ventilatory pump Composed of muscles, nerves that signal the controller, chest wall, and pleura that create negative thoracic pressure, airways and alveoli allowing flow from atmosphere and gas exchange Increased work of breathing, low tidal volumes Neuromuscular problems (eg, Guillain-Barré), decreased chest wall compliance, pneumothorax, pneumonia, bronchospasm (COPD, asthma) Gas exchanger Oxygen and carbon dioxide cross the pulmonary capillaries in the alveoli. Membrane destruction or interruption of the interface between the gas and capillaries by fluid or inflammatory cells limit gas exchange Increased respiratory drive, hypoxemia, chronic hypercapnia Emphysema, pneumonia, pulmonary edema, pleural effusion, hemothorax Adapted from Schwartzstein RM, Lewis A. Dyspnea. In: Broaddus V, Mason RC, Ernst JD, et al, editors. Murray & Nadel's textbook of respiratory medicine. Elsevier health sciences. 6th edition. Philadelphia: Saunders/Elsevier; 2015. Cardiovascular disease manifests as dyspnea by causing disruptions of the system that pumps oxygenated blood to tissues and then transports the carbon dioxide back to the lung. Decreases in cardiac output or increases in resistance limit oxygen delivery. Similarly, decreased oxygen carrying capacity in anemia plays a role in its presentation with dyspnea. Physical Examination A detailed physical examination also provides important guidance ( Table 2 ). Respiratory rate and oxygen saturation are obtained with vital signs. The clinician should assess the patient's work of breathing, looking for any tripoding or retractions. Crepitance in the chest may indicate subcutaneous air and pneumothorax. Lung sounds such as wheezing, rales, and rhonchi further guide the differential. Decreased sounds, hyperresonance, or egophony may also provide additional clues. Table 2 Physical examination findings and correlating diagnoses Symptom Differential Diagnosis Wheeze COPD/emphysema, asthma, allergic reaction, CHF (cardiac wheeze) Cough Pneumonia, asthma, COPD/emphysema Pleuritic chest pain Pneumonia, pulmonary embolism, pneumothorax, COPD, asthma Orthopnea Acute heart failure Fever Pneumonia, bronchitis, TB, malignancy Hemoptysis Pneumonia, TB, pulmonary embolism, malignancy Edema Acute heart failure, pulmonary embolism (unilateral) pulmonary edema Acute and chronic heart failure, end stage renal and liver diseases, ARDS (sepsis) Tachypnea alone pulmonary embolism, acidosis (including aspirin toxicity), anxiety Abbreviations: ARDS, acute respiratory distress syndrome; CHF, congestive heart failure; COPD, chronic obstructive pulmonary disease; TB, tuberculosis. Jugular venous distension, S3 gallop, and peripheral edema indicate that a patient has fluid overload. Conjunctival pallor, capillary refill, and temperature of extremities can provide clues about blood volume and general circulation. Pulses must also be assessed. Testing Multiple tests are available to narrow the differential diagnosis of acute dyspnea. When using tests to augment clinical decision making, be sure to weigh the information they may provide with any risks involved in performing the tests ( Table 3 ). Table 3 Diagnostic testing for dyspneic patients Test General Information Pros Cons Chest radiograph Often primary imaging Low radiation, can assess consolidation, pleural fluid, hyperinflation, pneumothorax, and subcutaneous air. Heart size is apparent Low sensitivity in acute dyspnea. In one series only 8 of 26 pneumonias diagnosed on CT met CXR criteria 37 Ultrasonography Multiple protocols to assess acute dyspnea No radiation, fast, reproducible bedside test, can be done on unstable patients in department and in semirecumbent position Requires some skill to acquire and interpret bedside images. Patient factors such as subcutaneous air, body habitus, and so forth may limit images D-dimer Marker of fibrinolytic activity. When measured by ELISA or second-generation latex agglutination can be used to rule out PE in selected patients Serum test readily available Requires risk assessment and clear clinical question. Also increased in consumptive coagulopathy, infection, malignancy, trauma, dissection, preeclampsia, and other cardiovascular disorders Arterial blood gas Provides additional information about ventilation (Pa co 2 ) to patients with reliable pulse oximetry and bicarbonate level available on BMP May be faster than general laboratory tests. Useful in assessing anxiety-induced hyperventilation 36 Limited evidence for routine use in undifferentiated dyspnea Electrocardiogram Initial cardiac assessment for assessing dyspnea Fast and inexpensive. Easy to compare with prior examinations. Specific for dysrhythmias or ACS limiting perfusion May be nonspecific in findings such as right heart strain and P pulmonale Troponin Serum indicator of myocardial damage Serum test readily available Can narrow differential to cardiac causes. PE with right heart strain may have increased troponin levels; this finding predicts worse outcomes BNP and proBNP Useful in assessing for acute heart failure Serum test readily available Limited in obesity, mitral regurgitation, flash pulmonary edema, and renal insufficiency. Context is essential Complete blood count Provides information about oxygen carrying capacity based on hemoglobin and hematocrit. White blood cell count may indicate infection Serum test readily available Nonspecific CT scan Provides detailed imaging of cardiorespiratory system. Use is increasing, but practitioners should maintain clinical context and consider whether other modalities can answer the clinical question Offers sensitive and specific results Significant radiation exposure, contrast nephropathy, intravenous contrast dye reactions Ventilation/perfusion scan Radiolabeled aerosol and albumin aggregates are used to study ventilation and perfusion. Read as negative or low, medium or high probability for pulmonary embolism Low in radiation Limited by underlying pulmonary disease and availability of isotopes Abbreviations: ACS, acute coronary syndrome; BMP, basic metabolic panel; BNP, B-type natriuretic peptide; CT, computed tomography; CXR, chest radiograph; ELISA, enzyme-linked immunosorbent assay; proBNP, pro–B-type natriuretic peptide. Ultrasonography provides valuable information about the origin of symptoms, and, often, diagnosis in the initial assessment of an acutely dyspneic patient. These images may be obtained during or shortly after initial assessment, potentially guiding therapy faster than laboratory tests or other imaging studies would be available. The Bedside Lung Ultrasonography in Emergency (BLUE) protocol offers one approach to differentiate several causes of respiratory failure ( Table 4 and Fig. 1 , Fig. 2 , Fig. 3 , Fig. 4 , Fig. 5 , Fig. 6 , Fig. 7 , Fig. 8 , Fig. 9 ). Table 4 BLUE protocol Ultrasonography Finding Ultrasonography Approach Description Clinical Meaning Image Assess for artifacts: A lines Anterior Subpleural air causes repeated linear artifacts parallel to the pleural line (horizontal) Air in lung: either normal or pneumothorax Fig. 1 Assess for artifacts: B lines Anterior Seven features: hyperechoic, well-defined, hydroaeric comet-tail artifacts arising from the pleural line. They spread upwards indefinitely and obscure A lines. When lung sliding is present, they move with the lung Represents an interface of 2 widely different transmissions of ultrasound waves: in this case, air and fluid. When 3 or more B lines are in a single interspace, they are B+ lines (or pulmonary rockets), indicating interstitial syndrome Fig. 2 Assess for lung sliding Anterior Absence of lung sliding occurs with a disruption of the normal sliding of viscera on parietal pleura or separation of the two. In M mode, absence of lung slide is seen as the stratosphere sign (also known as bar-code sign) Absence of lung sliding in the presence of A lines necessitates search for pneumothorax. Lung point is the ultrasonography finding in which lung slide is seen in the same view with the abolished lung slide and A lines in the same location, indicating the tip of the lung Fig. 3 (stratosphere sign) Fig. 4 (normal lung) Fig. 5 (lung point) Assess for alveolar consolidation or pleural effusion (posterolateral alveolar and/or pleural syndrome) Lateral subposterior The classic anechoic, dependent pattern may be inconsistent. Other findings include (1) quad sign: pleural effusion on expiration noted between the pleural and regular, lower lung lines (viscera). (2) Shred sign: tissuelike pattern seen with alveolar consolidation, with the upper border of lung line (or pleural line when there is no effusion) with an irregular lower border. (3) Sinusoid sign: movement of the lung line toward the pleural line in inspiration Pleural effusion: sinusoid, plus may have quad sign. Alveolar consolidation: tissuelike appearance or shred sign, absent lung line, absent sinusoidal sign Figs. 6 and 7 (pleural effusion) Fig. 8 (tissuelike lung) Fig. 9 (sinusoidal) Deep venous thrombosis Femoral veins Visualization of thrombus in the lumen or lack of compressibility is positive test Consider pulmonary embolus if positive — Adapted from Lichtenstein DA, Mezière GA. Relevance of lung ultrasound in the diagnosis of acute respiratory failure: the BLUE protocol. Chest 2008;134(1):117–25; and Courtesy of Petra Duran, MD, RDMS, University of Florida College of Medicine-Jacksonville (FL). Fig. 1 A lines. Fig. 2 B lines. Fig. 3 Stratosphere sign. Fig. 4 Normal lung. Fig. 5 Lung point. Fig. 6 Pleural effusion. Fig. 7 Pleural effusion. Fig. 8 Tissuelike lung. Fig. 9 Sinusoidal sign. Other protocols include assessments to assess for other cardiac causes of dyspnea. 4 , 5 , 6 Focused evaluation of global left ventricular function, diastolic function, right ventricular size, and any pericardial effusion facilitates rapid assessment for massive myocardial infarction, cardiac tamponade, and massive pulmonary embolism at the bedside. In addition, inferior vena cava measurement can be used to assess for right-sided heart failure and to estimate central venous pressure. Computed tomography (CT) use to evaluate acute dyspnea has increased in the last decade. 7 Risks include contrast reactions and nephropathy as well as radiation-induced cancers. 8 Recent American College of Physicians recommendations advocate avoidance of CT as an initial test to evaluate patients at low risk for pulmonary embolism (PE). 9 Further, nearly one-fourth of patients undergoing CT for PE evaluation have clinically significant incidental findings. Although CT may provide vital diagnostic information, clinicians must not only consider the scan's necessity but also plan appropriate follow-up for any clinically important incidental findings. 10 Always consider whether CT is necessary or whether less risky modalities, such as chest radiograph or ultrasonography, will answer pertinent questions. Differential diagnosis for acutely dyspneic patients Obstructive Dyspnea Consider the 35-year-old woman discussed earlier. Medics report tachypnea with very poor air movement during transport. As she rolls through the ambulance bay doors, you are already assessing her. Adept clinicians can spot respiratory distress from across the room. She is diaphoretic, her shoulders are held adjacent to her ears, and she is breathing extremely rapidly with minimal air movement. You decide to aggressively treat her for a severe asthma exacerbation, starting BIPAP ventilation with continuous nebulized albuterol and order adjunct therapies including intravenous steroids, intravenous magnesium, and intramuscular epinephrine. After 20 minutes at her bedside, she begins to breathe more comfortably with the BIPAP machine and repeat auscultation reveals diffuse wheezing and improved air movement. As she begins to improve, EMS returns with another patient. His breath sounds are audible to everyone in the resuscitation bay. He appears diaphoretic and panicked. Examination reveals stridor, periorbital edema, tachycardia, and hypotension. Immediate intervention for anaphylactic shock begins and, after 2 rounds of epinephrine, fluid boluses, antihistamines, and steroids, he too begins to look better. Wheezing, or musical respiratory sounds, typically result from partial airway obstruction. 11 Because this obstruction can result from inflammation, secretions, or even a foreign body, patients with noisy or whistling breathing need close evaluation to determine whether the noise is inspiratory or expiratory, and whether it is from the lower airways or the upper airways. Stridor from a swollen airway, foreign body, or other airway obstruction is imminently dangerous. Although patients in anaphylaxis may benefit from the nebulized beta-agonist treatment used to treat an asthma exacerbation, it is not sufficient to save their lives. As opposed to wheezing, which is a lower airway expiratory sound, stridor is an upper airway sound transmitted when there is obstruction to the inflow of air during inspiration. The obstruction may be fixed (food bolus; Fig. 10 ) or inflammatory (anaphylaxis), but in any situation must be emergently managed. Fig. 10 Food bolus in airway. National and world organizations define asthma "by the history of respiratory symptoms such as wheeze, shortness of breath, chest tightness and cough that vary over time and in intensity, together with variable expiratory airflow limitation." 12 The reversibility of airflow obstruction is the hallmark distinguishing asthma from other obstructive respiratory disorders. In contrast, chronic obstructive pulmonary disease (COPD)/emphysema is defined as "persistent airflow limitation that is usually progressive and associated with enhanced chronic inflammatory responses in the airways and the lungs." 12 These patients also frequently wheeze, but may have a different course of acute and chronic disease. Table 5 highlights the differences between these similar, at times overlapping, diseases. Table 5 Features suggesting asthma or COPD Favors Asthma Favors COPD Onset in childhood Onset in adulthood Symptoms vary over time Symptoms worse at night Symptoms may be triggered by allergens or exercise Symptoms persist even with treatment Daily symptoms, some days better than others Chronic cough and sputum unrelated to triggers Variable airflow obstruction Persistent airflow obstruction Normal lung function when asymptomatic Abnormal lung function when asymptomatic Atopy in self or family Smoker No progression over time Progression over time Adapted from Global Initiative for Asthma. Diagnosis of diseases of chronic airflow limitation: asthma, COPD and asthma-COPD overlap syndrome (ACOS). 2015. Available at: http://www.ginasthma.org/documents/14/Asthma%2C-COPD-and-Asthma-COPD-Overlap-Syndrome-%28ACOS%29 . Accessed May 1, 2015. Asthma is an obstructive disease resulting from increased airway resistance. It is a reversible but recurrent chronic inflammatory disorder that characteristically causes severe dyspnea, wheezing, and coughing. 13 There are 2 main problems in asthma: chronically inflamed airways and hyperresponsive airways. Intermittent airflow obstruction in symptomatic patients results in decreased ability to expire, leading to hyperinflation, stenting open the alveoli, and increasing the work of breathing. Early in an exacerbation, symptoms are bronchospastic secondary to smooth muscle contraction. As an episode progresses, inflammatory changes in the airways can cause increased airway resistance and lead to VQ mismatch ( Fig. 11 ). The severity of an exacerbation can be assessed clinically and should dictate how aggressively a patient is treated ( Table 6 ). Fig. 11 Pathophysiology of symptom development in asthma. Table 6 Severity of asthma exacerbation assessment Symptoms Mild Moderate Severe Near Death Breathless While walking While talking At rest Decreased effort Speaking In sentences In phrases In words Unable Alertness May be agitated Usually agitated Usually agitated Confused Respiratory Rate (breaths/min) Increased Increased >30 >30, imminent failure Accessory Muscle Use Usually not Commonly Usually Usually Wheeze Moderate Loud Loud or silent Silent Heart Rate (beats/min) 120 ± Saturation (%) >95 92–94 2 wk with whoop or post-tussive vomiting Bordetella pertussis Structural lung disease (eg, bronchiectasis) P aeruginosa , Burkholderia cepacia , S aureus Injection drug use S aureus, anaerobes, M tuberculosis , S pneumoniae Endobronchial obstruction Anaerobes, S pneumonia , H influenzae , S aureus In context of bioterrorism Bacillus anthracis (anthrax), Yersinia pestis (plague), Francisella tularensis (tularemia) Abbreviations: CA-MRSA, community-acquired methicillin-resistant Staphylococcus aureus ; HIV, human immunodeficiency virus; SARS, severe acute respiratory syndrome. From Mandell LA, Wunderink RG, Anzueto A, et al. Infectious Diseases Society of America/American Thoracic Society consensus guidelines on the management of community-acquired pneumonia in adults. Clin Infect Dis 2007;44(Suppl 2):S46; with permission. The American Thoracic Society along with the Infectious Disease Society of America's consensus statement offers 4 important principles in the initial management and evaluation of adult patients with bacterial HAP, VAP, or HCAP; the most important to be accomplished in the ED is to promptly treat with "appropriate and adequate therapy" to decrease mortality. 24 Circulatory Dyspnea After a brief delay, you see a 28-year-old woman with shortness of breath and chest pain. She smokes, uses hormonal birth control, and reports that her symptoms started when she came back from a business trip. Pulmonary embolism (PE) interferes with both ventilation and perfusion. It ultimately causes circulatory collapse because of obstruction of right ventricular outflow eventually causing increased pulmonary artery pressure and failure of the right then left ventricles. Before circulatory collapse, echocardiography can show signs of right ventricular (RV) strain, including dilatation of the right ventricle, RV hypokinesis, paradoxic septal wall motion, McConnell sign (hypokinesis of the free RV wall with sparing of the apex), and tricuspid regurgitation. 25 Dresden and colleagues 26 supported the use of ultrasonography in moderate-risk to high-risk patients to determine whether the patients were appropriate for anticoagulation while awaiting definitive imaging. Early anticoagulation is recommended to improve mortality and there is evidence to support anticoagulation before diagnosis in patients with a Wells score greater than 4 who will have a delay to diagnosis of more than 1 hour and 40 minutes. 27 , 28 The assessment of patients with dyspnea and concern for PE requires a series of risk stratification. One common method is to use Wells criteria 29 ( Box 1 ) in patients with suspicion for PE; although other stratification tools exist, none has been shown to be clearly superior. When there is low clinical suspicion for PE, PERC (pulmonary embolism rule-out criteria) 30 rules or D-dimer testing may be applied. 27 If PERC ( Box 2 ) is negative, or there is intermediate pretest probability for PE with negative high-sensitivity D-dimer, no further testing for PE is required. 27 When further testing is needed (positive D-dimer or high-sensitivity D-dimer not available), negative CT angiogram or low-probability VQ scan may be used to rule out PE. Box 1 Wells criteria for pulmonary embolism • Clinical signs and symptoms of deep vein thrombosis (DVT): +3 • PE is the main diagnosis or equally likely: +3 • Heart Rate greater than 100 beats/min: +1.5 • Immobilization >3 days or surgery in last 4 weeks: +1.5 • History of prior PE/deep venous thrombosis (DVT): +1.5 • Hemoptysis: +1 • Malignancy with treatment within 6 months or palliative: +1 Less than or equal to 1.5 = low risk, 1.3% chance of PE in ED population; 2 to 6 = moderate risk, 16.2% chance of PE in ED population; greater than 6 = high risk, 40.6% chance of PE in ED population Adapted from Wells PS, Anderson DR, Rodger M, et al. Excluding pulmonary embolism at the bedside without diagnostic imaging: Management of patients with suspected pulmonary embolism presenting to the emergency department by using a simple clinical model and D-dimer. Ann Intern Med 2001;135:99. Box 2 Pulmonary embolism rule-out criteria • Further work-up recommended if any of the following are present: ○ Age greater than or equal to 50 years ○ Pulse greater than or equal to 100 beats/min ○ Oxygen saturation less than 95% ○ Hemoptysis ○ Unilateral leg swelling ○ History of PE/DVT ○ Exogenous estrogen use ○ Surgery or trauma within 4 weeks that required hospitalization or intubation • If none are present, probability of PE is less than 2%. Adapted from Kline JA, Courtney DM, Babrhel C, et al. Prospective multicenter evaluation of the pulmonary embolism rule-out criteria. J Thromb Haemost 2008;6(5):773. In the next bed is a middle-aged woman with diabetes complaining of shortness of breath today. It was associated with some vague nausea and she says that she just does not feel good. Angina pectoris is cardiac chest pain in which oxygen demand outweighs myocardial oxygen supply; in this case caused by occlusion of coronary arteries. Although typically chest pain is a part of the presentation, dyspnea alone may be the initial complaint, termed an anginal equivalent. In one recent large series of patients undergoing stress testing, patients with dyspnea alone were at increased risk of death from cardiac causes. Patients asked simply whether they experienced shortness of breath were considered dyspneic. The subset with no prior known coronary artery disease had more than 4 times the risk of sudden cardiac death versus asymptomatic patients and more than twice the risk of those with typical angina. 31 Clinicians should consider past medical history and risk factors when assessing dyspnea for cardiac causes such as acute myocardial infarction and acute coronary syndrome. Appropriate testing includes bedside electrocardiogram, troponin, and chest radiograph. The department eventually settles down and you are able to do some charting until a young man comes in with visible respiratory distress. He is tall and thin, smokes regularly, and reports sudden onset of severely painful breathing. Pneumothorax occurs when air enters the plural space between the chest wall and the lung. Typically only a thin serous layer exists between the visceral and parietal pleura. Air enters this potential space only when there is damage to the lung or chest wall, or a gas-producing pleural space infection. The classic risk factors for bleb rupture causing spontaneous pneumothorax are tall men, although smoking has been suggested to increase the risk of rupture by damaging the pleural layer. Pneumothoraces may be identified by ultrasonography, chest radiograph, or CT. Treatment may be guided by cause, severity, comorbidities, interventions such as positive pressure ventilation, size of the pneumothorax, and patient's preference. Recent studies suggest that uncomplicated spontaneous pneumothorax in patients not undergoing positive pressure ventilation may be treated as successfully with needle aspiration as with other more invasive chest drains, regardless of size. 32 Tension pneumothorax is a serious event requiring immediate needle decompression to avert loss of cardiac output and arrest. However, recent review shows that the classic presentation of tension pneumothorax with hypotension, absent breath sounds, and deviated trachea may not be immediately seen in patients with spontaneous, unassisted respiration. 33 Because of the slower development of the accumulation of air and pressure variations, spontaneously breathing patients may compensate much longer and present atypically, as shown in Table 8 . Thus, clinicians must remain vigilant. Table 8 Findings in tension pneumothorax Unassisted Ventilation Positive Pressure Ventilation Spontaneous respiration with air passing through 1-way flap Assisted ventilation forces air through pleural defect into pleural space Compensatory mechanisms delay collapse: • Tachycardia and accessory muscle use caused by tachypnea, increased tidal volume, and negative movement of the opposite side of the chest • BP is maintained because of limits in the pressure of the pneumothorax on mediastinum and hemithorax Sudden hemodynamic and respiratory compromise: • Sedation may increase inspiratory pressure • Intrapleural pressure is increased throughout respiratory cycle Venous siphon resulting from negative intrathoracic pressure in the opposite side of the chest returns blood to the heart Decreased venous return leads to hypotension and cardiac arrest Adapted from Roberts DJ, Leigh-smith S, Faris PD, et al. Clinical presentation of patients with tension pneumothorax: a systematic review. Ann Surg 2015;261(6):1069. Compensatory Dyspnea This article focuses on the cardiopulmonary system as the source of the problem in acutely dyspneic patients. It is important to also consider that the appearance of shortness of breath, tachypnea, or other typical symptoms of dyspnea may result from changing metabolic demands. These patients may appear, on the surface, to be in respiratory distress; they may be tachypneic, tachycardic, even pale or diaphoretic. In these cases, the clinician's responsibility is to identify and fix the true problem in order to improve the respiratory symptoms. Severely anemic patients have limited oxygen carrying capacity. Their bodies therefore experience oxygen hunger, which can manifest as shortness of breath. Patients with dysfunctional hemoglobins secondary to irreversibly bound atoms or toxins may also be functionally anemic with the same symptoms. People's bodies attempt at all costs to maintain equilibrium. Therefore, in metabolic acidoses (such as diabetic ketoacidosis), chemoreceptors detect acidosis and stimulate the respiratory center to hyperventilate. Both the rate and the depth of ventilation often increase, leading to both tachypnea and hyperpnea, at times referred to as Kussmaul respirations. This compensatory response is crucial for survival and should not be mistaken for dyspnea. It is equally important to realize that an increase in alveolar ventilation is not always a compensatory response (to acidosis or to primary pulmonary disorders) and hypocapnia may cause primary respiratory alkalosis, from central nervous system compromise, toxins (eg, salicylates), anxiety, or pain. 34 In these patients, imaging rarely reveals a source of dyspnea, but clinical suspicion based on history and examination, including signs such as the fruity breath of ketonemia, the pallor of anemia, or the cyanosis of toxic hemoglobinopathies, directs providers toward appropriate laboratory testing and treatment. Diagnosis of Exclusion In addition, sometimes dyspnea is not dyspnea. Acute anxiety and panic disorder can present as shortness of breath, tachypnea, or hyperventilation. Patients with panic disorder often describe symptoms similar to those of patients with true airway obstruction despite their normal pulmonary function. It has been suggested that these patients have abnormal proprioception, experiencing dyspnea without abnormal stimuli. 35 However, patients with a history of pulmonary disease can also have pure panic episodes. Arterial blood gas may be useful in diagnosing anxiety-related hyperventilation. 36 Severe pain can also induce abnormal respiratory patterns. Like compensatory problems, pain and anxiety can be managed by managing their causes. Treat pain. Reduce stress and anxiety with words, behaviors, or, if necessary, medications. However, air hunger and difficulty breathing also make individuals anxious. Be sure to avoid premature diagnosis of a purely anxiety-based concern without first evaluating for more dangerous disorders. Obstructive Dyspnea Consider the 35-year-old woman discussed earlier. Medics report tachypnea with very poor air movement during transport. As she rolls through the ambulance bay doors, you are already assessing her. Adept clinicians can spot respiratory distress from across the room. She is diaphoretic, her shoulders are held adjacent to her ears, and she is breathing extremely rapidly with minimal air movement. You decide to aggressively treat her for a severe asthma exacerbation, starting BIPAP ventilation with continuous nebulized albuterol and order adjunct therapies including intravenous steroids, intravenous magnesium, and intramuscular epinephrine. After 20 minutes at her bedside, she begins to breathe more comfortably with the BIPAP machine and repeat auscultation reveals diffuse wheezing and improved air movement. As she begins to improve, EMS returns with another patient. His breath sounds are audible to everyone in the resuscitation bay. He appears diaphoretic and panicked. Examination reveals stridor, periorbital edema, tachycardia, and hypotension. Immediate intervention for anaphylactic shock begins and, after 2 rounds of epinephrine, fluid boluses, antihistamines, and steroids, he too begins to look better. Wheezing, or musical respiratory sounds, typically result from partial airway obstruction. 11 Because this obstruction can result from inflammation, secretions, or even a foreign body, patients with noisy or whistling breathing need close evaluation to determine whether the noise is inspiratory or expiratory, and whether it is from the lower airways or the upper airways. Stridor from a swollen airway, foreign body, or other airway obstruction is imminently dangerous. Although patients in anaphylaxis may benefit from the nebulized beta-agonist treatment used to treat an asthma exacerbation, it is not sufficient to save their lives. As opposed to wheezing, which is a lower airway expiratory sound, stridor is an upper airway sound transmitted when there is obstruction to the inflow of air during inspiration. The obstruction may be fixed (food bolus; Fig. 10 ) or inflammatory (anaphylaxis), but in any situation must be emergently managed. Fig. 10 Food bolus in airway. National and world organizations define asthma "by the history of respiratory symptoms such as wheeze, shortness of breath, chest tightness and cough that vary over time and in intensity, together with variable expiratory airflow limitation." 12 The reversibility of airflow obstruction is the hallmark distinguishing asthma from other obstructive respiratory disorders. In contrast, chronic obstructive pulmonary disease (COPD)/emphysema is defined as "persistent airflow limitation that is usually progressive and associated with enhanced chronic inflammatory responses in the airways and the lungs." 12 These patients also frequently wheeze, but may have a different course of acute and chronic disease. Table 5 highlights the differences between these similar, at times overlapping, diseases. Table 5 Features suggesting asthma or COPD Favors Asthma Favors COPD Onset in childhood Onset in adulthood Symptoms vary over time Symptoms worse at night Symptoms may be triggered by allergens or exercise Symptoms persist even with treatment Daily symptoms, some days better than others Chronic cough and sputum unrelated to triggers Variable airflow obstruction Persistent airflow obstruction Normal lung function when asymptomatic Abnormal lung function when asymptomatic Atopy in self or family Smoker No progression over time Progression over time Adapted from Global Initiative for Asthma. Diagnosis of diseases of chronic airflow limitation: asthma, COPD and asthma-COPD overlap syndrome (ACOS). 2015. Available at: http://www.ginasthma.org/documents/14/Asthma%2C-COPD-and-Asthma-COPD-Overlap-Syndrome-%28ACOS%29 . Accessed May 1, 2015. Asthma is an obstructive disease resulting from increased airway resistance. It is a reversible but recurrent chronic inflammatory disorder that characteristically causes severe dyspnea, wheezing, and coughing. 13 There are 2 main problems in asthma: chronically inflamed airways and hyperresponsive airways. Intermittent airflow obstruction in symptomatic patients results in decreased ability to expire, leading to hyperinflation, stenting open the alveoli, and increasing the work of breathing. Early in an exacerbation, symptoms are bronchospastic secondary to smooth muscle contraction. As an episode progresses, inflammatory changes in the airways can cause increased airway resistance and lead to VQ mismatch ( Fig. 11 ). The severity of an exacerbation can be assessed clinically and should dictate how aggressively a patient is treated ( Table 6 ). Fig. 11 Pathophysiology of symptom development in asthma. Table 6 Severity of asthma exacerbation assessment Symptoms Mild Moderate Severe Near Death Breathless While walking While talking At rest Decreased effort Speaking In sentences In phrases In words Unable Alertness May be agitated Usually agitated Usually agitated Confused Respiratory Rate (breaths/min) Increased Increased >30 >30, imminent failure Accessory Muscle Use Usually not Commonly Usually Usually Wheeze Moderate Loud Loud or silent Silent Heart Rate (beats/min) 120 ± Saturation (%) >95 92–94 2 wk with whoop or post-tussive vomiting Bordetella pertussis Structural lung disease (eg, bronchiectasis) P aeruginosa , Burkholderia cepacia , S aureus Injection drug use S aureus, anaerobes, M tuberculosis , S pneumoniae Endobronchial obstruction Anaerobes, S pneumonia , H influenzae , S aureus In context of bioterrorism Bacillus anthracis (anthrax), Yersinia pestis (plague), Francisella tularensis (tularemia) Abbreviations: CA-MRSA, community-acquired methicillin-resistant Staphylococcus aureus ; HIV, human immunodeficiency virus; SARS, severe acute respiratory syndrome. From Mandell LA, Wunderink RG, Anzueto A, et al. Infectious Diseases Society of America/American Thoracic Society consensus guidelines on the management of community-acquired pneumonia in adults. Clin Infect Dis 2007;44(Suppl 2):S46; with permission. The American Thoracic Society along with the Infectious Disease Society of America's consensus statement offers 4 important principles in the initial management and evaluation of adult patients with bacterial HAP, VAP, or HCAP; the most important to be accomplished in the ED is to promptly treat with "appropriate and adequate therapy" to decrease mortality. 24 Circulatory Dyspnea After a brief delay, you see a 28-year-old woman with shortness of breath and chest pain. She smokes, uses hormonal birth control, and reports that her symptoms started when she came back from a business trip. Pulmonary embolism (PE) interferes with both ventilation and perfusion. It ultimately causes circulatory collapse because of obstruction of right ventricular outflow eventually causing increased pulmonary artery pressure and failure of the right then left ventricles. Before circulatory collapse, echocardiography can show signs of right ventricular (RV) strain, including dilatation of the right ventricle, RV hypokinesis, paradoxic septal wall motion, McConnell sign (hypokinesis of the free RV wall with sparing of the apex), and tricuspid regurgitation. 25 Dresden and colleagues 26 supported the use of ultrasonography in moderate-risk to high-risk patients to determine whether the patients were appropriate for anticoagulation while awaiting definitive imaging. Early anticoagulation is recommended to improve mortality and there is evidence to support anticoagulation before diagnosis in patients with a Wells score greater than 4 who will have a delay to diagnosis of more than 1 hour and 40 minutes. 27 , 28 The assessment of patients with dyspnea and concern for PE requires a series of risk stratification. One common method is to use Wells criteria 29 ( Box 1 ) in patients with suspicion for PE; although other stratification tools exist, none has been shown to be clearly superior. When there is low clinical suspicion for PE, PERC (pulmonary embolism rule-out criteria) 30 rules or D-dimer testing may be applied. 27 If PERC ( Box 2 ) is negative, or there is intermediate pretest probability for PE with negative high-sensitivity D-dimer, no further testing for PE is required. 27 When further testing is needed (positive D-dimer or high-sensitivity D-dimer not available), negative CT angiogram or low-probability VQ scan may be used to rule out PE. Box 1 Wells criteria for pulmonary embolism • Clinical signs and symptoms of deep vein thrombosis (DVT): +3 • PE is the main diagnosis or equally likely: +3 • Heart Rate greater than 100 beats/min: +1.5 • Immobilization >3 days or surgery in last 4 weeks: +1.5 • History of prior PE/deep venous thrombosis (DVT): +1.5 • Hemoptysis: +1 • Malignancy with treatment within 6 months or palliative: +1 Less than or equal to 1.5 = low risk, 1.3% chance of PE in ED population; 2 to 6 = moderate risk, 16.2% chance of PE in ED population; greater than 6 = high risk, 40.6% chance of PE in ED population Adapted from Wells PS, Anderson DR, Rodger M, et al. Excluding pulmonary embolism at the bedside without diagnostic imaging: Management of patients with suspected pulmonary embolism presenting to the emergency department by using a simple clinical model and D-dimer. Ann Intern Med 2001;135:99. Box 2 Pulmonary embolism rule-out criteria • Further work-up recommended if any of the following are present: ○ Age greater than or equal to 50 years ○ Pulse greater than or equal to 100 beats/min ○ Oxygen saturation less than 95% ○ Hemoptysis ○ Unilateral leg swelling ○ History of PE/DVT ○ Exogenous estrogen use ○ Surgery or trauma within 4 weeks that required hospitalization or intubation • If none are present, probability of PE is less than 2%. Adapted from Kline JA, Courtney DM, Babrhel C, et al. Prospective multicenter evaluation of the pulmonary embolism rule-out criteria. J Thromb Haemost 2008;6(5):773. In the next bed is a middle-aged woman with diabetes complaining of shortness of breath today. It was associated with some vague nausea and she says that she just does not feel good. Angina pectoris is cardiac chest pain in which oxygen demand outweighs myocardial oxygen supply; in this case caused by occlusion of coronary arteries. Although typically chest pain is a part of the presentation, dyspnea alone may be the initial complaint, termed an anginal equivalent. In one recent large series of patients undergoing stress testing, patients with dyspnea alone were at increased risk of death from cardiac causes. Patients asked simply whether they experienced shortness of breath were considered dyspneic. The subset with no prior known coronary artery disease had more than 4 times the risk of sudden cardiac death versus asymptomatic patients and more than twice the risk of those with typical angina. 31 Clinicians should consider past medical history and risk factors when assessing dyspnea for cardiac causes such as acute myocardial infarction and acute coronary syndrome. Appropriate testing includes bedside electrocardiogram, troponin, and chest radiograph. The department eventually settles down and you are able to do some charting until a young man comes in with visible respiratory distress. He is tall and thin, smokes regularly, and reports sudden onset of severely painful breathing. Pneumothorax occurs when air enters the plural space between the chest wall and the lung. Typically only a thin serous layer exists between the visceral and parietal pleura. Air enters this potential space only when there is damage to the lung or chest wall, or a gas-producing pleural space infection. The classic risk factors for bleb rupture causing spontaneous pneumothorax are tall men, although smoking has been suggested to increase the risk of rupture by damaging the pleural layer. Pneumothoraces may be identified by ultrasonography, chest radiograph, or CT. Treatment may be guided by cause, severity, comorbidities, interventions such as positive pressure ventilation, size of the pneumothorax, and patient's preference. Recent studies suggest that uncomplicated spontaneous pneumothorax in patients not undergoing positive pressure ventilation may be treated as successfully with needle aspiration as with other more invasive chest drains, regardless of size. 32 Tension pneumothorax is a serious event requiring immediate needle decompression to avert loss of cardiac output and arrest. However, recent review shows that the classic presentation of tension pneumothorax with hypotension, absent breath sounds, and deviated trachea may not be immediately seen in patients with spontaneous, unassisted respiration. 33 Because of the slower development of the accumulation of air and pressure variations, spontaneously breathing patients may compensate much longer and present atypically, as shown in Table 8 . Thus, clinicians must remain vigilant. Table 8 Findings in tension pneumothorax Unassisted Ventilation Positive Pressure Ventilation Spontaneous respiration with air passing through 1-way flap Assisted ventilation forces air through pleural defect into pleural space Compensatory mechanisms delay collapse: • Tachycardia and accessory muscle use caused by tachypnea, increased tidal volume, and negative movement of the opposite side of the chest • BP is maintained because of limits in the pressure of the pneumothorax on mediastinum and hemithorax Sudden hemodynamic and respiratory compromise: • Sedation may increase inspiratory pressure • Intrapleural pressure is increased throughout respiratory cycle Venous siphon resulting from negative intrathoracic pressure in the opposite side of the chest returns blood to the heart Decreased venous return leads to hypotension and cardiac arrest Adapted from Roberts DJ, Leigh-smith S, Faris PD, et al. Clinical presentation of patients with tension pneumothorax: a systematic review. Ann Surg 2015;261(6):1069. Compensatory Dyspnea This article focuses on the cardiopulmonary system as the source of the problem in acutely dyspneic patients. It is important to also consider that the appearance of shortness of breath, tachypnea, or other typical symptoms of dyspnea may result from changing metabolic demands. These patients may appear, on the surface, to be in respiratory distress; they may be tachypneic, tachycardic, even pale or diaphoretic. In these cases, the clinician's responsibility is to identify and fix the true problem in order to improve the respiratory symptoms. Severely anemic patients have limited oxygen carrying capacity. Their bodies therefore experience oxygen hunger, which can manifest as shortness of breath. Patients with dysfunctional hemoglobins secondary to irreversibly bound atoms or toxins may also be functionally anemic with the same symptoms. People's bodies attempt at all costs to maintain equilibrium. Therefore, in metabolic acidoses (such as diabetic ketoacidosis), chemoreceptors detect acidosis and stimulate the respiratory center to hyperventilate. Both the rate and the depth of ventilation often increase, leading to both tachypnea and hyperpnea, at times referred to as Kussmaul respirations. This compensatory response is crucial for survival and should not be mistaken for dyspnea. It is equally important to realize that an increase in alveolar ventilation is not always a compensatory response (to acidosis or to primary pulmonary disorders) and hypocapnia may cause primary respiratory alkalosis, from central nervous system compromise, toxins (eg, salicylates), anxiety, or pain. 34 In these patients, imaging rarely reveals a source of dyspnea, but clinical suspicion based on history and examination, including signs such as the fruity breath of ketonemia, the pallor of anemia, or the cyanosis of toxic hemoglobinopathies, directs providers toward appropriate laboratory testing and treatment. Diagnosis of Exclusion In addition, sometimes dyspnea is not dyspnea. Acute anxiety and panic disorder can present as shortness of breath, tachypnea, or hyperventilation. Patients with panic disorder often describe symptoms similar to those of patients with true airway obstruction despite their normal pulmonary function. It has been suggested that these patients have abnormal proprioception, experiencing dyspnea without abnormal stimuli. 35 However, patients with a history of pulmonary disease can also have pure panic episodes. Arterial blood gas may be useful in diagnosing anxiety-related hyperventilation. 36 Severe pain can also induce abnormal respiratory patterns. Like compensatory problems, pain and anxiety can be managed by managing their causes. Treat pain. Reduce stress and anxiety with words, behaviors, or, if necessary, medications. However, air hunger and difficulty breathing also make individuals anxious. Be sure to avoid premature diagnosis of a purely anxiety-based concern without first evaluating for more dangerous disorders. Summary Acute dyspnea presents commonly to the ED and it is imperative that emergency physicians be prepared to stabilize patients' oxygenation and ventilation, which requires careful and efficient consideration of the differential diagnosis. Using cues from the history and physical examination, practitioners may guide the work-up and treatment to identify a parenchymal, obstructive, circulatory, or compensatory cause of dyspnea. Early use of bedside testing, including ultrasonography, may limit unnecessary tests and save time in determining the best treatment course. Thus ensuring both the best care for the patient and also the physician's ability to readily respond to the next case.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10048445/

Assessment of Genetic and Health Management of Tunisian Holstein Dairy Herds with a Focus on Longevity

In Tunisia, the recognition of the possibility of including longevity and disease resistance in dairy cattle selection objectives has been hypothesized as a useful strategy by both researchers and producers. However, in this paper, the state of the art, with a focus on health and longevity, is reviewed. Along the same lines, the heritability for the milk traits, fertility traits, and longevity of Tunisian Holstein dairy cows complies with the literature. Therefore, the influence of genetics on some diseases of the dairy cow was investigated. In addition, a decreasing efficiency in cow fertility has been observed over the last few years. The results showed that the risk of culling increased with common diseases. When analyzed with the Weibull model, functional lifespan was strongly influenced by milk yield; therefore, the risk increased with a reduced milk yield. In her first three lactations, the relative risk of selection increased gradually with lactation. Thus, the risk of thinning is highest at the beginning and end of the first feeding and the end of her second feeding. In conclusion, the risk of culling was reduced in parity. The factors that influence the life of the herd, such as health, husbandry, environmental conditions, and management, are often ignored when evaluating longevity. 1. Introduction For the modern dairy cow, advances in genetics and breeding for productivity have resulted in an increased incidence of health disorders and reduced longevity. However, to maintain farm sustainability, farmers need to optimize the balance between maximum production and minimum production costs [ 1 ]. Moreover, reduced profitability is associated with the costs of dairy herd health and fertility, which are also major causes of involuntary culling. Nevertheless, reducing the incidence of disease in dairy cows is of economic, social, and environmental importance. Therefore, dairy cattle selection around the world has focused on increasing milk production due to consumer demand and the impact of production on farm profit margins. This was extremely successful through the combination of genetic selection with improvements in nutrition and health management. On the other hand, Oltenacu and Broom [ 2 ] indicated that inefficiencies exist because the increased production has led to negative effects on health, reproduction, and longevity. Miglior et al. [ 3 ] insisted on the inclusion of functional traits, such as fertility, health, and longevity, and reported that these traits have both economic and socioeconomic impacts through improving animal welfare and the sustainability of dairy production. Brickell and Wathes [ 4 ] suggested that extending the productive life of cows reduces replacement costs, allows for more limited selection, and increases the potential milk yields from adult cows, thus improving milk production. Generally, the culling rate was higher for low-producing cows and older ages at first calving. In addition, higher-producing cows are found to be culled at an earlier age than low-producing ones. M'Hamdi et al. [ 5 ] concluded that cows at the start and end of the first lactation and at the end of all other lactations were at the highest risk of culling. Tunisian dairy herds reported an average culling rate of 15–22% [ 5 ], with more than 50% of the cullings resulting from involuntary causes such as infertility, mastitis, and lameness [ 5 , 6 ]. In addition, Heikkilä et al. [ 7 ] reported that mastitis and lameness are the costliest diseases for dairy farmers as they cause sharp declines in milk production and farm income. Agiri et al. [ 8 ] reported that the true herd lifespan of Tunisian Holstein-Friesian cattle averaged 41.99 months, corresponding to a production life of 3.5 years, and the number of cows culled after the first two lactations reported that only 7.14% of cattle remained, compared to 57% in their fifth lactation. The genetic improvement program of Tunisian Holstein dairy cow has been elaborated since the 1960s to strengthen this sector; recording, performances control, AI, and Herd-Book are implemented and performed by the «Office de l'Elevage et des Pâturages (OEP)» [ 9 ]. Large phenotypic databases are continuously evaluated in collaboration with universities to track the pedigree of herds of animals, register their performance, and use them to implement herd genetic tendencies and management. This paper reviews the main results of the genetic and health management of Tunisian Holstein cattle regarding the situation across the world. 2. Animal Health, Diseases, and Welfare The health and welfare of all animals in the dairy herd are paramount for profitable and efficient milk production. The disease is often measured by economic impact, but animal health is also part of animal welfare. The pain and discomfort caused by health problems affect an animal's well-being, so animals must be in good health to improve animal welfare [ 6 ]. Records of the incidence and prevalence of various diseases are made more readily available through farm record-keeping systems. Producers must be able to correctly identify specific animal health problems early to improve animal welfare and herd health. Unfortunately, in Tunisia, there is no systematic recording system of health events in dairy herds [ 5 ]. This lack of information concerning health problems and their impact on dairy cow productivity prohibits work on the genetic analysis of the health and diseases of the dairy herd [ 5 ]. However, evidence of culling can partially compensate for this lack of information. Examining data for screening can be a cost-effective alternative compared to the costs of collecting, storing, and analyzing data on health disorders. Mastitis and lameness are major problems in Tunisian herds. Mastitis is a production, food quality, and safety issue. From an animal welfare perspective, it is a localized and painful infection for cows that can cause systemic illness leading to fever, dehydration, depression, and even death, depending on the type of infection and the cow's resistance. However, mastitis has been assessed during somatic cell counting. A prospective study of 21 selected dairy farms in northern Tunisia was conducted by Mtaallah et al. [ 10 ] to assess the reduction in milk yield due to high somatic cell counts in bulk tanks and to find associations between the risk factors and asymptomatic mastitis. The authors found that the average somatic cell count of bulk milk was 626,103 cells/mL and the average milk loss due to the somatic cell count in bulk tanks was 524 kg per cow per year. The risk factors associated with high in-tank somatic cell counts include: (i) livestock risk factors (inadequate bedding area, inadequate cleaning of bedding and waste areas); and (ii) milking risk factors (tap washing with showers without adjustable flow and individual towels. No wiping, more than five milking shifts per cow-herd; no overspray before milking or milking healthy and mastitis cows at the same time; no teat dipping). M'Hamdi et al. [ 11 ] analyzed the data, consisting of 73,189 test-day records of somatic cell counts, for the three first lactations of 8350 Holstein cows calving between 1997 and 2003 in 114 dairy herds. The results showed that the milk yield (MY) was largest in the second control, at 25 ± 9 kg, and lowest in the tenth control (14 ± 6 kg). Lameness among dairy cows is widely recognized as one of the most serious (and costly) animal welfare issues affecting dairy cattle. Lameness is the third most common infection in cows in modern barns. Environmental factors (diet, stables, injuries) and genetic factors are responsible for this condition. Lameness is also recognized as an important welfare problem, causing pain, and impairing the cow's ability to exhibit normal behavior [ 12 , 13 ]. In a study conducted by Ferchichi et al. [ 14 ] in Tunisian dairy herds, the incidence of lame cows averaged 67%. However, Bouraoui et al. [ 15 ] showed that podal pathologies have an incidence equal to 38.71% (score > 2). They reported that the incidence of lameness was approximately 37% and 99% in the second and third parities, respectively. The authors found that the prevalence was higher in heifers than in primiparous cows and that lameness occurred more frequently in Winter and Autumn than in Summer and Spring because the animals were reared under temperate climate conditions, where the cows' environment may be wetter in Winter and Autumn. The rate of lameness increased by 2%. Regarding the most economically important disorders in dairy cows, mastitis, infertility, and claw and leg disorders are listed [ 16 ]. A Tunisian study was conducted on 35 dairy farms to assess the welfare quality of Tunisian Holstein cattle based on several animal welfare indicators validated by the European Welfare Quality Project. The avoidance distance (on the face and in the stall), physical condition, lameness, fertility, somatic cell count (SCC; cells/mL), and milk yield were assessed [ 17 ]. The main results showed that the SCC averaged 427.3 ± 90.12 10 3 , being the highest in Autumn and associated with milk yield. Milk yield increased with the number of lactations and varied by the lactation stage. Smaller farms had lower somatic cell counts. The same study reported that the body condition score (BCS) ranged between 1.25 and 4 (lactating cattle) using a BCS scale between one and five. Most cows presented a BC score of 2.5 (50% of cows); however, most dry cows presented a BCS of 2.75 (65% of cows), ranging between a BCS of 1.5 and 4. A BCS of two or less was classified as 'thin'. The mean number of lactating cows in this category on all farms was 18.9 ± 1.9%. As for lameness, the proportion decreased, with only 19 out of 350 cows (5.4%) showing moderate lameness. Lameness appears to have been the greatest welfare problem within the parameters investigated. In general, the avoidance distances are short, which is an indicator of good human-animal relations and may reflect good farming practices [ 17 ]. In dairy herds, some bacterial diseases ( Bovine tuberculosis, Campylobacter enteritis , Anthrax, Hemorrhagic septicemia, Mannheimia haemolytica , and Contagious bovine pleuropneumonia, etc.) are of paramount importance, particularly those considered zoonotic. Among these, Bovine tuberculosis (caused by infection with Mycobacterium bovis ) is perhaps the most problematic. Heritability estimates on the observation and responsibility scales varied between 0.06 and 0.18; the standard errors varied between 0.012 and 0.044. The presumption of inheritance was based on the tuberculin test response and the presence of tuberculosis lesions confirmed during slaughterhouse testing. These results demonstrate that gene truncation can achieve a significant improvement in tuberculosis resistance [ 18 ]. Bovine tuberculosis (bTB) is considered a major zoonosis in Tunisia and a break to the intensification of production. Tunisia has had a national bTB control program since 1984. It is based on the intradermal tuberculin skin testing of dairy cows and regular meat inspections at the slaughterhouse. Nonetheless, bTB remains prevalent, mainly in the private sector, where disease control is based on sparse veterinary practices and slaughterhouse testing without routine intradermal tuberculin skin testing [ 19 ]. In their synthesis on animal health and disease genetics, Berry et al. [ 20 ] concluded that the accurate quantification of genetic trends in most health traits is not possible due to the lack of the routine availability of accurate animal health records and data in most countries. Nonetheless, past genetic trends may be predicted based on the estimated genetic correlations with the production traits, and the impact of these correlations can be quantified using the knowledge of past breeding goals. 3. Genetic Evaluation of Longevity Due to its high economic value, longevity is an important part of dairy cattle breeding goals in many countries. Imbayarwo-Chikosi et al. [ 21 ] attribute the high economic value of longevity to the herd trait dynamics that depend on the degree of voluntary and involuntary selection. Reducing involuntary culling increases the chances of voluntary culling and keeping high-yielding cows for longer [ 21 ]. Longevity employs different trait definitions [ 22 , 23 ]. All of these were based on age at culling, or death (uncensored) or censorship, and survived to the indicated age or period during or between lactations [ 24 , 25 ]. Lifespan appears as a threshold whose expression is not continuous but has a distinct categorical phenotype. Therefore, there are complete and incomplete records for survival dates. Events such as culling, and death are uncensored as they may be known. At the same time, animals may have been lost in pursuit, and no culling or death events are known to have occurred. Animals may also be alive at the time of analysis, so only a lower bound on the productive lifespan is known. Appropriate modeling strategies for such data are required to account for these intrinsic traits without losing the important phenotypic, additive, and environmental dispersal information required for genetic evaluation [ 26 ]. According to Vacek et al. [ 27 ], different measures for longevity have been reported, such as the length of productive life, total milk production, herd life, the total number of lactations, and survival observed at a certain age, measured from birth or after first calving. Longevity is determined by the voluntary and involuntary culling decisions of individual farmers. In the process of making decisions on culling, the farmers or producers will consider many traits, such as production, health, fertility, and other functional traits such as milking speed, milking temperament, and calving ease [ 28 ]. Longevity has been excluded from breeding programs because the genetic evaluation of this trait is generally difficult. In several studies, researchers have attempted to address this issue using various models that have been proposed to assess cattle longevity [ 29 , 30 ]. Gene scoring systems are not standardized by country because lifespan can be measured in many ways. Many models have been developed for the genetic assessment of lifespan. Ducrocq et al. [ 30 ] have been adopted by many researchers worldwide to analyze the survival characteristics of dairy cows. This model attempts to estimate the probability that an animal will survive to time " t ", given that it survived to time " t − 1". A second genetic assessment model for survival is the multi-trait animal model (MT). Here, a trait is defined as survival (0 or 1) to a specified endpoint or within a specified time interval during the cattle's lifetime [ 31 ]. Veerkamp et al. [ 32 ] proposed a longitudinal generalization of multiple-traits models for survival achieved through a random regression (RR) model. Binary observations (0 = culled, 1 = survived) are assigned to each discrete unit in the cow's lifetime, such as per lactation or month after first calving. Furthermore, survival is generally considered a genetically different trait in different parities, even in different stages of lactation [ 33 ]. Linear animal and sire models have been used by several researchers [ 34 , 35 ]. In practice, the performance of a model depends on the definition of the survival trait and the quality of the data [ 36 ], and the strategy of processing the merit index, combining single-trait models or multivariate models for groups of traits [ 37 ]. In Tunisia, the lifespan of Holstein cows was expressed by the number of lactations initiated in dairy cows [ 8 ]. M'Hamdi et al. [ 5 ] showed that the lifespan of Tunisian dairy cows is considered persistent, defined as the probability that a cow will live to that age if given the chance to live to that age. He used the Weibull proportional hazards model. The Weibull distribution results from a combination of the simplicity of the Weibull survival function SUR 0 ( t ) = exp (−(ht) ρ) and its flexibility. Weibull regression can model constant (ρ = 1), increasing (ρN1), and decreasing (ρb1) hazards. Further analysis is greatly facilitated if h 0 ( t ) can be approximated. The following Weibull model was used: h ( t , z ) = h 0 ( t ) ∗ e ( h y s + p l s t + m l + y b + a g e 1 + s i r e ) where: h(t, z) = hazard function of the cow at time t ; h 0 ( t ) = Weibull baseline hazard function at time t ; hys = random time-dependent effect of the herd–year–season; plst = fixed time-dependent effect of the parity–stage of lactation; ml = fixed time-dependent effect of the class of milk production expressed as deviations within the herd–year; yb = fixed time-dependent effect of annual change in herd size; a ge 1 = time-independent effect of age at the first calving; and sire = random time-independent effect of the sire of the cow. The average length of productive life was 37.5 months, and the heritability was 0.19. Ben Salem et al. [ 38 ] calculated the longevity or true herd as the number of days from the first calving to the culling or censoring; they found an average value of 48.6 ± 28.2 months. Ajili et al. [ 8 ] analyzed the true herd life (THL) variations for the last lactation of a cow. Longevity corrected for voluntary culling is called functional longevity, whereas observed longevity is called true longevity or true herd life; they found an average THL of 41.9 months. The linear model used to study the herd life was: T H L i j k l m n = μ + H Y i + S K + N L i + a g e m + e i j k l m n where: THL ijklmn is the true herd life µ: is an overall mean, HYj is the fixed effect of the herd by calving year j, Sk is the fixed effect of calving season k, NL is the fixed effect of lactation number, Age m is the fixed effect of age at first calving, and e ijklmn is a residual effect. The longevity from the Tunisian studies varies between 37.5 [ 5 ] and 48.6 months [ 38 ]. These results agree with those observed by Morek-Kopeć et al. [ 39 ] in Polish Simmental Cattle (39.3 months) and are greater than the results of Kern et al. [ 40 ] in Brazilian Holstein cows (33.5 months). The heritability is low (0.12–0.19) and in the range of the literature worldwide, as reviewed by Imbayarwo-Chikosi et al. [ 36 ], as well as the recent results reported by Ghaderi- Zefrehei et al. [ 41 ] in Iran (0.086); Wiebelitz et al. [ 42 ], (0.05) and Van Pelt and Veerkamp [ 43 ], (0.01) in Germany; and Kern et al. [ 44 ] in Brazil (0.15). 4. Genetic Evaluation of Fertility Female fertility is one of the most important factors affecting the longevity of dairy cows. Many authors have reported antagonistic genetic relationships between reproductive and milk production traits [ 45 , 46 , 47 , 48 ]. There is abundant evidence linking the selection for milk production and infertility [ 49 ]. Higher milk yields are genetically correlated with longer calving intervals, longer days to first service, and lower conception rates at first service. An evaluation of the reproductive parameters in Tunisian Holsteins was conducted on 35 dairy farms [ 50 ]. The main fertility characteristics used were calving interval (CI), days open (DO), days to first visit (DFS), and visits per conception (NSC). The fertility characteristics averaged 444 ± 101.5, 154 ± 78.4, 82 ± 56.8 days, and 2.1 ± 1.0 for CI, DO, DFS, and NSC, respectively. The average age of the cow was 6.0 ± 1.0 years of age. The heritability estimates were low: 0.03, 0.06, and 0.09 for DFS, DO, and CI, respectively. A Bayesian analysis yielded similar results [ 51 ]. The posterior mean heritabilities for CI, DO, DFS, first service to conception interval (FSC), and NSC were 0.047, 0.03, 0.025, 0.024, and 0.069, respectively. The reproducibility of the same respective functions was 0.106, 0.094, 0.051, 0.036, and 0.17 ( Table 1 ). The estimated genetic correlation between the calving intervals and DO is the highest (0.85). The lowest estimated genetic correlation (0.25) was found between DFS and NCS. Makgahlela et al. [ 54 ] reported 0.074, 0.076, 0.044, 0.058, and 0.046 for CI, DO, FSC, DFS, and NCS, respectively. The heritability estimates for the reproductive traits in the current analysis are comparable to those of M'Hamdi et al. [ 11 ]; using a multi-trait animal model, these authors reported estimates of DFS, DO, and CI of 0.032, 0.041, and 0.063, respectively. However, the results of the current heritability estimation study are less than those of Sewalem et al. [ 28 ]; performing a bivariate analysis, we found heritability estimates of 0.08 and 0.05 for DFS and FSC, respectively. Yamazaki et al. [ 48 ] used a multi-trait linear model to analyze the reproductive traits of Japanese Holstein cattle, and the heritability values for DO were 0.07 in the first lactation, 0.06 in the second lactation, and 0.12 in the third lactation. Agili et al. [ 8 ] found consecutive deliveries and intervals from delivery to first service, with release days of 427, 90, and 163 days, respectively. The data included his 128.652 records collected between 1990 and 2004 on his 47,276 Holstein Friesian cattle in 142 herds. The same authors reported positive phenotypic correlations between the fertility traits and true herd life ( Table 2 ). The estimates of the genetic parameters for female fertility traits in Tunisian dairy herds are low, even lower than those reported in the literature. In short, to improve the reproductive performance of Tunisian Holstein breeds, selection should not only focus on fertility traits, but also on improving cattle reproductive management. The genetic correlations between health and fertility are good, and the selection for mastitis or lameness is associated with shorter calving intervals, shorter intervals from calving to first service, fewer inseminations to conception, and higher rates of non-reversion. This suggests that it is likely to lead to an increase [ 1 ]. Poor health conditions such as mastitis and lameness have been shown to adversely affect fertility [ 55 ]. Early lactation illness is thought to affect the cow's ability to show fever (thereby reducing detection) and may become pregnant after insemination. Huszenicza et al. [ 55 ] found that mastitis can impair the resumption of ovarian activity in postpartum cows. B. Mastitis cows had delayed luteal activity and decreased estrous behavior after 15–28 days of milking. 5. Association of Longevity with Fertility and Type Traits 5.1. Fertility Reproductive disorders reduce fertility, prevent conception, cause problems in delivering healthy calves, cause postpartum complications, increase inter-calving intervals, reduce milk yield, and affect overall life expectancy [ 56 ]. Among the many factors that influence the age of replacement in primiparous cows, reproductive disorders are particularly important [ 57 ]. Under Tunisian conditions, the optimal primal age was 23–27 months. Reducing the age of replacement to about 24 months may improve the 305-d and longevity yields and extend the swarm longevity [ 58 ]. Medium-performing cows tended to stay in the herd longer than low-performing or high-performing cows. Shorter-than-expected residence times in prolific cows can be explained by selection for reasons other than production (involuntary selection), such as poor health and reduced fertility. These results are comparable to those by Ducrocq [ 30 ] and Weigel et al. [ 26 ]. The phenotypic correlations between the actual herd life and the milk, fat, and protein yields ranged between −0.04 (open day) and 0.06 (calving interval) on the fertility parameters [ 8 ]. Environmental factors (year of calving, season of calving) and management factors (herd) are very important sources of variability in milk production, reproductive traits, and thus herd lifespan. A combination of clearly defined reproductive goals and better management (such as selection for low production, selection of non-yield traits, and diversified feed sources) improves the performance of Tunisian Holstein cattle [ 8 ]. Functional lifespan (FL) or productive lifespan (LPL) length is an important trait for measuring the overall functional fitness in cattle. The effect of age at first calf on the milk yield and actual herd life span was investigated in Tunisian Holstein cattle. Ajili et al. [ 8 ] investigated 33.407 first lactating records for cows born between 1987 and 2001 from 166 herds and found that the herd life expectancy was 38.6 months (SD = 24 months) and the mean age at first birth was 28.7 months (SD = 3.4 months). The backward mean heritability for age at first delivery was 0.08. 5.2. Type Traits Type traits are currently being measured as part of the genetic improvement programs aimed at linking trait types with milk production, conformation, reproduction, and longevity. The purpose of including type traits is to improve the conformation of cattle by providing better body and functional and reproductive structures so that they can meet the challenges of increased production [ 3 ]. Linear traits are used to select for longevity, mainly because all trait types are acquired early in cattle life and have moderate heritability [ 59 , 60 ]. Selecting the types of traits associated with herd longevity can be beneficial in reducing involuntary selection and increasing profitability [ 61 ]. Type traits that have a significant impact on a cow's longevity are those associated with the udder, feet, legs, and leg sections, for example, anterior attachment, texture, depth, posterior mammary attachment height, posterior mammary attachment width, median ligament, and bone quality. and trunk angle [ 28 ]. The selection of the hind mammary width and height, mammary structure, mammary cleft, loin strength, bone quality, and final score may improve the longevity and milk production [ 40 ]. Zavadilová and Štípková [ 35 ] found a positive genetic correlation between longevity and type traits, BCS (0.14–0.19), tail angle (0.15–0.21), and hook quality (0.05–0.19). A slightly weaker correlation was found in the results (−0.13–0.02). Both the true and functional longevity were tested, and the type traits showed stronger genetic correlations with functional longevity. Kern et al. [ 44 ] found similar results for the type traits. The correlation between lifespan and paw angle was in the same range of −0.18 to 0.08 [ 44 ] and −0.10 to 0.10 [ 35 ]. The correlation between the type characteristic and udder depth was positive, with the longevity characteristic 0.20–0.27 (that is, the taller the breast, the longer the longevity) [ 62 ], 0.04–0.11 [ 35 ], and 0.17–0.31 [ 45 ]. Zavadilová and Štípková [ 35 ] found a negative correlation with the median ligament type trait (−0.19). [ 61 ] found a positive correlation (0.28) for the breast-supporting trait, which is considered to be the same as the median ligament, and a positive correlation (0.17–0.29) for the posterior papillary trait. Setati et al. [ 62 ] observed a low heritability of longevity and moderate heritability of most types of traits, except for lump height and pre-papilla length. All of the phenotypic correlations between lifespan and linear traits were slightly positive (0.01–0.09). The genetic correlations between longevity and breast features and angles were moderate-to-high and positive (0.22 to 0.48). In conclusion, the positive genetic correlations and moderate heritability suggest that the selection of udder features and angles could improve longevity in dairy cows [ 63 , 64 ]. 5.1. Fertility Reproductive disorders reduce fertility, prevent conception, cause problems in delivering healthy calves, cause postpartum complications, increase inter-calving intervals, reduce milk yield, and affect overall life expectancy [ 56 ]. Among the many factors that influence the age of replacement in primiparous cows, reproductive disorders are particularly important [ 57 ]. Under Tunisian conditions, the optimal primal age was 23–27 months. Reducing the age of replacement to about 24 months may improve the 305-d and longevity yields and extend the swarm longevity [ 58 ]. Medium-performing cows tended to stay in the herd longer than low-performing or high-performing cows. Shorter-than-expected residence times in prolific cows can be explained by selection for reasons other than production (involuntary selection), such as poor health and reduced fertility. These results are comparable to those by Ducrocq [ 30 ] and Weigel et al. [ 26 ]. The phenotypic correlations between the actual herd life and the milk, fat, and protein yields ranged between −0.04 (open day) and 0.06 (calving interval) on the fertility parameters [ 8 ]. Environmental factors (year of calving, season of calving) and management factors (herd) are very important sources of variability in milk production, reproductive traits, and thus herd lifespan. A combination of clearly defined reproductive goals and better management (such as selection for low production, selection of non-yield traits, and diversified feed sources) improves the performance of Tunisian Holstein cattle [ 8 ]. Functional lifespan (FL) or productive lifespan (LPL) length is an important trait for measuring the overall functional fitness in cattle. The effect of age at first calf on the milk yield and actual herd life span was investigated in Tunisian Holstein cattle. Ajili et al. [ 8 ] investigated 33.407 first lactating records for cows born between 1987 and 2001 from 166 herds and found that the herd life expectancy was 38.6 months (SD = 24 months) and the mean age at first birth was 28.7 months (SD = 3.4 months). The backward mean heritability for age at first delivery was 0.08. 5.2. Type Traits Type traits are currently being measured as part of the genetic improvement programs aimed at linking trait types with milk production, conformation, reproduction, and longevity. The purpose of including type traits is to improve the conformation of cattle by providing better body and functional and reproductive structures so that they can meet the challenges of increased production [ 3 ]. Linear traits are used to select for longevity, mainly because all trait types are acquired early in cattle life and have moderate heritability [ 59 , 60 ]. Selecting the types of traits associated with herd longevity can be beneficial in reducing involuntary selection and increasing profitability [ 61 ]. Type traits that have a significant impact on a cow's longevity are those associated with the udder, feet, legs, and leg sections, for example, anterior attachment, texture, depth, posterior mammary attachment height, posterior mammary attachment width, median ligament, and bone quality. and trunk angle [ 28 ]. The selection of the hind mammary width and height, mammary structure, mammary cleft, loin strength, bone quality, and final score may improve the longevity and milk production [ 40 ]. Zavadilová and Štípková [ 35 ] found a positive genetic correlation between longevity and type traits, BCS (0.14–0.19), tail angle (0.15–0.21), and hook quality (0.05–0.19). A slightly weaker correlation was found in the results (−0.13–0.02). Both the true and functional longevity were tested, and the type traits showed stronger genetic correlations with functional longevity. Kern et al. [ 44 ] found similar results for the type traits. The correlation between lifespan and paw angle was in the same range of −0.18 to 0.08 [ 44 ] and −0.10 to 0.10 [ 35 ]. The correlation between the type characteristic and udder depth was positive, with the longevity characteristic 0.20–0.27 (that is, the taller the breast, the longer the longevity) [ 62 ], 0.04–0.11 [ 35 ], and 0.17–0.31 [ 45 ]. Zavadilová and Štípková [ 35 ] found a negative correlation with the median ligament type trait (−0.19). [ 61 ] found a positive correlation (0.28) for the breast-supporting trait, which is considered to be the same as the median ligament, and a positive correlation (0.17–0.29) for the posterior papillary trait. Setati et al. [ 62 ] observed a low heritability of longevity and moderate heritability of most types of traits, except for lump height and pre-papilla length. All of the phenotypic correlations between lifespan and linear traits were slightly positive (0.01–0.09). The genetic correlations between longevity and breast features and angles were moderate-to-high and positive (0.22 to 0.48). In conclusion, the positive genetic correlations and moderate heritability suggest that the selection of udder features and angles could improve longevity in dairy cows [ 63 , 64 ]. 6. Genetic Evaluation Enhanced with Genomic During the last century, dairy cattle breeding schemes were based on the phenotypic data processed according to quantitative genetic rules. Evaluation accuracy requires heavy and expensive phenotype recording and the progeny testing of bulls by observing the performances of their daughters. Indeed, the application of quantitative genetics to dairy cattle breeding between 1960 and 2007 was very successful, increasing the milk yield and profitability of the production systems [ 65 ]. In Holstein dairy cattle, milk production still increases by 110 kg per animal per year [ 66 ]. This process involved thousands of cows and took many years to complete. However, the use of molecular information to make selection decisions in breeding schemes was envisaged decades ago. In the 1990s, DNA analysis marked the beginning of the new field of genomics [ 67 ]. First, a genetic marker approach based on the concept of the detection linkage relationships between genomic regions and quantitative traits-quantitative trait loci (QTLs)was utilized in the marker-assisted selection (MAS) process. The usual MAS idea is a three-step process: (i) detect one or more QTLs; (ii) find the causative gene (causative mutation); and (iii) increase the frequency of favorable alleles by selection or introgression. Except for a few cases (halothane in pigs, Barolo in sheep), the impact of MAS in livestock breeding programs was minimal (10%), as a QTL above the selected significance level can usually explain only a fraction of the trait variance [ 67 ]. In the context of multi-trait breeding goals, such markers may have a less overall impact on the breeding goals, as stronger responses to one trait often appear at the expense of another trait. The MAS approach was successful for traits with simple genetic determinism but had disappointing results in more complex situations [ 68 ]. Second, the recent availability and affordability of high-density panels of single nucleotide polymorphism (SNP) markers have opened up this new possibility. Mészáro et al. [ 69 ] proposed what is now known as genomic selection. In general, 45,000 SNPs are used in bulls and 3000 SNPs are used in genotyping cattle, heifers, and calves. Genomic selection has revolutionized dairy farming in the last decade and is considered a success story [ 68 ]. High-density SNP genotyping is used in two ways: Genome-Wide Associations studies (GWSA) to identify genetic markers (SNPs) or genomic regions (QTLs) associated with traits. This technology has been largely used to detect the genetic associations with farm animal traits, mainly with low heritability, to achieve faster genetic progress. Hundreds of markers were identified to be associated with production, functional, and novel traits in dairy cattle. Implementing a GWA study on the Holstein breed, Nayeri et al. [ 70 ] identified interesting markers associated with longevity ( SYT10 on chromosome 5, ADAMTS3 on chromosome 6, NTRK2 on chromosome 8, and DERL1 and SNTG1 on chromosome 14). Tiezzi et al. [ 71 ] concluded that the most significant SNP associated with longevity traits were found on Bos taurus autosome 18 ( BTA18 ) and were mostly located within the QTL regions associated with mastitis. In contrast, for the lifetime profit index, the strongest associations were detected on BTA14 and BTA18 . They also observed an association between BTA6 and BTA20 for lactation persistency. For clinical mastitis, which is related to longevity, Bermotiene [ 72 ] observed in Holstein's first lactation an association between genetic variation and the regions on chromosomes 2, 14, 20, and 29. In Tunisia, no QTL mapping experiment has ever been conducted on dairy cattle. Recently, a study aimed to analyze the polymorphism of the gene responsible for the biosynthesis of leptin using the polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) technique. Indeed, leptin is a glycoprotein that is involved in the defense mechanisms of the mammary gland of dairy cows. The level of this protein secretion rises significantly in response to viral or bacterial infections [ 73 ]. A total of 160 blood samples were collected from dairy Holstein breed cows. The results showed the presence of two alleles A and B and three genotypes AA, AB, and BB, with a dominance of the allele A. The results indicate that animals carrying the BB gene could contribute to a reduction in somatic cells in cattle. Based on this observation, they are more resistant to mastitis. This peptide can be considered a candidate gene for udder health [ 74 ]. The whole-Genome selection presents numerous advantages versus the traditional quantitative approach: (i) genomic information is not affected by environment; (ii) it can be available at an early age; (iii) it can be obtained on all selected candidates; (iv) it enhances the reliability in predicting the mature phenotype; and (v) the genomic approach helps us to select for a wide range of traits and, in turn, saves time and effort [ 75 ]. The 'development of inexpensive high-throughput genotyping platforms of SNP markers revolutionized dairy cattle breeding" [ 76 ]. In the last decade, genomic selection has been very successful and was quickly adopted in the largest populations worldwide; genetic trends have doubled, especially for traits with low heritability. Several countries have already adopted genomic selection in the dairy cattle industry and an international genomic evaluation model across countries was implemented (Genomic MACE). Nine years after the adoption of genomic selection in the United States of America (US), the results were very successful. A dramatic reduction in generation interval (GI) was observed. The GI of the Sire of offspring decreased from seven years to less than 2.5 years, and the GI of the fathers of daughters decreased from about four years to almost 2.5 years. The difference is relatively stable. The most dramatic responses to genomic selection were observed for the inherited traits of low daughter pregnancy rate (DPR), productive life (PL), and SCS. Genetic propensities have gone from near-zero to large and favorable, resulting in rapid genetic improvements in fertility, longevity, and health; however, these traits have eroded over time. These results demonstrate a positive effect of genomic selection in US dairy cows, although this technique has only recently been used. Based on a four-way selection model, the genetic gains per year ranged between approximately 50–100% for yield-related traits and 3- to 4-fold increases for low-heritability traits [ 76 ]. Similar results were observed in other countries and dairy cattle populations. The unique genomic research program initiated on cattle in Tunisia was implemented on the phylogenic insights into the history of Tunisian native cattle. Furthermore, Ben Jemaa et al. [ 76 ], in a genome-wide characterization of local cattle breeds from the central and western Mediterranean, pinpointed the admixed origin of the genome of the Tunisian native cattle population with the two main European and African ancestries. However, Ben Jemaa [ 76 ] suggested the implementation of a genomic selection program for Tunisian dairy cattle. 7. Conclusions The main conclusions of this review are: The genetic parameters observed for the Tunisian Holstein dairy cattle are in agreement with the literature used in the present manuscript. During the last few decades of the twentieth century, the application of quantitative genetics to dairy cattle breeding was very successful, increasing the milk yield and profitability of the production systems. At present, herd management focuses on the balance of functional traits (particularly health, fertility, longevity, and welfare) against production. The performance of an evaluation model depends on the definition of the trait, the quality of the data, and the strategy of processing the merit index, combining single-trait models or multivariate models for groups of traits. However, the revolution in genotyping provided by high-density SNP coupled with a cost reduction has resulted in large databases of individuals with genome-wide genotypic data. Moreover, in the last decade, genomic selection has been very successful and was quickly adopted in the largest populations worldwide; genetic trends have doubled, especially for traits with low heritability. In Tunisia, Genomic research is sporadic, with no implementation of a genomic selection program to date. Despite the success of several countries in increasing the annual genetic gain of different traits of economic interest, we recommend to Tunisia, as it is a member of the ICAR consortia, to reinforce the phenotypic recording system and extend it to functional parameters, mainly longevity, and initiate a genomic selection program. The implementation and validation of the genetic evaluation process, in the INTERBULL framework, would enable Tunisia and developing countries in a similar situation to overcome the gap caused by genomic selection.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2784286/

Bacillus anthracis Physiology and Genetics

Bacillus anthracis is a member of the Bacillus cereus group species (also known as the "group 1 bacilli"), a collection of Gram-positive spore-forming soil bacteria that are non-fastidious facultative anaerobes with very similar growth characteristics and natural genetic exchange systems. Despite their close physiology and genetics, the Bacillus cereus group species exhibit certain species-specific phenotypes, some of which are related to pathogenicity. B. anthracis is the etiologic agent of anthrax. Vegetative cells of B. anthracis produce anthrax toxin proteins and a poly-D-glutamic acid capsule during infection of mammalian hosts and when cultured in conditions considered to mimic the host environment. The genes associated with toxin and capsule synthesis are located on the B. anthracis plasmids, pXO1 and pXO2, respectively. Although plasmid content is considered a defining feature of the species, pXO1- and pXO2-like plasmids have been identified in strains that more closely resemble other members of the B. cereus group. The developmental nature of B. anthracis and its pathogenic (mammalian host) and environmental (soil) lifestyles of make it an interesting model for study of niche-specific bacterial gene expression and physiology.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7858201/

Importance of livestock diseases identified using participatory epidemiology in the highlands of Ethiopia

Livestock are a major cornerstone for livelihoods of smallholders in the highlands of Ethiopia. However, infectious diseases are a key constraint in livestock production affecting productivity and food security. This study aimed to understand livestock producers' perception on importance and epidemiology of livestock diseases. Participatory tools using focus group discussions were used to identify and rank livestock species, major livestock diseases, and their epidemiological patterns in smallholder systems. The study covered 17 districts in four regions, with two or three kebeles (smallest administrative unit) included in each district. To capture views of women and men, separate focus group discussions (FGD) were conducted. Data collected through proportional piling were used in generalized linear model analysis. Taking sheep as a reference, cattle were significantly the most preferred species with the odds of cattle scoring higher than sheep being 2.08 times ( P = 0.000). Sheep and poultry were equally the second most important livestock, followed by goat and equine. There were no statistically significant differences between men and women FGD groups and between the four regions in their preference for livestock species. Twelve out of 28 identified livestock diseases were mentioned by at least 10% of FGDs in their list of five most important diseases. Blackleg, foot, and mouth disease, lumpy skin disease, anthrax, and bovine pasteurellosis were ranked as the top five diseases by both men and women focus groups. Reasons for high scores included suddenness of death, zoonosis risks, animal age groups and species affected, frequency of occurrence, rate of transmission in herds, curability, and inefficient vaccination. Seasonality and agro-ecology were also factors associated with disease occurrence. The study also documents that adult men and women are more involved in activities related to disease transmission compared with other family members. When asked about impacts of diseases, loss of livelihood, displacement, and infection with zoonoses were mentioned, with women considered equally affected as men. In conclusion, participatory tools allowed understanding disease priorities of and their perceived epidemiology in smallholder systems. The perceptions and priorities of men and women are very similar, and both need to be involved in designing health management interventions. Based on the findings of this study, the intervention strategies listed in the 2013 animal health strategy of Ethiopia (MoA and ILRI 2013 ) seem pertinent and need to be implemented to overcome the challenges of diseases. Introduction Rural Ethiopia is home for 87% of the country's 100 million people. The smallholder farmers in the highland mixed livestock-crop production system rely heavily on livestock for their livelihoods, livestock serving as source of cash income for incidental/daily expenses (poultry, egg and milk sale), big expenses (fattened cattle and sheep/goat), quality protein supplementing cereal-based diet, and power for crop farming. The country is endowed with diverse animal genetic resources and agroecology. However, the contribution of the livestock sector to household livelihoods and the national economy is well below potential (e.g., Aklilu et al. 2013 ). Livestock provides only 16% of the total GDP (equivalent to 30% of agricultural GDP) and generates 14% of the country's foreign exchange earnings (CSA 2009 ). Ethiopia is endemic to a number of livestock diseases, including OIE-listed diseases such as contagious bovine pleuropneumonia (CBPP), lumpy skin disease (LSD), foot and mouth disease (FMD), Newcastle disease (NCD), Peste des petits ruminants (PPR), sheep and goat pox (SGP), and African horse sickness (AHS) (NEPAD 2005 ). Besides affecting the production of livestock products, livestock diseases have other impacts in the highland mixed crop-livestock system, namely their impact on work performance of oxen which could lead to severe food insecurity and poverty. Mortality rates of up to 49% in oxen, 50% in local cows, 81% in crossbred cows, 52% in local calves/heifers, and 60% in crossbred calves/heifers have been reported in four regional regions in Ethiopia (Genbremedhin et al. 2017 ). Livestock diseases reduce income and affect livelihoods of livestock keepers and jeopardize food security at local and national level. Despite the substantial export demand and the countries potential, the presence and prevalence of a number of trade-limiting transboundary livestock diseases has denied the country access to international market and makes it vulnerable to trade bans. A 2005 report (NEPAD 2005 ) estimated annual loss of about Birr 1.5–2.5 billion from the export market due to animal diseases. Diseases cause huge economic loss to producers and livestock exporters. A study of five cattle exporting enterprises with 4321 bulls kept for export at quarantines found infection rates of 12.9% for FMD and 8.0% for CBPP and estimated financial losses of about USD 241, 2341 due to the diseases between November 2013 and May 2014 (Birhanu 2014 ). Animal diseases also have an important impact on human health, with 60% of human infectious diseases being of animal origin. Ethiopia is endemic to a wide range of such zoonotic diseases (MoA and ILRI 2013 ). The disease problem is rooted, among other causes, in limited understanding of the epidemiology of diseases in the various geographic regions, agroecological zones, and seasons. Lack of epidemiological information hampers development of disease control strategies. Epidemiological studies using conventional veterinary science methods have been limited due to the limitations in laboratory and logistic resources. A participatory approach, participatory epidemiology (PE), has been proven a suitable approach in such situations and has helped to improve understanding of diseases, and options for disease control through the systematic use of participatory methods (Catley et al. 2012 ). The approach is particularly valuable in the conditions of developing countries where smallholders' production objectives and management practices vary and in areas where diverse agroecological and seasonal variations exist and enables to consider smallholders priorities and their indigenous knowledge in designing health interventions. The objective of this study was to understand livestock disease patterns in four highland regions of Ethiopia and the roles and indigenous knowledge of different household members including women and men household members using participatory epidemiology approach. The results of this study would be used to design effective disease control interventions across the different agro-ecologies and identify the roles of women and men farmers in implementing interventions. Materials and methods Sampling and data collection The study was conducted in four regions Ethiopia, namely Amhara, Oromia, SNNPR (Southern Nations and Nationalities and Peoples region), and Tigray. In 17 districts, 37 kebeles (the smallest administrative unit) were selected; details on the agroecological characteristics are shown in Table 1 . Two focus group discussions (FGD), each consisting of six to eight men and six to eight women members, were held in each kebele . However, men and women FGDs were not held in two and three of the kebeles , respectively, and two women FGDs were held in one of the kebeles , resulting in a total of 70 FGDs. Table 1 Sampling structure for focus group discussions Region No. of districts No. of kebeles Agro-ecological zones Altitude (m) Rainfall (mm) Temperature (°C) Amhara 2 4 Moist highland 2636 1137 14.8 2 4 Moist mid-highland 1912 1081 18.9 2 5 Wet mid-highland 2023 1486 17.6 Oromia 1 3 Dry mid-highland 1761 837 19.5 2 4 Moist highland 2583 1137 14.2 2 4 Wet mid-highland 1892 1759 18.4 SNNPR 1 2 Moist highland 2588 1276 14 1 3 Moist mid-highland 1948 1213 18 Tigray 1 2 Dry highland 2436 581 16.3 3 6 Dry mid-highland 2036 612 18.6 Overall 17 37 – – – – Participatory tools were used to identify priority livestock species and diseases and their epidemiological patterns. Beans were used for proportional piling to elicit farmers' priorities and perceptions. For this, participants were asked to allocate 100 beans according to preferences for alternative outcomes (livestock species and diseases). The FGDs were first asked to discuss and identify the five most important livestock species by allocating 100 beans to the five species. Then five most important livestock diseases were identified similarly and the local/traditional name of each disease or syndrome, and its clinical signs were described by the FGD groups, which were used by the local veterinarians who facilitated the FGDs to translate the diseases/disease syndromes into their equivalent scientific names. The seasonal distribution of each of the five diseases and the livestock age groups affected by the diseases were then discussed by dividing the scores for each disease into four seasons according to the importance of the disease in the four seasons. Effects of diseases on the household in general and household members individually and activities and household members involved in disease transmission were discussed and described, and contribution of each family member was scored out of 20 beans. Disease coping mechanisms were also discussed. Data analysis A mixed-methods approach to data analysis was used. Qualitative information based on notes taken on the discussions during the FGDs was coded into categories, for example to define and group different diseases or reasons to prioritize diseases and livestock species. For quantitative information, descriptive statistics including proportions of FGD groups' scorings for the different species and diseases were disaggregated by gender of the FGD groups, regions, agroecologies, and seasons. Statistical inferential analyses were employed to test the significance of farmers' relative preferences to the different livestock species and their perceptions on the relative importance of diseases and to demonstrate the applicability of results beyond the sample respondents. Farmers' preferences and perceptions (considered as the dependent variables) were measured as the number of beans allocated to each of the categories of the independent variables, namely the five species, diseases, four agroecologies, and seasons. The analyses were conducted using a generalized linear model procedure in SPSS version 20 (2011), fitting a natural logarithmic transformation of the data to meet the normal distribution assumption for linear model analysis. The P-P plot test showed the data distribution was normalized after transformation (Fig. 1 ). The odds ratio model as suggested by Abeyasekera ( 2001 ) and Agresti ( 1996 ) was fitted to compare the farmers' preferences for the different species in reference to sheep. The interpretation of results was based on the odds of farmers preferring each species compared with sheep. The choice of sheep as a reference category was due to its high economic importance in the highlands of Ethiopia as a major source of income. Chi-squared analyses were also used for pairwise comparison of variations across regions, gender of FGD groups, agroecologies, and seasons as appropriate. Fig. 1 Test for normality of FGD groups' proportional piling scores using P-P plots before ( a ) and after ( b ) natural logarithm transformation of the data Sampling and data collection The study was conducted in four regions Ethiopia, namely Amhara, Oromia, SNNPR (Southern Nations and Nationalities and Peoples region), and Tigray. In 17 districts, 37 kebeles (the smallest administrative unit) were selected; details on the agroecological characteristics are shown in Table 1 . Two focus group discussions (FGD), each consisting of six to eight men and six to eight women members, were held in each kebele . However, men and women FGDs were not held in two and three of the kebeles , respectively, and two women FGDs were held in one of the kebeles , resulting in a total of 70 FGDs. Table 1 Sampling structure for focus group discussions Region No. of districts No. of kebeles Agro-ecological zones Altitude (m) Rainfall (mm) Temperature (°C) Amhara 2 4 Moist highland 2636 1137 14.8 2 4 Moist mid-highland 1912 1081 18.9 2 5 Wet mid-highland 2023 1486 17.6 Oromia 1 3 Dry mid-highland 1761 837 19.5 2 4 Moist highland 2583 1137 14.2 2 4 Wet mid-highland 1892 1759 18.4 SNNPR 1 2 Moist highland 2588 1276 14 1 3 Moist mid-highland 1948 1213 18 Tigray 1 2 Dry highland 2436 581 16.3 3 6 Dry mid-highland 2036 612 18.6 Overall 17 37 – – – – Participatory tools were used to identify priority livestock species and diseases and their epidemiological patterns. Beans were used for proportional piling to elicit farmers' priorities and perceptions. For this, participants were asked to allocate 100 beans according to preferences for alternative outcomes (livestock species and diseases). The FGDs were first asked to discuss and identify the five most important livestock species by allocating 100 beans to the five species. Then five most important livestock diseases were identified similarly and the local/traditional name of each disease or syndrome, and its clinical signs were described by the FGD groups, which were used by the local veterinarians who facilitated the FGDs to translate the diseases/disease syndromes into their equivalent scientific names. The seasonal distribution of each of the five diseases and the livestock age groups affected by the diseases were then discussed by dividing the scores for each disease into four seasons according to the importance of the disease in the four seasons. Effects of diseases on the household in general and household members individually and activities and household members involved in disease transmission were discussed and described, and contribution of each family member was scored out of 20 beans. Disease coping mechanisms were also discussed. Data analysis A mixed-methods approach to data analysis was used. Qualitative information based on notes taken on the discussions during the FGDs was coded into categories, for example to define and group different diseases or reasons to prioritize diseases and livestock species. For quantitative information, descriptive statistics including proportions of FGD groups' scorings for the different species and diseases were disaggregated by gender of the FGD groups, regions, agroecologies, and seasons. Statistical inferential analyses were employed to test the significance of farmers' relative preferences to the different livestock species and their perceptions on the relative importance of diseases and to demonstrate the applicability of results beyond the sample respondents. Farmers' preferences and perceptions (considered as the dependent variables) were measured as the number of beans allocated to each of the categories of the independent variables, namely the five species, diseases, four agroecologies, and seasons. The analyses were conducted using a generalized linear model procedure in SPSS version 20 (2011), fitting a natural logarithmic transformation of the data to meet the normal distribution assumption for linear model analysis. The P-P plot test showed the data distribution was normalized after transformation (Fig. 1 ). The odds ratio model as suggested by Abeyasekera ( 2001 ) and Agresti ( 1996 ) was fitted to compare the farmers' preferences for the different species in reference to sheep. The interpretation of results was based on the odds of farmers preferring each species compared with sheep. The choice of sheep as a reference category was due to its high economic importance in the highlands of Ethiopia as a major source of income. Chi-squared analyses were also used for pairwise comparison of variations across regions, gender of FGD groups, agroecologies, and seasons as appropriate. Fig. 1 Test for normality of FGD groups' proportional piling scores using P-P plots before ( a ) and after ( b ) natural logarithm transformation of the data Results and discussion Priority of livestock species The majority of the focus group discussion (FGD) groups ranked cattle as their most important livestock species. The average scores allocated by men and women FGDs were 38.0 and 40.5 for cattle, 18.6 and 19.4 for sheep, 15.5 and 16.3 for poultry, 14.9 and 15.6 for goats, and 14.3 and 13.4 for equines. The scores in Amhara, Oromia, SNNPR, and Tigray regions were 41.5, 38.3, 39.3, and 37.4 for cattle, 18.5, 18.3, 19.6, and 21.4 for sheep, 12.6, 18.7, 16.6, and 16.8 for poultry, 12.9, 12.8, 16.1, and 19.4 for goat and 15.9, 15.0, 13.9, and 9.6 for equine. The generalized linear model analysis (Table 2 ) indicated cattle were significantly the most preferred species with the odds of cattle scoring higher than sheep being 2.08 times ( P = 0.000). Sheep and poultry were equally the second most important livestock, followed by goat and equine. There were no statistically significant differences between men and women FGD groups in their preference for livestock species (Table 2 ). There was also no significant variation in the preferences of farmers for cattle, sheep, and poultry in the four regions studied. However, goats received significantly higher scores in Tigray region compared to Oromia (odds ratio = 0.66, P = 0.009) and Amhara (odds ratio = 0.67, P = 0.005) resulting in goats being more important than poultry in Tigray. Similarly, compared with Tigray, equines were 1.65 ( P = 0.008) and 1.55 ( P = 0.024) times more likely to receive higher scores in Amhara and Oromia, respectively, where they are more important than goats. Table 2 The odds (Exp(B)) of men and women FGD groups allocating higher scores to the different livestock species in reference to sheep in Ethiopia B Std. error Sig. Exp(B) Intercept 2.966 0.0655 0.000 19.414 Cattle 0.736 0.0715 0.000 2.088 Equine − 0.370 0.1116 0.001 0.691 Goat − 0.217 0.1068 0.042 0.805 Poultry − 0.175 0.0974 0.072 0.839 Sheep 0 a 1 Cattle—Men FGD − 0.064 0.0417 0.125 0.938 Cattle—Women FGD 0 a 1 Equine—Men FGD 0.062 0.1214 0.612 1.063 Equine—Women FGD 0 a 1 Goat—Men FGD − 0.048 0.1200 0.689 0.953 Goat—Women FGD 0 a 1 Poultry—Men FGD − 0.049 0.1037 0.636 0.952 Poultry—Women FGD 0 a 1 Sheep—Men FGD −0.045 0.0948 0.632 0.956 Sheep—Women FGD 0 a 1 a Reference category = sheep; women FGD The reasons provided by respondents for ranking cattle as the most important species reflected the production objectives of farmers in mixed crop-livestock systems, the most important reasons being draught power/income from fattened cattle and milk production, mentioned by 82.9% and 84.3% of the respondents, respectively. There was no difference in the reasons provided by men and women FGD groups and across the four regions, except that significantly smaller percentage of FGD groups cited milk (52.9%) and income from fattening (47.1%) in Tigray and drought power (40%) in SNNPR. Manure was also mentioned as important function of cattle by 34.3% of women and 37.1% of male FGDs, and by 43.5%, 40.0% and 41.2% of the respondents in Amhara, Oromia, and Tigray regions, respectively. Unlike in the pastoral system, socio-cultural values, and asset/saving functions of cattle were less important functions of cattle in the highland mixed crop-livestock system. Priority of diseases Twenty-eight diseases or disease syndromes were identified by the FGD groups. Reflecting the importance of cattle, most of the diseases identified were cattle diseases, though some of them also affect small ruminants. Two poultry-specific (New castle disease and salmonellosis, though salmonellosis also affects ruminants) and one equine-specific disease (African horse sickness) were also identified. However, only 12 diseases were mentioned by at least 10% of the FGD groups (Fig. 2 ). The 12 diseases, which were considered for further analysis, were blackleg, food and mouth disease (FMD), lumpy skin disease (LSD), anthrax, bovine pasteurellosis, mastitis, endoparasitosis, ectoparasitosis, fasciolosis, rabies, foot rot, and trypanosomiasis. Fig. 2 Frequency of FGD groups who mentioned 28 diseases among their top five diseases in four regions of Ethiopia. (CBPP = contagious bovine pleuropneumonia, CCPP = contagious caprine pleuropneumonia, FMD = foot and mouth disease, LSD = lumpy skin disease, NCD = New castle disease) The top five diseases based on their average scores were blackleg (18.9), FMD (13.8), LSD (12.5), anthrax (10.6), and bovine pasteurellosis (8.8). Statistical analysis (Table 3 ) also confirmed these were the five priority diseases. Blackleg was significantly the most important disease. Farmers are more likely to allocate higher score to blackleg compared with the reference LSD (odds ratio = 1.42; P = 0.021). However, farmers would allocate statistically similar scores for anthrax, FMD, bovine pasteurellosis, and LSD (Table 3 ). Table 3 The odds of farmers allocating higher scores for diseases in reference to LSD (lumpy skin disease) and men allocating higher scores in reference to women the highlands of Ethiopia Β Std. error Sig. ( P ) Odds ratio Intercept 2.593 0.1240 0.000 13.371 Anthrax − 0.219 0.1981 0.269 0.803 Blackleg 0.351 0.1517 0.021 1.421 Bovine pasteurellosis − 0.365 0.2175 0.094 0.694 Ectoparasite − 1.067 0.3813 0.005 0.344 Endoparasite − 1.186 0.4245 0.005 0.306 Fasciolosis − 1.439 0.5375 0.007 0.237 FMD − 0.155 0.1907 0.418 0.857 Foot-rot − 2.105 1.0260 0.040 0.122 Mastitis − 1.158 0.4140 0.005 0.314 Ovine pasteurellosis − 2.511 1.5328 0.101 0.081 Rabies − 2.071 0.9918 0.037 0.126 Trypanosomiasis − 2.105 1.0260 0.040 0.122 LSD 0 a 1 Men FGD scoring b Anthrax 0.452 0.1830 0.013 1.572 Blackleg − 0.009 0.1240 0.942 0.991 Bovine pasteurellosis − 0.107 0.2673 0.689 0.898 Ectoparasite − 0.507 0.6987 0.468 0.602 Endoparasite − 0.058 0.5914 0.922 0.944 Fasciolosis 0.018 0.7331 0.981 1.018 FMD − 0.156 0.2227 0.483 0.855 Foot-rot − 0.214 1.6218 0.895 0.807 Mastitis 0.047 0.5460 0.932 1.048 Ovine pasteurellosis 0.234 1.9486 0.905 1.263 Rabies − 0.271 1.6225 0.867 0.763 Trypanosomiasis 0.728 1.1311 0.520 2.070 LSD − 0.147 0.1898 0.439 0.863 a Lumpy skin disease (LSD) is the reference category. FGD focus group discussion groups b The reference category: Men FGDs disease scoring were compared with women FGDs scoring for each disease Men and women respondents had similar perceptions of diseases, their priority diseases being statistically similar except men allocated significantly higher scores to the importance of anthrax in reference to LSD (Table 3 ). Women and men groups' scores were 19 and 18.3 for blackleg, 11.46 and 9.8 for FMD, 13.4 and 11.5 for LSD, 10.7 and 16.9 for anthrax, and 9.29 and 8.34 for bovine pasteurellosis. There were slight variations across the four regions in disease priorities. Among the most frequently mentioned diseases, those that were common across the regions were blackleg, FMD, LSD, anthrax, bovine pasteurellosis, and ectoparasitosis. Mastitis was mentioned in Oromia only and cowdriosis (heartwater) and Newcastle disease (NCD) in SNPPR only. In reference to the list of five top diseases across regions (Table 3 ), the list in Amhara and Oromia excluded bovine Pasteurellosis and included trypanosomiasis and mastitis, respectively. In SNNPR, the list included mastitis and fasciolosis and excluded FMD and anthrax. The top five in Tigray excluded LSD and ectoparasites. The FGD groups considered a range of criteria to rank diseases (Table 4 ). The reasons for anthrax were the sudden death it inflicts and its potential effect on human health as it is a zoonotic disease, reasoned by 81.2% and 50.0% of the FGDs. Although not supported by the existing modern veterinary knowledge, 7.1% of the respondents considered LSD as a zoonotic disease. Blackleg was perceived to infect animals with good body condition such as well fattened animals. LSD was associated with high morbidity rate resulting in reduced milk production and work performance of oxen by almost all respondents and high transmission rate by about a third of the respondents who ranked the disease. Table 4 Percentage of farmers providing reasons for top ranking diseases in the highlands of Ethiopia Reasons for ranking Anthrax Blackleg Bovine pasteurellosis FMD LSD Zoonotic disease 50.0 6.2 0.0 0.0 7.1 Affects all age groups/species 43.8 6.2 0.0 75.0 14.3 Affects animals in good condition 6.2 56.2 100.0 75.0 0.0 Sudden death 81.2 37.5 66.7 0.0 7.1 High morbidity/effect on productivity 0.0 37.5 0.0 25.0 92.9 High mortality 43.8 18.8 0.0 25.0 7.1 High transmission rate 0.0 12.5 0.0 25.0 35.7 Affects skin quality 0.0 0.0 0.0 0.0 21.4 Not common in locality 0.0 6.2 0.0 0.0 0.0 Agroecological and seasonal distribution of diseases Agroecology is the major determinant of disease epidemiology. Based on a combined criterion of the proportion of respondents ranking the diseases 1st to 3rd and those not including the disease in their top five list, Anthrax, FMD, and bovine pasteurellosis were the most important diseases in the dry highlands (Table 5 ). LSD and blackleg in the moist lowlands and blackleg and LSD in the wet highlands were the most important diseases. Table 5 Per cent of respondents ranking the top five diseases in different agro-ecological zones of Ethiopia Agroecology 1 Rank 2 Anthrax Blackleg Bovine pasteurellosis FMD LSD Dry highland 0 3 38.1 a 4 38.1 a 28.6 a 19.0 a 90.5 b 1 38.1 a 14.3 a, b 14.3 a, b 19.0 a, b – 2 19.0 a, b 9.5 a, b 42.9 b 9.5 a, b – 3 4.8 a 23.8 a 4.8 a 23.8 a – Moist highland 0 46.9 a, b 12.5 c 78.1 b 28.1 a, c 25.0 a, c 1 21.9 a, b 18.8 a, b – – 34.4 a 2 28.1 a 31.2 a 12.5 a, b – 9.4 a, b 3 – 15.6 a, b 9.4 a, b 28.1 b 9.4 a, b Wet highland 0 56.2 a, b 31.2 a, b 62.5 a, b 75.0 b 18.8 a 1 6.2 a, b 43.8 b – – 18.8 a, b 2 – 25.0 a – – 18.8 a 3 12.5 a – – 6.2 a – 1 Result for moist lowland zone is not presented. None of the respondents included anthrax and LSD in the top five diseases list, whereas all respondents in this zone ranked Blackleg, pasteurellosis, and FMD as the 2nd, 4th, and 5th most important diseases, respectively 2 Ranks derived from the number of beans (scores) allocated to the diseases 3 The disease was not included in farmers' lists of top five diseases 4 Different superscript letters within row denotes percent of respondents ranking the diseases differ significantly from each other at the 0.05 level Seasons of the year in the highlands of Ethiopia are generally divided into four, although there are some variations across agroecological zones and geographic locations (regions). Generally, the four seasons are the dry/cold (September to November, dry/cold), the dry/hot (December to February, dry/hot), the short rainy (March to May, short rainy season), and the long rainy (June to August, long rainy season) seasons. Interviewed farmers observed a higher prevalence of anthrax in dry/cold and short rainy seasons, blackleg in season dry/cold, short rainy seasons, and long rainy seasons, pasteurellosis in season dry/hot and short rainy seasons, FMD in dry/hot and short rainy seasons, and LSD in dry/cold season and long rainy seasons (Fig. 3 ). Fig. 3 Average scorings of farmers for the prevalence level of five top diseases in four seasons of the year in the highlands of Ethiopia Transmission and impacts of diseases Women and men respondents' understanding of disease transmission pathways was similar. Adults were more responsible than the rest of the family members for the transmission of diseases. Adult men played more role in the transmission of non-zoonotic diseases, but both men and women were equally responsible for zoonotic diseases (Table 6 ). Adult men scored more for transmission activities involving animal movement, disposal of dead animals, and handling of contaminated tools, whereas women scored more in activities related to managing sick animals. However, both men and women are equally responsible for the transmission of zoonotic diseases. Table 6 Respondents' scorings (out of 20) of the roles of family members in the transmission of diseases in the highlands of Ethiopia Transmission activity Adult men Adult women Young males Young females Children Non-zoonotic diseases Moving animals to distant places 11.0 0.0 9.0 0.0 0.0 Burying dead animals 10.0 4.0 4.7 1.3 0.0 Handling contaminated tool 14.0 6.0 0.0 0.0 4.0 Eating/drinking raw meat/milk 9.0 2.3 2.3 2.3 4.0 Eating without washing hands 4.0 6.7 2.7 5.0 1.7 Feeding 12.0 6.7 0.3 1.0 0.0 Watering 6.0 4.0 4.5 2.2 3.2 Herding 7.3 3.4 6.0 2.3 0.8 Housing animals with humans 4.0 4.0 4.0 4.0 4.0 Exposing to insect bite during herding 10.3 3.8 3.5 1.8 1.8 Managing sick animals 6.5 7.8 2.3 3.2 0.2 Marketing 9.2 4.7 5.8 0.6 0.0 Assisting mating 7.0 12.0 0.0 1.0 0.0 Agricultural activities 15.7 0.0 4.3 0.0 0.0 General farm work (livestock + agriculture) 8.7 5.8 2.5 1.7 1.3 Overall 9.0 4.7 3.5 1.8 1.4 Zoonotic diseases Cleaning Barns 0.0 6.0 6.0 8.0 0.0 Eating/drinking raw meat/milk 6.8 4.8 3.0 2.3 2.0 Feeding 5.0 10.0 1.0 4.0 0.0 Handling contaminated feed 3.0 13.0 2.0 2.0 0.0 Herding 5.6 4.5 4.1 2.3 3.4 Watering 4.1 5.1 3.6 2.8 4.4 Housing animals with humans 7.0 4.0 6.0 3.0 0.0 Managing sick animals 10.8 6.0 0.8 2.0 0.0 Marketing 9.8 4.6 5.0 0.6 0.0 Assisting mating 10.0 10.0 Slaughtering 6.0 4.2 1.0 0.6 8.2 Slaughtering or cooking 9.2 4.2 4.2 3.0 0.0 Renting animals form neighbors 9.0 6.5 4.5 0.0 0.0 Treating sick animals 13.4 4.5 4.4 0.8 1.0 General farm work (livestock + agriculture) 6.5 9.0 2.5 2.0 0.0 Overall 7.1 6.2 3.9 2.4 1.4 The major effects of diseases were loss of household income, resulting from lower animal productivity and higher mortality, and zoonotic risks (Table 7 ). Anthrax was recognized by 77.5 of the respondents as the most important zoonotic disease. Loss of oxen and other valuable animals could also impact agricultural activities and saving/insurance since livestock serve as capital store in rural Ethiopia. In extreme cases, farmers could be displaced from their ancestral land and migrate to towns or become dependent on government handouts and children would be out of school. Table 7 Percentage of FGD groups reporting the different impacts of diseases on households in Ethiopia Impacts Anthrax Blackleg Bovine pasteurellosis FMD LSD Overall Reduced income 77.5 69.8 83.3 81.7 73.2 76.7 Mortality 45.0 55.6 50.0 51.7 39.0 49.2 Reduced productivity 42.5 61.9 72.2 65.0 73.2 62.9 Time for caring sick animals 12.5 12.7 13.9 8.3 9.8 11.2 Cost of treatment 22.5 15.9 8.3 11.7 19.5 15.4 Reduced market value 10.0, 7.9 – 3.3 34.1 10.4 Malnutrition 10.0 14.3 2.8 20.0 9.8 0.12 Impact on human health 77.5 14.3 19.4 – 14.6 22.1 School drop outs 25.0 7.9 – 18.3 9.8, 12.5 Psychological/social impacts – 6.4 5.6 1.7 2.4 0.84 Renting-out land/renting-in oxen due to loss of oxen 20.0 7.9 2.8 18.3 12.2 12.5 Migration 15.0 9.5 – 18.3 9.8 11.2 Women and men adult members were equally more affected by diseases than the other members of the family (Fig. 4 ), the reason provided by the respondents being that adults had to work harder to supplement their reduced income and animal source foods due to death of oxen and milking cows. Adult members were also more at risk of zoonotic diseases, as they were more responsible for caring, slaughtering, and cooking meat from animals unsuspected of being infected with zoonotic diseases. However, the perception of the FGD groups varied. Twenty-eight percent of FGD groups believed lower livestock productivity and higher mortality would affect women more since they are directly responsible for providing food for the family including the husband and care for young children who depend on cow milk. Women are also more affected by zoonotic diseases as they are responsible for the management of sick animals. The rest provided either equal (24.9% of FGDs) or more scores to male adults (46.9% of FGDs). Fig. 4 FGD groups' scoring (out of a total score of 20) of the impacts of diseases on family members in the highlands of Ethiopia Male and women farmers had similar understanding of the effects of diseases on the different classes of animals (Table 8 ). Farmers understood that anthrax affect all age groups of livestock but inflicts high morbidity in adults. Incidence of anthrax was low in young animals up to the age of 3 months as they were kept indoor and less likely to be exposed to the pathogen. However, young grazing animals aged 3–12 months and animals in good body condition/with heavy muscle were severely affected. LSD and blackleg were also understood by farmers to affect mostly adult animals, as they are exposed to the external environment, and animals in good body condition in the case of blackleg. On the contrary, the farmers observed that FMD occurs in all age groups, but it is very severe among younger age groups (calves) due to their low ability to cope with the disease. Table 8 Animal age groups most affected by five most important diseases as perceived by women and men farmers in Ethiopia Scorings of women and male FGD groups Disease Gender of FGD members Newborn Young Adult female Adult male Anthrax Men 2.3 8.2 10.5 14.1 Women 4.7 7.5 11.3 14.5 Blackleg Men 2.9 6.3 11.0 14.6 Women 1.3 5.9 12.1 19.3 Bovine Pasteurellosis Men 0.0 12.0 12.0 16.0 Women 0.0 8.0 12.0 18.0 FMD Men 14.7 4.5 9.0 8.5 Women 14.7 4.5 9.0 8.5 Lumpy skin disease Men 1.6 4.6 12.1 13.3 Women 3.7 6.0 12.1 13.9 FGD Focus group discussion, FMD foot and mouth disease Disease coping mechanisms Farmers coping mechanisms against diseases include both traditional and modern practices. The traditional practices include treating sick animals with herbs such as Endod/Phytolacca dodecandra against anthrax, incision on the brisket and mandibular vein of animals affected with anthrax, incision on the hind quarter, and burning the affected part with hot metal/branding in the case of blackleg, use of 'holy water', applying crushed millipedes mixed with honey on the tongue of animals affected with FMD. There are also traditional superstitious practices. The modern practices mentioned were treatment of sick animals in veterinary clinics, vaccination, and health management practices including isolation of sick animals and improved feeding and housing management for sick animals. Most of the farmers used modern treatment and a combination of modern and traditional treatment (Table 9 ). Improved health management was the least practiced intervention. Table 9 Percentage of respondents using different coping mechanisms against five top livestock diseases in Ethiopia Coping mechanisms Anthrax Blackleg Bovine pasteurellosis FMD LSD Overall A * 23.1 a, b 27.6 a, b 52.8 b, c 16.9 a 65.9 c 34.80 B 7.7 a – 2.8 a 8.5 a 9.8 a 5.60 A + B 7.7 a 6.9 a 2.8 a 8.5 a 9.8 a 7.30 C – 3.4 a, b 5.6 a, b 18.6 b 2.4 a, b 6.90 A + C 51.3 a, b 50.0 b 19.4 c 23.7 a, c 9.8 c 31.80 A + C + B 2.6 a 12.1 a 13.9 a 5.1 a 2.4 a 7.30 C + B 7.7 a, b, c – 2.8 a, b, c 18.6 b – 6.40 * a Modern veterinary services, including vaccination and treatment in veterinary clinics b Health management, including isolation of sick animals, feeding, and housing of sick animals, c Traditional healing, incisions, branding, use of herbs, and some superstitious ones Priority of livestock species The majority of the focus group discussion (FGD) groups ranked cattle as their most important livestock species. The average scores allocated by men and women FGDs were 38.0 and 40.5 for cattle, 18.6 and 19.4 for sheep, 15.5 and 16.3 for poultry, 14.9 and 15.6 for goats, and 14.3 and 13.4 for equines. The scores in Amhara, Oromia, SNNPR, and Tigray regions were 41.5, 38.3, 39.3, and 37.4 for cattle, 18.5, 18.3, 19.6, and 21.4 for sheep, 12.6, 18.7, 16.6, and 16.8 for poultry, 12.9, 12.8, 16.1, and 19.4 for goat and 15.9, 15.0, 13.9, and 9.6 for equine. The generalized linear model analysis (Table 2 ) indicated cattle were significantly the most preferred species with the odds of cattle scoring higher than sheep being 2.08 times ( P = 0.000). Sheep and poultry were equally the second most important livestock, followed by goat and equine. There were no statistically significant differences between men and women FGD groups in their preference for livestock species (Table 2 ). There was also no significant variation in the preferences of farmers for cattle, sheep, and poultry in the four regions studied. However, goats received significantly higher scores in Tigray region compared to Oromia (odds ratio = 0.66, P = 0.009) and Amhara (odds ratio = 0.67, P = 0.005) resulting in goats being more important than poultry in Tigray. Similarly, compared with Tigray, equines were 1.65 ( P = 0.008) and 1.55 ( P = 0.024) times more likely to receive higher scores in Amhara and Oromia, respectively, where they are more important than goats. Table 2 The odds (Exp(B)) of men and women FGD groups allocating higher scores to the different livestock species in reference to sheep in Ethiopia B Std. error Sig. Exp(B) Intercept 2.966 0.0655 0.000 19.414 Cattle 0.736 0.0715 0.000 2.088 Equine − 0.370 0.1116 0.001 0.691 Goat − 0.217 0.1068 0.042 0.805 Poultry − 0.175 0.0974 0.072 0.839 Sheep 0 a 1 Cattle—Men FGD − 0.064 0.0417 0.125 0.938 Cattle—Women FGD 0 a 1 Equine—Men FGD 0.062 0.1214 0.612 1.063 Equine—Women FGD 0 a 1 Goat—Men FGD − 0.048 0.1200 0.689 0.953 Goat—Women FGD 0 a 1 Poultry—Men FGD − 0.049 0.1037 0.636 0.952 Poultry—Women FGD 0 a 1 Sheep—Men FGD −0.045 0.0948 0.632 0.956 Sheep—Women FGD 0 a 1 a Reference category = sheep; women FGD The reasons provided by respondents for ranking cattle as the most important species reflected the production objectives of farmers in mixed crop-livestock systems, the most important reasons being draught power/income from fattened cattle and milk production, mentioned by 82.9% and 84.3% of the respondents, respectively. There was no difference in the reasons provided by men and women FGD groups and across the four regions, except that significantly smaller percentage of FGD groups cited milk (52.9%) and income from fattening (47.1%) in Tigray and drought power (40%) in SNNPR. Manure was also mentioned as important function of cattle by 34.3% of women and 37.1% of male FGDs, and by 43.5%, 40.0% and 41.2% of the respondents in Amhara, Oromia, and Tigray regions, respectively. Unlike in the pastoral system, socio-cultural values, and asset/saving functions of cattle were less important functions of cattle in the highland mixed crop-livestock system. Priority of diseases Twenty-eight diseases or disease syndromes were identified by the FGD groups. Reflecting the importance of cattle, most of the diseases identified were cattle diseases, though some of them also affect small ruminants. Two poultry-specific (New castle disease and salmonellosis, though salmonellosis also affects ruminants) and one equine-specific disease (African horse sickness) were also identified. However, only 12 diseases were mentioned by at least 10% of the FGD groups (Fig. 2 ). The 12 diseases, which were considered for further analysis, were blackleg, food and mouth disease (FMD), lumpy skin disease (LSD), anthrax, bovine pasteurellosis, mastitis, endoparasitosis, ectoparasitosis, fasciolosis, rabies, foot rot, and trypanosomiasis. Fig. 2 Frequency of FGD groups who mentioned 28 diseases among their top five diseases in four regions of Ethiopia. (CBPP = contagious bovine pleuropneumonia, CCPP = contagious caprine pleuropneumonia, FMD = foot and mouth disease, LSD = lumpy skin disease, NCD = New castle disease) The top five diseases based on their average scores were blackleg (18.9), FMD (13.8), LSD (12.5), anthrax (10.6), and bovine pasteurellosis (8.8). Statistical analysis (Table 3 ) also confirmed these were the five priority diseases. Blackleg was significantly the most important disease. Farmers are more likely to allocate higher score to blackleg compared with the reference LSD (odds ratio = 1.42; P = 0.021). However, farmers would allocate statistically similar scores for anthrax, FMD, bovine pasteurellosis, and LSD (Table 3 ). Table 3 The odds of farmers allocating higher scores for diseases in reference to LSD (lumpy skin disease) and men allocating higher scores in reference to women the highlands of Ethiopia Β Std. error Sig. ( P ) Odds ratio Intercept 2.593 0.1240 0.000 13.371 Anthrax − 0.219 0.1981 0.269 0.803 Blackleg 0.351 0.1517 0.021 1.421 Bovine pasteurellosis − 0.365 0.2175 0.094 0.694 Ectoparasite − 1.067 0.3813 0.005 0.344 Endoparasite − 1.186 0.4245 0.005 0.306 Fasciolosis − 1.439 0.5375 0.007 0.237 FMD − 0.155 0.1907 0.418 0.857 Foot-rot − 2.105 1.0260 0.040 0.122 Mastitis − 1.158 0.4140 0.005 0.314 Ovine pasteurellosis − 2.511 1.5328 0.101 0.081 Rabies − 2.071 0.9918 0.037 0.126 Trypanosomiasis − 2.105 1.0260 0.040 0.122 LSD 0 a 1 Men FGD scoring b Anthrax 0.452 0.1830 0.013 1.572 Blackleg − 0.009 0.1240 0.942 0.991 Bovine pasteurellosis − 0.107 0.2673 0.689 0.898 Ectoparasite − 0.507 0.6987 0.468 0.602 Endoparasite − 0.058 0.5914 0.922 0.944 Fasciolosis 0.018 0.7331 0.981 1.018 FMD − 0.156 0.2227 0.483 0.855 Foot-rot − 0.214 1.6218 0.895 0.807 Mastitis 0.047 0.5460 0.932 1.048 Ovine pasteurellosis 0.234 1.9486 0.905 1.263 Rabies − 0.271 1.6225 0.867 0.763 Trypanosomiasis 0.728 1.1311 0.520 2.070 LSD − 0.147 0.1898 0.439 0.863 a Lumpy skin disease (LSD) is the reference category. FGD focus group discussion groups b The reference category: Men FGDs disease scoring were compared with women FGDs scoring for each disease Men and women respondents had similar perceptions of diseases, their priority diseases being statistically similar except men allocated significantly higher scores to the importance of anthrax in reference to LSD (Table 3 ). Women and men groups' scores were 19 and 18.3 for blackleg, 11.46 and 9.8 for FMD, 13.4 and 11.5 for LSD, 10.7 and 16.9 for anthrax, and 9.29 and 8.34 for bovine pasteurellosis. There were slight variations across the four regions in disease priorities. Among the most frequently mentioned diseases, those that were common across the regions were blackleg, FMD, LSD, anthrax, bovine pasteurellosis, and ectoparasitosis. Mastitis was mentioned in Oromia only and cowdriosis (heartwater) and Newcastle disease (NCD) in SNPPR only. In reference to the list of five top diseases across regions (Table 3 ), the list in Amhara and Oromia excluded bovine Pasteurellosis and included trypanosomiasis and mastitis, respectively. In SNNPR, the list included mastitis and fasciolosis and excluded FMD and anthrax. The top five in Tigray excluded LSD and ectoparasites. The FGD groups considered a range of criteria to rank diseases (Table 4 ). The reasons for anthrax were the sudden death it inflicts and its potential effect on human health as it is a zoonotic disease, reasoned by 81.2% and 50.0% of the FGDs. Although not supported by the existing modern veterinary knowledge, 7.1% of the respondents considered LSD as a zoonotic disease. Blackleg was perceived to infect animals with good body condition such as well fattened animals. LSD was associated with high morbidity rate resulting in reduced milk production and work performance of oxen by almost all respondents and high transmission rate by about a third of the respondents who ranked the disease. Table 4 Percentage of farmers providing reasons for top ranking diseases in the highlands of Ethiopia Reasons for ranking Anthrax Blackleg Bovine pasteurellosis FMD LSD Zoonotic disease 50.0 6.2 0.0 0.0 7.1 Affects all age groups/species 43.8 6.2 0.0 75.0 14.3 Affects animals in good condition 6.2 56.2 100.0 75.0 0.0 Sudden death 81.2 37.5 66.7 0.0 7.1 High morbidity/effect on productivity 0.0 37.5 0.0 25.0 92.9 High mortality 43.8 18.8 0.0 25.0 7.1 High transmission rate 0.0 12.5 0.0 25.0 35.7 Affects skin quality 0.0 0.0 0.0 0.0 21.4 Not common in locality 0.0 6.2 0.0 0.0 0.0 Agroecological and seasonal distribution of diseases Agroecology is the major determinant of disease epidemiology. Based on a combined criterion of the proportion of respondents ranking the diseases 1st to 3rd and those not including the disease in their top five list, Anthrax, FMD, and bovine pasteurellosis were the most important diseases in the dry highlands (Table 5 ). LSD and blackleg in the moist lowlands and blackleg and LSD in the wet highlands were the most important diseases. Table 5 Per cent of respondents ranking the top five diseases in different agro-ecological zones of Ethiopia Agroecology 1 Rank 2 Anthrax Blackleg Bovine pasteurellosis FMD LSD Dry highland 0 3 38.1 a 4 38.1 a 28.6 a 19.0 a 90.5 b 1 38.1 a 14.3 a, b 14.3 a, b 19.0 a, b – 2 19.0 a, b 9.5 a, b 42.9 b 9.5 a, b – 3 4.8 a 23.8 a 4.8 a 23.8 a – Moist highland 0 46.9 a, b 12.5 c 78.1 b 28.1 a, c 25.0 a, c 1 21.9 a, b 18.8 a, b – – 34.4 a 2 28.1 a 31.2 a 12.5 a, b – 9.4 a, b 3 – 15.6 a, b 9.4 a, b 28.1 b 9.4 a, b Wet highland 0 56.2 a, b 31.2 a, b 62.5 a, b 75.0 b 18.8 a 1 6.2 a, b 43.8 b – – 18.8 a, b 2 – 25.0 a – – 18.8 a 3 12.5 a – – 6.2 a – 1 Result for moist lowland zone is not presented. None of the respondents included anthrax and LSD in the top five diseases list, whereas all respondents in this zone ranked Blackleg, pasteurellosis, and FMD as the 2nd, 4th, and 5th most important diseases, respectively 2 Ranks derived from the number of beans (scores) allocated to the diseases 3 The disease was not included in farmers' lists of top five diseases 4 Different superscript letters within row denotes percent of respondents ranking the diseases differ significantly from each other at the 0.05 level Seasons of the year in the highlands of Ethiopia are generally divided into four, although there are some variations across agroecological zones and geographic locations (regions). Generally, the four seasons are the dry/cold (September to November, dry/cold), the dry/hot (December to February, dry/hot), the short rainy (March to May, short rainy season), and the long rainy (June to August, long rainy season) seasons. Interviewed farmers observed a higher prevalence of anthrax in dry/cold and short rainy seasons, blackleg in season dry/cold, short rainy seasons, and long rainy seasons, pasteurellosis in season dry/hot and short rainy seasons, FMD in dry/hot and short rainy seasons, and LSD in dry/cold season and long rainy seasons (Fig. 3 ). Fig. 3 Average scorings of farmers for the prevalence level of five top diseases in four seasons of the year in the highlands of Ethiopia Transmission and impacts of diseases Women and men respondents' understanding of disease transmission pathways was similar. Adults were more responsible than the rest of the family members for the transmission of diseases. Adult men played more role in the transmission of non-zoonotic diseases, but both men and women were equally responsible for zoonotic diseases (Table 6 ). Adult men scored more for transmission activities involving animal movement, disposal of dead animals, and handling of contaminated tools, whereas women scored more in activities related to managing sick animals. However, both men and women are equally responsible for the transmission of zoonotic diseases. Table 6 Respondents' scorings (out of 20) of the roles of family members in the transmission of diseases in the highlands of Ethiopia Transmission activity Adult men Adult women Young males Young females Children Non-zoonotic diseases Moving animals to distant places 11.0 0.0 9.0 0.0 0.0 Burying dead animals 10.0 4.0 4.7 1.3 0.0 Handling contaminated tool 14.0 6.0 0.0 0.0 4.0 Eating/drinking raw meat/milk 9.0 2.3 2.3 2.3 4.0 Eating without washing hands 4.0 6.7 2.7 5.0 1.7 Feeding 12.0 6.7 0.3 1.0 0.0 Watering 6.0 4.0 4.5 2.2 3.2 Herding 7.3 3.4 6.0 2.3 0.8 Housing animals with humans 4.0 4.0 4.0 4.0 4.0 Exposing to insect bite during herding 10.3 3.8 3.5 1.8 1.8 Managing sick animals 6.5 7.8 2.3 3.2 0.2 Marketing 9.2 4.7 5.8 0.6 0.0 Assisting mating 7.0 12.0 0.0 1.0 0.0 Agricultural activities 15.7 0.0 4.3 0.0 0.0 General farm work (livestock + agriculture) 8.7 5.8 2.5 1.7 1.3 Overall 9.0 4.7 3.5 1.8 1.4 Zoonotic diseases Cleaning Barns 0.0 6.0 6.0 8.0 0.0 Eating/drinking raw meat/milk 6.8 4.8 3.0 2.3 2.0 Feeding 5.0 10.0 1.0 4.0 0.0 Handling contaminated feed 3.0 13.0 2.0 2.0 0.0 Herding 5.6 4.5 4.1 2.3 3.4 Watering 4.1 5.1 3.6 2.8 4.4 Housing animals with humans 7.0 4.0 6.0 3.0 0.0 Managing sick animals 10.8 6.0 0.8 2.0 0.0 Marketing 9.8 4.6 5.0 0.6 0.0 Assisting mating 10.0 10.0 Slaughtering 6.0 4.2 1.0 0.6 8.2 Slaughtering or cooking 9.2 4.2 4.2 3.0 0.0 Renting animals form neighbors 9.0 6.5 4.5 0.0 0.0 Treating sick animals 13.4 4.5 4.4 0.8 1.0 General farm work (livestock + agriculture) 6.5 9.0 2.5 2.0 0.0 Overall 7.1 6.2 3.9 2.4 1.4 The major effects of diseases were loss of household income, resulting from lower animal productivity and higher mortality, and zoonotic risks (Table 7 ). Anthrax was recognized by 77.5 of the respondents as the most important zoonotic disease. Loss of oxen and other valuable animals could also impact agricultural activities and saving/insurance since livestock serve as capital store in rural Ethiopia. In extreme cases, farmers could be displaced from their ancestral land and migrate to towns or become dependent on government handouts and children would be out of school. Table 7 Percentage of FGD groups reporting the different impacts of diseases on households in Ethiopia Impacts Anthrax Blackleg Bovine pasteurellosis FMD LSD Overall Reduced income 77.5 69.8 83.3 81.7 73.2 76.7 Mortality 45.0 55.6 50.0 51.7 39.0 49.2 Reduced productivity 42.5 61.9 72.2 65.0 73.2 62.9 Time for caring sick animals 12.5 12.7 13.9 8.3 9.8 11.2 Cost of treatment 22.5 15.9 8.3 11.7 19.5 15.4 Reduced market value 10.0, 7.9 – 3.3 34.1 10.4 Malnutrition 10.0 14.3 2.8 20.0 9.8 0.12 Impact on human health 77.5 14.3 19.4 – 14.6 22.1 School drop outs 25.0 7.9 – 18.3 9.8, 12.5 Psychological/social impacts – 6.4 5.6 1.7 2.4 0.84 Renting-out land/renting-in oxen due to loss of oxen 20.0 7.9 2.8 18.3 12.2 12.5 Migration 15.0 9.5 – 18.3 9.8 11.2 Women and men adult members were equally more affected by diseases than the other members of the family (Fig. 4 ), the reason provided by the respondents being that adults had to work harder to supplement their reduced income and animal source foods due to death of oxen and milking cows. Adult members were also more at risk of zoonotic diseases, as they were more responsible for caring, slaughtering, and cooking meat from animals unsuspected of being infected with zoonotic diseases. However, the perception of the FGD groups varied. Twenty-eight percent of FGD groups believed lower livestock productivity and higher mortality would affect women more since they are directly responsible for providing food for the family including the husband and care for young children who depend on cow milk. Women are also more affected by zoonotic diseases as they are responsible for the management of sick animals. The rest provided either equal (24.9% of FGDs) or more scores to male adults (46.9% of FGDs). Fig. 4 FGD groups' scoring (out of a total score of 20) of the impacts of diseases on family members in the highlands of Ethiopia Male and women farmers had similar understanding of the effects of diseases on the different classes of animals (Table 8 ). Farmers understood that anthrax affect all age groups of livestock but inflicts high morbidity in adults. Incidence of anthrax was low in young animals up to the age of 3 months as they were kept indoor and less likely to be exposed to the pathogen. However, young grazing animals aged 3–12 months and animals in good body condition/with heavy muscle were severely affected. LSD and blackleg were also understood by farmers to affect mostly adult animals, as they are exposed to the external environment, and animals in good body condition in the case of blackleg. On the contrary, the farmers observed that FMD occurs in all age groups, but it is very severe among younger age groups (calves) due to their low ability to cope with the disease. Table 8 Animal age groups most affected by five most important diseases as perceived by women and men farmers in Ethiopia Scorings of women and male FGD groups Disease Gender of FGD members Newborn Young Adult female Adult male Anthrax Men 2.3 8.2 10.5 14.1 Women 4.7 7.5 11.3 14.5 Blackleg Men 2.9 6.3 11.0 14.6 Women 1.3 5.9 12.1 19.3 Bovine Pasteurellosis Men 0.0 12.0 12.0 16.0 Women 0.0 8.0 12.0 18.0 FMD Men 14.7 4.5 9.0 8.5 Women 14.7 4.5 9.0 8.5 Lumpy skin disease Men 1.6 4.6 12.1 13.3 Women 3.7 6.0 12.1 13.9 FGD Focus group discussion, FMD foot and mouth disease Disease coping mechanisms Farmers coping mechanisms against diseases include both traditional and modern practices. The traditional practices include treating sick animals with herbs such as Endod/Phytolacca dodecandra against anthrax, incision on the brisket and mandibular vein of animals affected with anthrax, incision on the hind quarter, and burning the affected part with hot metal/branding in the case of blackleg, use of 'holy water', applying crushed millipedes mixed with honey on the tongue of animals affected with FMD. There are also traditional superstitious practices. The modern practices mentioned were treatment of sick animals in veterinary clinics, vaccination, and health management practices including isolation of sick animals and improved feeding and housing management for sick animals. Most of the farmers used modern treatment and a combination of modern and traditional treatment (Table 9 ). Improved health management was the least practiced intervention. Table 9 Percentage of respondents using different coping mechanisms against five top livestock diseases in Ethiopia Coping mechanisms Anthrax Blackleg Bovine pasteurellosis FMD LSD Overall A * 23.1 a, b 27.6 a, b 52.8 b, c 16.9 a 65.9 c 34.80 B 7.7 a – 2.8 a 8.5 a 9.8 a 5.60 A + B 7.7 a 6.9 a 2.8 a 8.5 a 9.8 a 7.30 C – 3.4 a, b 5.6 a, b 18.6 b 2.4 a, b 6.90 A + C 51.3 a, b 50.0 b 19.4 c 23.7 a, c 9.8 c 31.80 A + C + B 2.6 a 12.1 a 13.9 a 5.1 a 2.4 a 7.30 C + B 7.7 a, b, c – 2.8 a, b, c 18.6 b – 6.40 * a Modern veterinary services, including vaccination and treatment in veterinary clinics b Health management, including isolation of sick animals, feeding, and housing of sick animals, c Traditional healing, incisions, branding, use of herbs, and some superstitious ones Discussion Livestock play multiple roles in the livelihoods of smallholder farmers in the highlands of Ethiopia. Thus, most farmers keep all types of livestock species since each species serves a different function and such relative importance of livestock species determines farmers' relative preferences for the different species. The preference for cattle over other species in the current study is to be expected. In the highland mixed crop-livestock system, cattle play a unique role besides providing animals and animal products for home consumption and sale, namely providing oxen power. This was also confirmed by 82.9% respondents in the current study who mentioned drought power as their reason for prioritizing cattle over other species. Sickness or death of oxen could result in extra costs for renting oxen, or worse, having to renting out crop land and could lead to absolute poverty. Land holding and ownership of oxen are among the significant determinants to avoid rural poverty in Ethiopia (Bogale et al. 2005 ). Accordingly, most of the diseases identified by the FGD groups in the current study were cattle diseases including two diseases among the top five diseases affecting cattle only, emphasizing the role of healthy cattle. The livestock health situation in Ethiopia is comparable or worse than in other African countries. Across Africa, disease outbreaks have been increasing. For instance, FMD (and also pasteurellosis) outbreaks have been increasing from 378 outbreaks from 26 countries in 2009, to 454 outbreaks from 24 countries in 2010, and to 902 outbreaks from 28 countries in 2011. A similar trend was also reported for pasteurellosis. In 2011, Ethiopia reported the highest number of FMD fatalities (721 = 0.0009% of the ruminant population), the highest number of pasteurellosis outbreaks (570 = 0.00071%), and the highest number of deaths from pasteurellosis (1633 = 0.002%) among the AU-IBAR countries ( http://www.au-ibar.org/foot-and-mouth-disease-in-ruminants ). The priority diseases identified by the FGDs in the current study correspond to previous findings in the same regions. Dereje and Shibiru ( 2016 ) reported that out of 1632 examined disease outbreaks from 2008 to 2013 in 17 districts in East Wollega zone of Oromia region in Ethiopia (one of the current study sites), 21.8%, 21.6%, 15.6%, 9.3%, 7.8%, and 7.2% were found to be positive for bovine pasteurellosis, blackleg, LSD, FMD, CBPP, and anthrax. Similar disease priorities have been reported from SNNPR region (blackleg, mastitis, lumpy skin disease, and tick infestation; Albe et al. 2016 ) and Tigray region (blackleg, anthrax, LSD, mastitis, and FMD in order of importance; Tekle 2007 ). Pastoralists and farmers traditionally describe livestock diseases by their local names, and some of the diseases are locally known by the same name in more than one region, though the communities in the different regions speak different languages. For instance, blackleg ( Gangraena emphysematosa caused by Clostridium chauvoei ) is known as Aba Gorba both in Oromia and SNNPR, and bovine pasteurellosis as Gororsa in Amhara, Oromia and SNNPR regions. Local veterinarians are well acquainted with the local disease nomenclatures which help to translate farmers perceptions into modern veterinary nomenclatures. According to Dereje and Shibiru (2016), lack of attention for prophylactic vaccination, misuse of veterinary drugs, lack of proper management, and poor outbreak reporting system could have contributed to the high prevalence of some of the diseases (bovine pasteurellosis, blackleg, and LSD). Similarly, a review of animal health service in the four highland regions of Ethiopia (similar to the location of the current study), Hooper ( 2016 ) has shown that most of the animal health services are provided by the government, while the private sector involvement is insignificant mainly serving as drug shops. The network of government animal health service is well structured with rural animal health posts serving a group of two to five kebeles and district level clinics coordinating the rural health posts and diagnostics and surveillance laboratories at region level. However, while the district clinics are relatively well staffed and equipped, the rural health posts are often ill equipped and staffed, and the situation is challenging for the rural staff to meet their responsibilities. Further, the regional laboratories are not functioning at their full capacity or not meeting their goals. The relative importance of diseases would vary depending on the criteria used for ranking or when ranking is done from different perspectives. For instance, the livestock sector analysis for the livestock master plan of Ethiopia (Shapiro et al. 2017 ) ranked livestock diseases based on three criteria, namely the impact on households and livelihood framework, impact on markets and value chains, and impact on intensification pathways in the production systems. The scores provided using the three criteria were again weighted, respectively, according to the share of the households in the affected production systems, total value added generated from the sub-chain affected, and the animal population in production systems affected. Among the top 10 diseases ranked based on their incidence, the top three livestock diseases according to the three criteria were, respectively, FMD, CBPP, and tuberculosis; FMD, LSD, and brucellosis; and brucellosis, FMD, and tuberculosis. Ranking of diseases may also vary depending on the category of respondents. These priority diseases included the priority diseases of the farmers in the current study but included other diseases. Priority of diseases could also differ depending on the respondents. Hassen et al. ( 2014 ) interviewing smallholders and professionals in North Gondar, Amhara region, found LSD as the most important disease for smallholder farmers, being mentioned by 46.3% of the respondents, while anthrax and gastro-intestinal tract parasites were more frequently mentioned by professionals in the public and private services, and FMD was frequently mentioned by both professionals and farmers. Livestock disease priorities may vary depending on various determinants. In the current study, variations were found in diseases women and men farmers prioritize, and variation in the relative importance of diseases across agro-ecological zones and seasons. Livestock development projects commonly target the male household heads for identifying constraints and designing development strategies. The approach followed in the current study, namely separate FGDs with both women and men groups in the same kebele , enabled to assess the perceptions and priorities of both women and men farmers, providing a more in-depth view of disease importance. The results showed that women may have similar perceptions, understanding, and priorities as men, though there are some differences between the genders including slight variation in their perceptions regarding the relative importance of anthrax. We found that agroecology and seasons determine the epidemiology of diseases. The high importance of LSD in the wet highlands in the current study is supported by the characteristics of this disease to be closely associated with prolonged and heavy rains which favor an increase in the population of the vector (Regassa 2003 ). Similarly, LSD was reported in Uganda by 67.5% of respondents in semi-humid tropics (above 1000 mm) and 86.1% in warm humid tropics (1000–1500 mm) but only by 20% in warm semi-arid climate receiving 500–1000 mm annual rainfall (Ocaido et al. 2009 ). Farmers associate disease incidence with seasons. The high incidence of anthrax during the drier seasons compared with the long rainy season was, according to respondents' reasoning, due to the high temperature and feed shortage which forced the animals to graze close to the ground and contract the disease agents. Farmers associated blackleg, although it occurs throughout the year, with wet seasons (long and short rainy seasons) where feed is more available and body condition of the animals are improved and are more vulnerable to the disease. The respondent farmers seem to be highly knowledgeable about blackleg epidemiology which corresponds to the epidemiology of clostridial diseases (PRIMEFACT 2007 ) in that it generally affects the best conditioned animals, with most losses occurring during wet seasons when clostridial spores are washed out and where there is an abundance of feed. Farmers associate LSD with the wet season as a result of increased stagnant water bodies and contact between different village herds in communal grazing lands. FMD is believed to occur when animals are in free range during the months of October to January. Bovine pasteurellosis is mostly associated with seasons of hot weather (February to May) and workload such as plowing. The use of participatory epidemiology tool for disease surveillance, survey, prioritization, and control has been increasing, especially since 2012 in Africa and Asia, according to a comprehensive inventory of participatory epidemiology activities (Alberto et al. 2017 ). Unlike the criteria used to rank diseases in conventional studies which rely mainly on incidence or prevalence of diseases (Hooper 2016 ; Shapiro et al. 2017 ), farmers consider a range of criteria to rank diseases. For instance, in the current study, the main underlying reasons mentioned by the respondents for allocating the highest scores the suddenness of death, zoonosis, and the types of animals the disease affects. For instance, blackleg was ranked high since it infects animals with good body condition such as well fattened animals which could fetch tens of thousands of Birr (the local currency) and thus cause great economic loss for households affected by the disease. Farmers also consider the diverse function of cattle in ranking diseases such as effect of diseases on work performance of oxen, as was the case in the current study in ranking LSD, which may not be considered in conventional studies. Proportional piling rather than direct ranking was used in the current study. Proportional piling is a suitable approach for working with illiterate farmers as it is a visual assessment tool. Its advantage over ranking is that the results are numerically meaningful providing a distance measure between preferences for different items. The limitation of the approach could be forced scoring when small number of seeds/pebbles/beans are used. However, this shortcoming can be overcome by using a larger number (100) of seeds/pebbles/beans since this increase respondents' flexibility in expressing their strength of preference for one item over another (Abeyasekera 2001 ). A striking similarity is observed between the ranking of diseases by farmers in the current study and by clinicians and heads of district offices of livestock agencies in a previous study in the same four regions (Hooper 2016 ) where all the five top diseases from the current study were indicated as the most common diseases except trypanosomiasis which was mentioned by more professionals (33) than FMD (27). This confirms that farmers are knowledgeable about diseases. A similar argument has been made that local farmers are knowledgeable about common disease that affect their poultry, have good diagnostic abilities, and can recognize clinical signs, and recommendations have been made that there should be improved veterinary outreach to share livestock health information through active community participation, and participatory epidemiology should be incorporated into state and national disease surveillance systems (Jibril et al. 2015 ). Indigenous disease treatment knowledge and practices identified through participatory epidemiology need to be investigated and utilized as alternative remedies. For instance, application of hot water or hot iron on the affected portion of animals infected with blackleg in the current study is shared by farmers in other parts of Africa (Mahachi 2016 ). Nevertheless, the current study also indicated gaps in farmers' knowledge of the epidemiology of diseases. For instance, some FGD groups considered FMD and LSD as zoonotic diseases and superstitious practices were reported as remedies for diseases. The high incidence of diseases reported by the FGDs in this study could be attributed to the veterinary service situation in the country. According to the 2013 animal health strategy and vision for Ethiopia (MoA and ILRI 2013 ), growth of private animal health service delivery is constrained by absence of an enabling policy environment, a system of sanitary mandates does not exist, only 45% of the country is served by the animal health delivery systems. Although the National Veterinary Institute is producing a wide range of vaccines, some essential vaccines are not produced or are not produced in sufficient quantity and quality to achieve sufficient vaccination cover to interrupt disease transmission, and disease surveillance and reporting is poor and irregular, with only about 30–35% of the districts submitting disease outbreak reports (MoA and ILRI 2013 ). Conclusions Cattle are the most important livestock for smallholder farmers in the highlands of Ethiopia. The top five diseases afflicting cattle in the highlands of Ethiopia are blackleg, anthrax, LSD, FMD, and bovine pasteurellosis. However, the relative importance of diseases varies across agro-ecological zones and seasons. This study confirmed that men and women farmers have similar and equal understanding of the epidemiology of diseases and are equally affected by diseases. It is therefore commendable to include both male and female household members in designing health interventions. Farmers use both modern and alternative traditional remedies to prevent diseases and treat sick animals. However, vaccination and health management are rarely used. The PE research approach used in this study proved to be reliable approach to get insights into farmers' knowledge of disease epidemiology, which corresponded to a large extent to results from conventional studies. Based on the findings of this study, the intervention strategies listed in the 2013 animal health strategy (MoA and ILRI 2013 ) seem pertinent and need to be implemented to overcome the challenges of diseases in the country.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7022184/

Optimization of a Plasmodium falciparum circumsporozoite protein repeat vaccine using the tobacco mosaic virus platform

Significance RTS,S/AS01 is a circumsporozoite protein (CSP)-based malaria vaccine that confers partial protection against malaria in endemic areas. Recent reports have elucidated structures of monoclonal antibodies that bind to the central (NPNA) repeat region of CSP and that inhibit parasite invasion. Antigen configuration and copy number of CSP repeats displayed on a tobacco mosaic virus (TMV) particle platform were studied. A TMV vaccine containing CSP repeats displayed as a loop induced 10× better antibody titer than a nearly full-length CSP in mice. In rhesus model, this translated to a 5× improvement in titer. Rhesus antibodies potently inhibited parasite invasion up to 11 mo after vaccination. An optimized epitope-focused, repeat-only CSP vaccine may be sufficient or better than the existing CSP vaccines. General Methods For detailed methods, please refer to SI Appendix . Vaccines. Mapping of protective regions of CSP was performed using Escherichia coli -derived GST fusion protein vaccines containing the N-terminal (NT-CSP; residues Y 26 -P 104 ), repeat (Repeat-CSP; residues L 94 -N 284 ), and C-terminal region (CT-CSP; residues K 274- L 387 ) of CSP ( SI Appendix , Fig. S2 ). Repeat antigenicity and immunogenicity was optimized using TMV capsid proteins produced in E. coli . TMV particles displaying NPNAx3, -x4, -x5, -x7, -x10, or -x20 on an exposed loop (TMV-Loop) were produced. The NPNAx5 also displayed on the N terminal (TMV-NT) and C terminal (TMV-CT) of the TMV particle. Particles were characterized for purity, identity, endotoxin content, and presence of TMV-like particles was determined by electron microscopy. mAbs and Antigenicity. Fab sequences of mAb 580 ( 35 ), mAb 317 ( 29 ), and mAb CIS43 ( 31 ) were grafted onto human IgG1 Fc and the resulting full-length IgGs were produced in HEK293 cells. An enzyme linked immunosorbent assay (ELISA) was conducted with TMV repeat particles or FL-CSP coated on plates (100 ng per well) and concentration of mAb that resulted in OD = 1 was determined as a measure of mAb-binding potency. Mouse Immunogenicity and Challenge. C57BL/6 (H-2b) mice ( n = 10 per group) were used for serological comparisons. Functionality of antisera in mice was determined by challenge with Plasmodium berghei FL-CSP transgenic sporozoites ( 45 ). Protection was defined as an absence of parasites in the mouse blood ( 46 ). Vaccines were formulated in the oil-in-water adjuvant, AddaVax (InvivoGen) or in the ALFQ that contains the immune-modulators, 3D-PHAD and QS21 ( 43 ). Mice were vaccinated three times with 1.25 to 2.5 µg protein (kept constant within an experiment) at 3-wk intervals, intramuscularly in alternating thigh muscles. Serological assays used (NPNA) 6 peptide, FL-CSP protein, TMV particles with NPNAx3, and NPNAx20 repeats as plate antigens. ELISA end-point titer was defined as the serum dilution that gave an OD 415 = 1. An IgG2c/IgG1 subclass ELISA was also performed. Rhesus Immunogenicity. Rhesus monkeys of Indian origin ( n = 6 per group) were vaccinated three times at 3-wk intervals, with 20 µg TMV-NPNAx5, 40 µg TMV-NPNAx5, or 40 µg FL-CSP; all antigens were adjuvanted with 1 mL ALFQ. Three doses were given on days 0, 28, and 60, in alternating thigh muscles. Monkeys were bled 2 wk after each vaccination and then monthly up to 6 mo after the last vaccine dose. A subset of monkeys, the 20 µg TMV-NPNAx5 ( n = 3) and 40 µg TMV-NPNAx5 ( n = 2) group, and for the 40 µg FL-CSP group ( n = 2), were bled again at week 47 after the third dose to access durability. In another study, rhesus of Chinese origin were vaccinated three times with 40 µg TMV-NPNAx5 ( n = 6), 40 µg TMV-NPNAx20 ( n = 4), or 40 µg FL-CSP ( n = 6) and sera was analyzed 2 wk after third dose. Sera were analyzed for function using an inhibition of liver stage development assay (ILSDA) ( 47 ). Repeat-specific ELISA and avidity assays were performed by the International Malaria Serology Reference Center using human ELISA methodology, as previously described for the RTS,S/AS01, MAL071 clinical trial ( 28 ). Ethics Statement. Animal procedures were conducted in compliance with the Animal Welfare Act and other federal statutes and regulations relating to animals and experiments involving animals, and adhere to principles stated in the Guide for the Care and Use of Laboratory Animals , National Research Councl Publication, 2011 edition ( 48 ). Vaccines. Mapping of protective regions of CSP was performed using Escherichia coli -derived GST fusion protein vaccines containing the N-terminal (NT-CSP; residues Y 26 -P 104 ), repeat (Repeat-CSP; residues L 94 -N 284 ), and C-terminal region (CT-CSP; residues K 274- L 387 ) of CSP ( SI Appendix , Fig. S2 ). Repeat antigenicity and immunogenicity was optimized using TMV capsid proteins produced in E. coli . TMV particles displaying NPNAx3, -x4, -x5, -x7, -x10, or -x20 on an exposed loop (TMV-Loop) were produced. The NPNAx5 also displayed on the N terminal (TMV-NT) and C terminal (TMV-CT) of the TMV particle. Particles were characterized for purity, identity, endotoxin content, and presence of TMV-like particles was determined by electron microscopy. mAbs and Antigenicity. Fab sequences of mAb 580 ( 35 ), mAb 317 ( 29 ), and mAb CIS43 ( 31 ) were grafted onto human IgG1 Fc and the resulting full-length IgGs were produced in HEK293 cells. An enzyme linked immunosorbent assay (ELISA) was conducted with TMV repeat particles or FL-CSP coated on plates (100 ng per well) and concentration of mAb that resulted in OD = 1 was determined as a measure of mAb-binding potency. Mouse Immunogenicity and Challenge. C57BL/6 (H-2b) mice ( n = 10 per group) were used for serological comparisons. Functionality of antisera in mice was determined by challenge with Plasmodium berghei FL-CSP transgenic sporozoites ( 45 ). Protection was defined as an absence of parasites in the mouse blood ( 46 ). Vaccines were formulated in the oil-in-water adjuvant, AddaVax (InvivoGen) or in the ALFQ that contains the immune-modulators, 3D-PHAD and QS21 ( 43 ). Mice were vaccinated three times with 1.25 to 2.5 µg protein (kept constant within an experiment) at 3-wk intervals, intramuscularly in alternating thigh muscles. Serological assays used (NPNA) 6 peptide, FL-CSP protein, TMV particles with NPNAx3, and NPNAx20 repeats as plate antigens. ELISA end-point titer was defined as the serum dilution that gave an OD 415 = 1. An IgG2c/IgG1 subclass ELISA was also performed. Rhesus Immunogenicity. Rhesus monkeys of Indian origin ( n = 6 per group) were vaccinated three times at 3-wk intervals, with 20 µg TMV-NPNAx5, 40 µg TMV-NPNAx5, or 40 µg FL-CSP; all antigens were adjuvanted with 1 mL ALFQ. Three doses were given on days 0, 28, and 60, in alternating thigh muscles. Monkeys were bled 2 wk after each vaccination and then monthly up to 6 mo after the last vaccine dose. A subset of monkeys, the 20 µg TMV-NPNAx5 ( n = 3) and 40 µg TMV-NPNAx5 ( n = 2) group, and for the 40 µg FL-CSP group ( n = 2), were bled again at week 47 after the third dose to access durability. In another study, rhesus of Chinese origin were vaccinated three times with 40 µg TMV-NPNAx5 ( n = 6), 40 µg TMV-NPNAx20 ( n = 4), or 40 µg FL-CSP ( n = 6) and sera was analyzed 2 wk after third dose. Sera were analyzed for function using an inhibition of liver stage development assay (ILSDA) ( 47 ). Repeat-specific ELISA and avidity assays were performed by the International Malaria Serology Reference Center using human ELISA methodology, as previously described for the RTS,S/AS01, MAL071 clinical trial ( 28 ). Ethics Statement. Animal procedures were conducted in compliance with the Animal Welfare Act and other federal statutes and regulations relating to animals and experiments involving animals, and adhere to principles stated in the Guide for the Care and Use of Laboratory Animals , National Research Councl Publication, 2011 edition ( 48 ). Results Rationale for Repeat-Only CSP Vaccines. Regions of CSP that confer maximal protection in a transgenic mouse challenge model were mapped using subunit proteins produced in E. coli ( SI Appendix , Figs. S2 and S3 ). The CSP-repeat region, C-terminal region, and FL-CSP Abs reacted with the sporozoite surface by immunofluorescence assay, but N-terminal Abs were nonreactive ( SI Appendix , Fig. S3 B ). No ILSDA activity was observed for the C terminal; Abs to the N terminal showed midlevel inhibition (∼1 log reduction); and CSP repeat Abs showed the highest (∼2 log) reduction in parasite mRNA burden ( SI Appendix , Fig. S3 C ). While repeat region Abs had lower average avidity than the C-terminal Abs, sterile protection was only detected when vaccines contained the poly-NPNA CSP repeat region ( SI Appendix , Fig. S3 D and E ). Repeat-Only TMV Vaccine Design. The N and C termini of the TMV capsid point outwards from an assembled TMV disk ( Fig. 1 A ) and majority of previous vaccine design efforts have used these N and C termini to display epitopes ( 41 , 49 – 52 ). Dedeo et al. ( 53 ) reported a "circular permutant" of TMV where the N and C termini were re-engineered to the inner pore and a short loop was placed to close the sequence gap. We used this circular permutant of TMV to display the NPNA repeats on the exposed loop ( Fig. 1 B and C ). TMV circular permutants displaying various NPNA copy numbers from NPNAx3 to NPNAx20 were designed and predicted tertiary structures [Rosetta Protein Prediction Server ( 54 , 55 ); http://robetta.bakerlab.org ] illustrated the expected increasing conformational complexity and flexibility at higher copy numbers ( Fig. 2 A ). To determine the best location of the NPNAx5 antigen within the TMV monomer, we utilized the native fold of TMV to display NPNAx5 on the N terminus (TMV-NT) or the C terminus (TMV-CT). Overlaid predicted minimum energy structures illustrated that N- and C-terminal display on native TMV fold would also lead to increased antigen flexibility compared to the loop display (TMV-Loop) ( Fig. 2 B ). TMV vaccine antigens were >95% pure, reactive to a repeat-specific mouse mAb 2A10 ( 56 ) ( SI Appendix , Fig. S4 ), and they contained TMV-like disks and rods as visualized by electron microscopy ( Fig. 2 ). Fig. 1. Predicted TMV capsid structures. ( A ) The native fold of the TMV monomer displaying NPNAx3 (circled) at the N terminus. ( B ) Circular permutant of TMV, where the N and C termini have been repositioned to the pore and an exposed loop was used to display NPNAx3. ( C ) Model of a TMV disk showing the exposed loops with NPNAx3 (based on 3KML crystal structure). All structures were predicted using Rosetta prediction server and pictures generated using PyMOL software. Fig. 2. Design and production of TMV particles. ( A ) Rosetta structure predictions showing poly-NPNA of various lengths on the exposed loop (NPNAx3 circled). ( B ) Overlaid, minimum energy structures of NPNAx5 antigen (arrows) on the exposed loop of the circular permutant (Loop) or on the N terminal (NT) and C terminal (CT) of the native TMV fold, as predicted by Rosetta protein prediction server. Shown below each structure is the electron micrograph (40,000×) of the respective vaccine preparation. Antigenicity of TMV Vaccines. The ability of multiple CSP repeat-specific mAbs to bind to the TMV-NPNA constructs was determined. All three mAbs (580, CIS43, and 317) bound to FL-CSP was high potency (1,049 ng/mL for NPNAx20; Spearman's correlation 60% mice with high titer (>10 4 ) were protected compared to ∼10% protection at lower titers. Fig. 4. Optimization of immunogenicity of TMV vaccines/AddaVax in C57BL/6 mouse challenge model. Mice ( n = 10) were vaccinated thrice with 1.25 µg TMV-NPNAx3, -x4, -x5, or FL-CSP vaccines ( A – C ) or with 2.5 µg TMV-NPNAx5, -x7, -x10, and -x20 vaccines ( D – F ) or with 1.25 µg TMV-NPNAx5-Loop, -NT, -CT, or FL-CSP ( G – I ) using AddaVax adjuvant. All vaccines were given at 3-wk intervals ELISA and challenge was at 2 wk after the third dose. ( A , D , and G ) Geometric mean ELISA titer using (NANP) 6 peptide coat. ( B , E , and H ) Mean avidity index using either a repeat peptide coat ( Left ) or FL-CSP coat ( Right ). ( C , F , and I ) Survival curves of mice postchallenge. ( J and K ) Mean OD 415 ratio for NPNAx3/NPNAx20 coated plates for mouse sera tested at 1:8,000 dilution. Error bars are ±SEM; red symbols are protected mice and asterisks indicate the level of statistical significance. **** P 100,000 titer ( Fig. 6 A and SI Appendix , Fig. S6 F ) and the avidity index was >60% ( Fig. 6 B ). Sera had to be diluted 1:200 to discern differences between ILSDA activities between groups. At this serum dilution, >90% mean ILSDA activity was observed in both TMV-NPNAx5 groups compared to 90% ILSDA activity. However, after 12 wk, anti–FL-CSP ILSDA activity rapidly dropped to 10,000 and these Abs continued to show >80% ILSDA activity up to the 47 wk after the third dose time point examined ( Fig. 6 E ). This high level of parasite inhibitory activity was also confirmed using ILSDA on sera from individual monkeys ( SI Appendix , Fig. S8 ). In a subsequent rhesus study, 40 µg TMV-NPNAx5 ( n = 6), 40 µg TMV-NPNAx20 ( n = 4), or 40 µg FL-CSP ( n = 6) formulated in ALFQ were compared in Chinese-origin rhesus macaques ( Fig. 6 F – H ). Three doses of the vaccine were given 4 wk apart. Compared to the FL-CSP titer at 2 wk after the third dose, the repeat titers of TMV-NPNAx5 was 2.3× higher and that of TMV-NPNAx20 was 1.6× higher ( SI Appendix , Fig. S6 G ). There was no difference between the titer of TMV-NPNAx5 and TMV-NPNAx20 groups ( Fig. 6 F ). Unlike the results in Indian rhesus, the avidity difference between FL-CSP and TMV vaccines was not statistically significant in Chinese rhesus ( Fig. 6 G ). However, ILSDA activity at 1:300 serum dilution showed a clear superiority of the TMV vaccines over FL-CSP ( Fig. 6 H ). The data confirmed the equivalence of NPNAx5 and NPNAx20 vaccines and superiority over FL-CSP in the rhesus model. ELISA and avidity assays for rhesus sera were performed by the International Malaria Serology Reference Center using a standardized human repeat ELISA. The same assay was used to report the immunological end-points for RTS,S/AS01 as part of the MAL071 clinical trial ( 28 ). While a direct comparison of ELISA between rhesus and humans can be confounded by cross-reactivity differences in secondary Abs, the geometric mean of titer and mean avidity index for the standard dose of RTS,S/AS01 was shown as a red line to provide context to the nonhuman primate data reported here for the NPNAx5 vaccine ( Fig. 6 A – G ). Rationale for Repeat-Only CSP Vaccines. Regions of CSP that confer maximal protection in a transgenic mouse challenge model were mapped using subunit proteins produced in E. coli ( SI Appendix , Figs. S2 and S3 ). The CSP-repeat region, C-terminal region, and FL-CSP Abs reacted with the sporozoite surface by immunofluorescence assay, but N-terminal Abs were nonreactive ( SI Appendix , Fig. S3 B ). No ILSDA activity was observed for the C terminal; Abs to the N terminal showed midlevel inhibition (∼1 log reduction); and CSP repeat Abs showed the highest (∼2 log) reduction in parasite mRNA burden ( SI Appendix , Fig. S3 C ). While repeat region Abs had lower average avidity than the C-terminal Abs, sterile protection was only detected when vaccines contained the poly-NPNA CSP repeat region ( SI Appendix , Fig. S3 D and E ). Repeat-Only TMV Vaccine Design. The N and C termini of the TMV capsid point outwards from an assembled TMV disk ( Fig. 1 A ) and majority of previous vaccine design efforts have used these N and C termini to display epitopes ( 41 , 49 – 52 ). Dedeo et al. ( 53 ) reported a "circular permutant" of TMV where the N and C termini were re-engineered to the inner pore and a short loop was placed to close the sequence gap. We used this circular permutant of TMV to display the NPNA repeats on the exposed loop ( Fig. 1 B and C ). TMV circular permutants displaying various NPNA copy numbers from NPNAx3 to NPNAx20 were designed and predicted tertiary structures [Rosetta Protein Prediction Server ( 54 , 55 ); http://robetta.bakerlab.org ] illustrated the expected increasing conformational complexity and flexibility at higher copy numbers ( Fig. 2 A ). To determine the best location of the NPNAx5 antigen within the TMV monomer, we utilized the native fold of TMV to display NPNAx5 on the N terminus (TMV-NT) or the C terminus (TMV-CT). Overlaid predicted minimum energy structures illustrated that N- and C-terminal display on native TMV fold would also lead to increased antigen flexibility compared to the loop display (TMV-Loop) ( Fig. 2 B ). TMV vaccine antigens were >95% pure, reactive to a repeat-specific mouse mAb 2A10 ( 56 ) ( SI Appendix , Fig. S4 ), and they contained TMV-like disks and rods as visualized by electron microscopy ( Fig. 2 ). Fig. 1. Predicted TMV capsid structures. ( A ) The native fold of the TMV monomer displaying NPNAx3 (circled) at the N terminus. ( B ) Circular permutant of TMV, where the N and C termini have been repositioned to the pore and an exposed loop was used to display NPNAx3. ( C ) Model of a TMV disk showing the exposed loops with NPNAx3 (based on 3KML crystal structure). All structures were predicted using Rosetta prediction server and pictures generated using PyMOL software. Fig. 2. Design and production of TMV particles. ( A ) Rosetta structure predictions showing poly-NPNA of various lengths on the exposed loop (NPNAx3 circled). ( B ) Overlaid, minimum energy structures of NPNAx5 antigen (arrows) on the exposed loop of the circular permutant (Loop) or on the N terminal (NT) and C terminal (CT) of the native TMV fold, as predicted by Rosetta protein prediction server. Shown below each structure is the electron micrograph (40,000×) of the respective vaccine preparation. Antigenicity of TMV Vaccines. The ability of multiple CSP repeat-specific mAbs to bind to the TMV-NPNA constructs was determined. All three mAbs (580, CIS43, and 317) bound to FL-CSP was high potency (1,049 ng/mL for NPNAx20; Spearman's correlation 60% mice with high titer (>10 4 ) were protected compared to ∼10% protection at lower titers. Fig. 4. Optimization of immunogenicity of TMV vaccines/AddaVax in C57BL/6 mouse challenge model. Mice ( n = 10) were vaccinated thrice with 1.25 µg TMV-NPNAx3, -x4, -x5, or FL-CSP vaccines ( A – C ) or with 2.5 µg TMV-NPNAx5, -x7, -x10, and -x20 vaccines ( D – F ) or with 1.25 µg TMV-NPNAx5-Loop, -NT, -CT, or FL-CSP ( G – I ) using AddaVax adjuvant. All vaccines were given at 3-wk intervals ELISA and challenge was at 2 wk after the third dose. ( A , D , and G ) Geometric mean ELISA titer using (NANP) 6 peptide coat. ( B , E , and H ) Mean avidity index using either a repeat peptide coat ( Left ) or FL-CSP coat ( Right ). ( C , F , and I ) Survival curves of mice postchallenge. ( J and K ) Mean OD 415 ratio for NPNAx3/NPNAx20 coated plates for mouse sera tested at 1:8,000 dilution. Error bars are ±SEM; red symbols are protected mice and asterisks indicate the level of statistical significance. **** P 100,000 titer ( Fig. 6 A and SI Appendix , Fig. S6 F ) and the avidity index was >60% ( Fig. 6 B ). Sera had to be diluted 1:200 to discern differences between ILSDA activities between groups. At this serum dilution, >90% mean ILSDA activity was observed in both TMV-NPNAx5 groups compared to 90% ILSDA activity. However, after 12 wk, anti–FL-CSP ILSDA activity rapidly dropped to 10,000 and these Abs continued to show >80% ILSDA activity up to the 47 wk after the third dose time point examined ( Fig. 6 E ). This high level of parasite inhibitory activity was also confirmed using ILSDA on sera from individual monkeys ( SI Appendix , Fig. S8 ). In a subsequent rhesus study, 40 µg TMV-NPNAx5 ( n = 6), 40 µg TMV-NPNAx20 ( n = 4), or 40 µg FL-CSP ( n = 6) formulated in ALFQ were compared in Chinese-origin rhesus macaques ( Fig. 6 F – H ). Three doses of the vaccine were given 4 wk apart. Compared to the FL-CSP titer at 2 wk after the third dose, the repeat titers of TMV-NPNAx5 was 2.3× higher and that of TMV-NPNAx20 was 1.6× higher ( SI Appendix , Fig. S6 G ). There was no difference between the titer of TMV-NPNAx5 and TMV-NPNAx20 groups ( Fig. 6 F ). Unlike the results in Indian rhesus, the avidity difference between FL-CSP and TMV vaccines was not statistically significant in Chinese rhesus ( Fig. 6 G ). However, ILSDA activity at 1:300 serum dilution showed a clear superiority of the TMV vaccines over FL-CSP ( Fig. 6 H ). The data confirmed the equivalence of NPNAx5 and NPNAx20 vaccines and superiority over FL-CSP in the rhesus model. ELISA and avidity assays for rhesus sera were performed by the International Malaria Serology Reference Center using a standardized human repeat ELISA. The same assay was used to report the immunological end-points for RTS,S/AS01 as part of the MAL071 clinical trial ( 28 ). While a direct comparison of ELISA between rhesus and humans can be confounded by cross-reactivity differences in secondary Abs, the geometric mean of titer and mean avidity index for the standard dose of RTS,S/AS01 was shown as a red line to provide context to the nonhuman primate data reported here for the NPNAx5 vaccine ( Fig. 6 A – G ). Discussion Mathematical modeling reveals that the high antigenic complexity of CSP negatively impacts efficient affinity maturation of B cells ( 34 , 57 ). The 10× higher titer and higher avidity in mice of TMV-NPNAx5 displayed on a loop, compared to the FL-CSP and N- and C-terminally tethered NPNAx5, could be the result of reduced antigen complexity due to the diminished antigen flexibility. Furthermore, 17 monomers of TMV capsid displayed ∼85 NPNA repeats per NPNAx5 disk, potentially augmenting B cell cross-linking by particulate display ( 42 , 58 ). The greater longevity of the low-dose TMV-NPNAx5 group in the rhesus experiment is also possible evidence of further augmentation of the B cell response through lower antigen dosing ( Fig. 6 D ). Upon increasing the loop size from NPNAx3 to NPNAx5, antibody avidity was unchanged but the titer increased. The shorter NPNAx3 loop may be overly restricted, which is consistent with the low potency of mAbs 580 and CIS43 binding. Interestingly, there was no significant improvement of vaccine performance between NPNAx5 and NPNAx20 in mice or rhesus, and there was evidence to indicate that NPNAx20 antibodies had lower titer and avidity compared to NPNAx5 group in mice. It is possible that at NPNA copy number around 20, increases the proportion of B cells that undergo apoptosis due to hypercross-linking ( 59 – 61 ). Overall in the mouse model, the NPNAx5 antigen displayed on a loop consistently showed optimal immunogenicity and presumably represented a balance between flexibility and copy number. Across Plasmodium species, CSP repeat sequences allow the formation of short structured motifs within a flexible repeat region ( 40 , 62 ). For example, proline and alanine within the poly-NPNA have low side-chain entropy and the asparagine residue can hydrogen-bond with the peptide back-bone. Plasmodium vivax CSP repeats, GNQPGANGA or GQPAGDRA(A/D), also contain residues with low side-chain entropy (glycine, proline, and alanine) and residues that hydrogen bond to the main chain (glutamine, asparagine, and aspartate). The partially constrained structural flexibility of the CSP repeats may be an escape mechanism, activating a diverse repertoire of B cells, only a fraction of which develop into plasma cells that produce high affinity and functional antibodies. The structure of RTS,S is not known, but it contains 19 NANP repeats expressed as an N-terminal fusion to the hepatitis B particle, and while optimal copy number may vary between display platforms ( 25 , 27 ), our results suggest that lowering repeat copy number and looped presentation may improve immunogenicity and protection of the current vaccine. Vaccine design screening studies using in vivo immunogenicity and protection readouts can require a large number of animals. An in vitro screen using a panel of well-characterized mAbs could greatly facilitate this effort ( 63 ). The positive association of mAb 580 potency with copy number prompted us to hypothesize that diminished abundance of low-affinity epitopes in a vaccine could translate to elicitation of polyclonal antibodies of higher average avidity. However, this was not the case with TMV-NPNAx3, which showed a higher proportion of constrained conformation antibodies, but the mouse titers were suboptimal and avidity did not differ from NPNAx5. mAbs CIS43 or 580 potency for TMV-NPNAx3 was poor, suggesting a lack of the minimal protective epitope displayed, and poor mAb 580 potency could be due to the inability of the NPNAx3 loop to sustain cross-linking of a low-affinity antibody via homotypic interactions ( 57 , 64 ) or additive binding ( 65 ). TMV-NPNAx5 may also reflect the right balance between limiting low-affinity mAb epitope abundance and efficient binding to high-affinity CSP mAbs. It is possible that an excess of mAb 580 binding epitopes in NPNAx20 or FL-CSP as compared to NPNAx5, could have resulted in lower-avidity polyclonal antibodies observed in mice. Furthermore, the potency of mAb 580 for TMV-CT was threefold better than for TMV-Loop and the avidity of TMV-CT Abs was also lower than TMV-Loop, but no different from FL-CSP Abs. Taken together, these data give us a glimpse at potency of mAb binding positively or negatively predicting immunogenicity outcomes for CSP vaccines. A future study using a panel of neutralizing and nonneutralizing antibodies can downselect mAbs that definitively predict CSP vaccine outcomes. Access to RTS,S and sera from protected/nonprotected volunteers could greatly facilitate the development of more precise assays to screen for improved CSP vaccines, similar to those that led to improving the RSV vaccine ( 66 ). FL-CSP contains 19 copies of NPNA and yet the most striking discordance between high mAb potency and low vaccine performance was that for FL-CSP. One reason could be that FL-CSP contains competing epitopes from the flanking N- and C-terminal regions of CSP. The 277-amino acid long FL-CSP vaccination induces antibodies against epitopes spread all across the CSP molecule ( 43 ). In contrast, the TMV response was highly focused to a 20-amino acid target ( Fig. 5 D ). Repeat Ab titers and avidity has been associated with protection ( 28 , 67 ), but several studies show C-terminal mAbs have little neutralizing activity ( 4 ). Although polymorphisms within the C-terminal region were implicated in escape from RTS,S-mediated protection ( 6 ), our data showed no discernable antiparasitic effects of C-terminal Abs despite their high avidity ( SI Appendix , Fig. S3 ). Nonfunctional B cell epitopes in vaccines could compete for immunological space during the Ab induction and maturation phases ( 68 ) and nonfunctional Abs can sterically block functional Abs from binding to the target ( 69 ). Our results, and clinical experience with repeatless CSP vaccines ( 70 ), argue that a repeat-only, particulate CSP vaccine, combined with a molecular adjuvant like AS01/ALFQ, needs to be assessed in humans. A flat TMV disk could be sterically less restrictive than VLPs that require sphere closure. Up to an 80-amino acid insert (NPNAx20) was successfully displayed on TMV, but only a 20-amino acid insert (NPNAx5) may be required to protect against P. falciparum , thus leaving potential space within the TMV particle for additional vaccine epitopes, such as the junctional epitope ( 30 , 31 ) or the repeat region of P. vivax CSP. Expression and purification of TMV was accomplished here in E. coli ; however, an optimized TMV-based malaria vaccine could also be manufactured in planta . A 10× improvement in repeat titers for TMV-NPNAx5 over FL-CSP was observed in mice but protection in a rechallenge study was better for TMV-NPNAx20. In rhesus of Indian origin, TMV-NPNAX5 showed a 5× improvement in titer over FL-CSP but rhesus of Chinese origin showed a 2.3× difference. Unlike mice, the Chinese origin rhesus showed no difference between TMV-NPNAx5 and TMV-NPNAx20 vaccines, confirming our long-standing position that immunological comparison in rhesus remain on critical path prior to advancing second generation vaccine candidates to humans ( 44 , 71 , 72 ). TMV-NPNAx5/ALFQ antibodies in rhesus macaques inhibited sporozoite invasion for up to a year, leading us to hypothesize that an optimized epitope-based vaccine could protect against malaria for up to a year and may be sustained thereafter by an annual booster vaccination. Based on the evidence provided here, TMV-NPNAx5 and TMV-NPNAx20 can be considered as candidates for testing the first TMV-based vaccine in humans. Supplementary Material Supplementary File

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3148622/

Biological warfare agents

The recent bioterrorist attacks using anthrax spores have emphasized the need to detect and decontaminate critical facilities in the shortest possible time. There has been a remarkable progress in the detection, protection and decontamination of biological warfare agents as many instrumentation platforms and detection methodologies are developed and commissioned. Even then the threat of biological warfare agents and their use in bioterrorist attacks still remain a leading cause of global concern. Furthermore in the past decade there have been threats due to the emerging new diseases and also the re-emergence of old diseases and development of antimicrobial resistance and spread to new geographical regions. The preparedness against these agents need complete knowledge about the disease, better research and training facilities, diagnostic facilities and improved public health system. This review on the biological warfare agents will provide information on the biological warfare agents, their mode of transmission and spread and also the detection systems available to detect them. In addition the current information on the availability of commercially available and developing technologies against biological warfare agents has also been discussed. The risk that arise due to the use of these agents in warfare or bioterrorism related scenario can be mitigated with the availability of improved detection technologies. Anthrax Anthrax is a disease of humans and animals, caused by the bacterium Bacillus anthracis , a gram-negative, facultative, anaerobic, nonmotile and spore forming bacterium. It derives its name from the Greek word for coal "anthrakis" because it causes black coal like skin lesions. This disease has been known since antiquity and affects humans who tend animals or come in contact with animals or their infected products. The bacterium often penetrates the body via wounds in the skin and may also infect humans as aerosol or ingestion. The ability of sporulation and resistance of the spores to harsh environmental conditions like heat and humidity, disinfectants and UV radiation makes anthrax the most important biological warfare agent. It is reported that the spores can survive up to 40 years in water or soil. Spores are formed only in the presence of oxygen. Experiments involving detonation of bombs containing Ba. anthracis spores were conducted by US and its allies in the Gruinard Island during the period 1943. After the experiments, live spores were detected in the soil at various depths and took 40 years and tons of formaldehyde to disinfect the island and declare it free of anthrax spores.[ 5 ] Anthrax is prevalent in many parts of the world, and historical accounts have documented outbreaks of the disease in livestock and humans. Herbivores can become infected with anthrax by grazing in pastures that are contaminated with spores. The animals develop bacteremia and contaminate the environment with vegetative cells, which can subsequently sporulate and form spores. When the animal dies due to anthrax, the animal carcasses become infectious, and when the animal decomposes, the spores can again mix with the soil and the transmission cycle continues. Human anthrax naturally occurs in three clinical forms: cutaneous anthrax when spores come in contact with the skin and develop into black lesions, this occurs mostly with the handling of infected animals; gastrointestinal anthrax that occurs by the consumption of infected animal products and undercooked/raw meat, respiratory anthrax that is caused by the inhalation of spores through respiration. The virulence factors of Ba. anthracis consist of toxin and an antiphagocytic capsular polypeptide containing D -glutamic acid.[ 6 ] The toxin is composed of three proteins: protective antigen (PA), edema factor (EF) and lethal factor (LF). In India, the incidence of the disease in animals is extremely low and few districts of Tamil Nadu, Andhra Pradesh and Karnataka report cases annually and are considered endemic to this disease. Sporadic cases from humans are also reported from this region. Anthrax in all forms can lead to septicemia and death if untreated; early treatment of cutaneous anthrax is usually cured but in all other forms the recovery depends on the early detection and treatment. The case fatality rates range from 25 to 75% for gastrointestinal anthrax and can reach up to 90–100% in case of inhalation anthrax. The natural source of infection for humans, in naturally acquired disease, is infected livestock and wild animals or contaminated animal products. Human to human transmission is extremely rare and reported only with cutaneous anthrax. Cutaneous anthrax is the most common manifestation of infection with Ba. anthracis . Inhalation anthrax occurs in persons working in certain occupations where spores may be released into the air from contaminated animal products, such as wool industry. The occupation high-risk groups include those who domesticate animals, workers who come in contact with livestock, e.g., veterinarians, animal handlers, abattoir workers and laboratory technicians who handle infected specimens. Anthrax can be treated in its earliest stages with penicillin (often delivered intravenously), tetracycline, ciprofloxacin or doxycycline. Evidence of antibiotic effectiveness comes from use in human and animals infected with Ba. anthracis . Antibiotics related to ciprofloxacin including chloramphenicol, erythromycin and some cephalosporins were also found to be effective. The antibiotic treatment is highly effective in early stage of the disease and becomes largely ineffective in later stages with severe complications. Attenuated spore based vaccine is available for use in animals. Recently, recombinant proteins, mainly the recombinant protective antigen, are used effectively to protect human anthrax and are widely available in many countries. The drawbacks with the recombinant protein based vaccines are short protective efficacy, need of yearly booster doses, side effects and poor tolerance in individuals. Plague Plague is an infectious disease of animals and humans caused by the bacteria, Yersinia pestis .[ 7 ] Y. pestis is a gram-negative, nonmotile, facultative anaerobe, on Wayson or Giemsa staining shows a bipolar or safety pin-like appearance. It is a zoonotic infection involving the rodents and humans. Millions of people in Europe, Asia and African continents died of plague in the earlier centuries due to flea infested rats inhabiting human homes and ports. Several port towns and cities were devastated by plague outbreak and epidemics used to spread across countries and continents. Today, the modern antibiotics and better sanitary conditions have effectively reduced the natural outbreaks of plague although small numbers of cases are reported from rural communities. World Health Organization (WHO) annually reports around 1000–3000 cases of plague from different countries around the world. In India, active plague foci are known to exist in Gujarat, Himachal Pradesh, Maharashtra, Karnataka and Tamil Nadu. Humans are infected from plague from the bite of rodent flea that carries the plague bacteria or by handling an infected animal. In the natural infection involving rodent flea bites, a person develops bubos or swelling of lymph nodes in groin and underarms. Further, after the appearance of bubos, the hematogenous spread of bacteria results in septicemia and infection of lungs and results in pneumonic plague. In a biological warfare or bioterrorist event, the plague bacilli are inhaled as aerosol and directly result in the pneumonic plague without the involvement of bubonic plague. The pneumonic plague is highly contagious and spreads from one individual to other by airborne droplets. The incubation period of conventional plague is 2–10 days, which may be considerably reduced to 8–12 hours in the case of pneumonic plague. The clinical sign of plague is a very painful, usually swollen and often hot to touch lymph node called a bubo. The bubos are accompanied by fever, extreme exhaustion and a history of possible exposure to rodents, rodent fleas, wild rabbits, sick or dead carnivores. Initial clinical manifestations include fever, headache, and general illness, followed by development of bubos. The disease progresses rapidly and the bacteria invade the blood stream, producing severe illness and further develop into overwhelming pneumonia with high fever, cough bloody sputum and chills. In untreated cases, the mortality rate can reach more than 50%. Brucellosis Brucellosis is one of the world's major zoonosis, caused by bacteria of the genus Brucella and can be transferred from animals to human. Brucellosis is a an important human disease in many parts of the world, especially in the Mediterranean countries of Europe, North and East Africa, Middle East, South and Central Asia, Central and South America. It is a serious public health problem in India as well as in other developing countries. In India, it is a neglected disease and closely associated with the agrarian society and mostly linked to the animal husbandry. The genus Brucella consists of eight species, designated on the basis of host preference, biochemical and molecular characteristics as follows: Brucella melitensis (goat and sheep), Brucella abortus (cattle), Brucella suis (pigs, reindeer), Brucella canis (dogs), Brucella ovis (sheep), Brucella neotomae (desert wood rats), Brucella pinnipediae (seal/otter) and Brucella cetaceae (porpoise/whale).[ 8 ] Br. pinnipediae and Br. cetaceae have been reported from marine mammals. Recently, one more species of Brucella , namely, Brucella mocroti has been isolated from common voles, Microtus arvalis .[ 9 ] The disease has been known by several terms, including Malta fever, undulant fever, Gibraltar fever, Bang's disease typhomalarial fever due to its resemblance to malaria and typhoid fevers. The causative organisms are gram-negative, facultative intracellular bacteria in the shape of cocci, coccobacilli or short rods, 0.5–0.7 μ m in diameter and 0.6–1.5 μ m in length. It is a partially acid fast bacterium that lacks endospores or native plasmids. Optimal pH condition for growth of this bacterium is between 6.6 and 7.4. All Brucella strains lose viability when incubated at 56°C; however, a minimum of 1 hour incubation at a temperature over 70°C may be required to ensure complete inactivation. It is a systemic infection that can involve any organ or organ system of the body; symptoms are nonspecific, generally occurring within 2–3 weeks of infection. It affects people of all age groups and sex. The infection in humans usually manifests itself as an acute febrile illness which may persist and progress to a chronically incapacitating disease with severe complications. The transmission of Brucella infection and its prevalence in a region depends upon several factors like food habits, methods of processing milk and milk products, social customs, husbandry practices, climatic conditions, socioeconomic status and environment hygiene. The disease has been recognized as one of the most common laboratory acquired infections; it has been reported to occur in clinical research and production laboratories. Brucella spp. is also considered as important biothreat agent and is characterized as category B pathogens in the bacterial list of biological warfare agents. Its significance as a potential agent of bioterrorism was acknowledged early, and the pathogen remains on the category B bio-defence research lists of both Center for Disease Control and Prevention (CDC) and the National Institute of Allergy and Infectious Diseases (NIAID).[ 10 ] Human brucellosis is a multi-system disease that may present with a broad spectrum of clinical manifestations and diagnosis requires microbiological confirmations by means of isolation from blood culture or demonstration of the presence of specific antibodies by serological tests. Blood culture provides definite proof of brucellosis but may not provide a positive result for all patients even under ideal conditions. Brucella is a slow growing organism and cultures are rarely positive and should be kept at least 45 days before the culture can be concluded negative. Serological tests are used in the diagnosis of brucellosis, which can be classified depending on the antigens used, as conventional tests (i.e., those using suspensions of whole cells as antigens) and tests using antigenic extracts (i.e., outer membrane proteins, cytoplasmic and periplasmic proteins). Outer membrane proteins (OMPs) are structural constituents of the cell and are unlikely to function as virulence factors. They induce immune responses useful in diagnosis[ 11 ] and in conferring protective immunity. Glanders Burkholderia mallei is the etiological agent of glanders, a disease of equines, i.e., horses, mules and donkeys. Bu. mallei is a nonmotile, aerobic gram-negative coccobacillus, which may or may not be oxidase positive and grow on Macconkey Agar. During World War I, the German military used this agent as a biological weapon against horses and mules that were primary forms of transport during the war.[ 2 ] The disease has been eradicated from most part of the developed world due to compulsory slaughter of infected or seropositive horses and mules. Reports on naturally acquired cases are mostly those of infections acquired from laboratory and many cases of laboratory acquired human glanders have been reported. Equines are the primary reservoirs of this rare illness and natural infection has been reported from eastern Europe, Middle East, Asia including India and Africa. The disease can be present as either a cutaneous or systemic disease. The incubation period ranges from 1 to 14 days depending on the virulence of the strain. Patients with cutaneous infection have nodules with accompanying localized lymphadenitis. The systemic illness usually manifests itself either as brancho or lobar pneumonia. Bacteremia may also occur, resulting in lesions being seen in liver and spleen. Infection in humans with Bu. mallei is often fatal without antimicrobial therapy. There is little or no clinical experience in the treatment of Bu. mallei infection in humans with the modern antimicrobial agents due to the rareness of this disease. The antimicrobial susceptibility studies with various strains and limited animal experiments suggest that ciprofloxacin ad doxycycline could be used for prophylaxis.[ 12 ] No vaccine against glanders is available for human use. Melioidosis Burkholderia pseudomallei is a recognized biothreat agent and the causative agent of melioidosis in animals and humans.[ 13 ] Melioidosis is prevalent in Southeast Asia, Northern Australia, and some other tropical and subtropical regions and also endemic in few states of India (Kerala, Maharashtra, Tamil Nadu, Orissa, Assam, West Bengal and Tripura). Bu. pseudomallei is an oxidase-positive, aerobic gram-negative bacillus that is straight or slightly curved. The organism grows on most standard laboratory media, such as sheep blood and chocolate and MacConkey agars, and it produces a characteristic musty odor. It is a resilient bacterium that can survive in disinfectant and antiseptic solution and extreme temperatures and grows very slowly in salt medium. The presence of Bu. pseudomallei in soil is an important factor in determining disease incidence in a given endemic area. This organism was first reported as causing human infections in 1911 by Whitmore from individuals living in Rangoon, Myanmar, and in earlier medical literature was called Whitmore's Disease. The diagnosis of this disease is very important for effective treatment and the populations at risk within endemic areas rarely have access to appropriate health care. Bu. pseudomallei is a natural inhabitant of soil and water in the tropics and subtropics, but can also survive in dry atmospheric conditions. It is present in rubber plantations, cleared fields, cultivated and irrigated agricultural sites as well as drains and ditches. There are several established modes of transmission with the patient population. The possible modes are ingestion of contaminated drinking water, inhalation and inoculation through the skin lesions from the contaminated soil. The rural poor and agricultural workers are likely to be at greatest risk of infection. Person to person transmission of Bu. pseudomallei , especially between patient and siblings or one of their playmates, is common; vertical transmission from mother to child is also possible. It can be transmitted by direct contact with infected rodents or infected food, soil, water, excreta; transmission is also possible through use of contaminated injection needle. Factors that may influence the distribution of Bu. pseudomallei in the environment may include physical factors such as rainfall, humidity, UV radiation, temperature, soil composition, vegetation, use of fertilizers, soil disturbance such as excavation and ploughing. The disease prevalence in nature has a significant seasonal and a strong linear correlation with rainfall. It is recognized as the most common cause of severe community-acquired sepsis in parts of northeast Thailand, where melioidosis accounts for 20% of community-acquired bacteremias and associated with a mortality of approximately 50%.[ 14 ] The clinical symptoms of melioidosis are so varied that the disease has been termed "the great mimicker". The clinical manifestation of melioidosis can be divided into asymptomatic, chronic, acute and subacute infections.[ 15 ] Chronic melioidosis is characterized by osteomyelitis (inflammation of the bone) and pus-filled abscesses in the skin, lungs or other organs. Acute melioidosis takes one of three forms: a localized skin infection that may spread to nearby lymph nodes; an infection of the lungs associated with high fever (102°F/38.9°C), headache, chest pain, coughing, septicemia characterized by disorientation, difficulty in breathing, severe headache, and an eruption of pimples on the head or trunk; the third form is most common among drug addicts and may be rapidly fatal. The clinical manifestation of melioidosis may be observed as in apparent infection, asymptomatic pulmonary infiltration, acute localized supportive infection, acute pulmonary infection, acute septicemic infection or chronic supportive infection. The highest mortality occurs in patients with the septicemic form of melioidosis, which is characterized by dissemination of the bacteria from blood to various organs. The clinical course of septicemic melioidosis is associated with liver and spleen and often includes rapid deterioration and death that occur within a few days after hospitalization. The host determinants such as diabetes mellitus, chronic alcoholism and renal failure represent Bu. pseudomallei . Diversity of the clinical manifestations necessitates the isolation and identification of causative organism for a definitive diagnosis of melioidosis. The causative organism, Bu. pseudomallei , is amenable to antibiotic treatment. For acute or chronic infections, parenteral administration of imipenem or ceftazidime for 2–4 weeks, followed by oral therapy with amoxicillin, clavulanate or a combination of doxycycline and trimethoprim-sulfamethoxazole for 3–6 months is recommended. Due to development of resistance with ceftazidime therapy, combination therapy is usually recommended for initial treatment. Bu. pseudomallei organisms are frequently resistant to aminoglycosides, first and second generation cephalosporins, and fluoroquinolone antimicrobials in vitro , and so, drugs in these classes are not likely to be good prophylactic agents. There are no data published on prophylactic agents for Bu. pseudomallei although it is likely that the agents used for oral therapy would be useful prophylactically. Botulism Botulism is a rare, naturally occurring disease caused by the bacteria Clostridium botulinum . It produces seven immunologically distinct toxins, which are designated by the letters A–G. Similar toxins are also produced by other closely related bacterial species, Clostridium baratii and Clostridium butyricum . These toxins can produce clinical syndromes such as food borne botulism (as intoxication), wound botulism (infection and toxin production) and botulism caused by toxin production after clostridial colonization of the intestine of infants (infant botulism) or older children and adults.[ 16 ] These toxins spread through the blood stream and affect the neuromuscular junctions, where they inhibit the release of acetylcholine. Their action on the cholinergic system at the presynaptic motor neuron terminal by blocking acetylcholine transmission across the neuromuscular junction causes neuromuscular blockade, resulting in flaccid paralysis. The toxins also affect the adrenergic system without any significant consequences. The lethal dose of the botulinum toxin is very low and is also the most potent toxin known to mankind; in humans, the exact dose is calculated based on the extrapolations made from the primate experiments. The lethal dose for the crystalline purified toxin type A is 0.009–0.01 mg for a man of 70 kg, when introduced intravenously; 0.80–0.90 mg when injected intravenously; 0.80–0.90 mg when introduced inhalationally; and 70 mg when introduced orally.[ 17 ] The toxins are highly heat sensitive and most of the toxins get inactivated by heating to 85°C for 5 minutes. C. botulinum forms spores under stress and these spores can survive standard cooking and food processing procedures. However, the conditions that can allow the spores to germinate, i.e., anaerobic condition, nonacidic pH, low salt and sugar content, are not commonly encountered in the food and may be the reason for few food borne botulism cases. Home canned foods have long been associated with botulism and actual incidence of botulism is not available in most of the countries. In case of the food borne botulism, the presumptive evidence can be obtained by collecting 3–5 day food history from the patient, evidence of consumption of home canned food, as the preformed toxins are ingested in the food borne infection, and the patient rapidly progresses to respiratory failure and mechanical ventilation despite administration of supportive and specific therapy. Wound botulism is caused by the contamination of wound with C. botulinum spores from the environment and further germination of the spore and production of toxin in the anaerobic condition prevalent in the wounds. This type of botulism is very common among the drug addicts who inject black tar heroin, a specific preparation of heroin into tissues. Infant botulism results from the adsorption of the toxin produced by the bacteria that have colonized the intestine of infants aged below 1 year. The main reason for the colonization of infant intestine was suspected to be use of contaminated honey. The clinical presentation resembles that of adult, with inability to suck and swallow, weakened voice, ptosis and floppy neck and it progresses to generalized flaccidity and respiratory compromise. A licensed vaccine with human source antitoxin reduces the median hospitalization from 6 to 3 weeks; speedy recovery is possible with intensive care and antitoxin therapy. The adult intestinal botulism is produced from adsorption of toxin produced in situ by rarely occurring intestinal colonization. The patients have anatomical or functional bowel abnormality or due to using antimicrobials, which may permit protection of fastidious C. botulinum from competitive normal bowel microbial flora. The botulinum toxin is also used for cosmetic or therapeutic purpose, and the dose of toxin used is very low that it cannot cause systemic disease. The licensed toxins contain very low concentration of toxin, which cannot produce any symptoms. With respect to inhalation botulism, only one case was reported among German laboratory workers handling this bacterial toxin in 1962, and mostly the symptoms resemble those of food borne botulism. The deliberate dissemination of botulism is possible through aerosol route and can produce outbreak of inhalation botulism. In the event of a bioterrorist attack, the most likely event is the contamination of food and water source with the purified or partially purified botulinum toxin. In the earlier days, the mortality due to botulism was very high even after the administration of equine antitoxin. The current mortality rates of 3–5% in natural infection is mostly attributed to the use of highly effective modern intensive care techniques, mainly the mechanical ventilation, and most of the deaths result from respiratory failure. The only specific treatment available for botulism is antitoxin. The antitoxin can arrest the progression of paralysis and decrease the duration of paralysis and needs good ventilation support system. The ideal antitoxin treatment should start within 24 hours of onset of symptoms and the antitoxin binds only to the toxin molecule that has not reached the nerve endings. There are many adverse effects like anaphylaxis, hypersensitivity reaction and serum sickness associated with the use of antitoxin. The sensitivity of the serum or antitoxin should be ascertained before administration; one vial of botulism antitoxin produces serum levels of toxin type specific antibodies capable of neutralizing serum toxin concentrations many fold in excess of those reported in the patients. When large numbers of persons are infected simultaneously from single food source, it has been observed that the appearance of symptoms and the susceptibility to the toxin varies with the host response and the concentration of toxin ingested. The botulinum toxin cannot be absorbed through intact skin, but can be absorbed through mucosal surfaces, eye and nonintact skin. Person to person transmission has never been reported for these bacteria and no licensed vaccine is available for human use. In laboratories equipped and capable of diagnosing botulism, serum and other body fluids are examined for the presence of toxin. The most commonly used method for toxin detection is enzyme-linked immunosorbent assay (ELISA) and mouse inoculation bioassay. The vomitus, stool or debrided tissue from infected wounds are subjected to culture for the isolation of these bacteria and the isolates are further confirmed based on biochemical and molecular characters and tested for the protection of toxin. Variola Virus The smallpox (variola) virus is the largest of the animal viruses. The virus particles are brick-shaped to ovoid and morphologically indistinguishable from the less pathogenic, closely related vaccinia virus. The variola virus contains double-stranded DNA and has a complex structure. Two lipoprotein membrane layers surround the dumbbell-shaped nucleoid. The nucleoid is embedded in an ellipsoid body, forming the thick center of the virion and double membrane surrounds the virus particle. The variola virus is highly contagious and very virulent, with a case fatality rate of 30% in unvaccinated persons. As extreme biohazards involved in working with the smallpox virus have limited research on this organism, the WHO decided to destroy the virus and the virus has been stored only in two places, one each in the United States and former Soviet Union, for research purpose, if need arises in the future.[ 18 ] Poxviruses are divided into four different groups. Group 1 comprises variola, vaccinia, cowpox, ectromelia, rabbitpox and monkeypox viruses. The variola virus exists as one of these two strains: variola major causing severe smallpox and variola minor causing mild smallpox or alastrim. Both these strains cannot be differentiated based on immunologic methods and vary only in the clinical presentation. Vaccination against smallpox is performed with the vaccinia virus, which has many antigenic structures in common with the smallpox agent. It was used worldwide in a live vaccine against smallpox and served as a laboratory model for the poxviruses. The origin of the vaccinia virus remains unclear. It is different from Jenner's cowpox virus and may be a mutant of the variola and alastrim viruses. At least seven distinct major variola virus antigens can be recognized by immunodiffusion techniques, and 17 polypeptide chains can be identified. Hemagglutinating, complement-fixing, and neutralizing antibodies may be produced in response to antigens of the smallpox virus. Neutralizing antibodies are directed against two antigens in the surface membrane of the virus particle. Complement-fixing antibodies react with a family antigen common to each subgroup of the unclassified poxviruses. The incubation period is 12 days and the clinical presentation is fever and headache. The infection causes pus-filled vesicles throughout the body, and the mortality rate is around 30% in the nonimmunized group with a mortality rate of about 3% in the immunized group.[ 19 ] Today, smallpox is considered a serious threat due to several reasons, the most important being that most of the human population is not immunized against this disease and susceptible to smallpox as the universal vaccination ended in 1980. Second, the spread of smallpox is rapid due to its high infectivity and the mode of spread through aerosol route can be used for the intentional spread of this disease.[ 4 ] Detection The use of anthrax spores in letters in a bioterrorist event in 2001 has emphasized the need for immediate detection and identification of biothreat agents from environmental samples as well as from affected persons.[ 20 ] The technologies so developed need to be rapid, accurate and have to unequivocally confirm the presence of these agents in varied matrices. An ideal detection system for a biothreat agent should not only be capable of detecting the agents in very low concentration, but also have the possibility to detect in various matrices. In addition, it should be portable, user-friendly and capable of detecting multiple threat agents. Among all the detection systems available for use, none of them satisfy all these criteria and the selection of methodology should be situation specific. Development of detection systems that can detect the biological agents in concentration at which they can cause disease in humans is a challenge, and due to lack of sensitivity of many of the available antigen and antibody based systems, research is focused on development of nucleic acid based sensors that are much more sensitive but need complex sample preparation. The Polymerase Chain Reaction (PCR) based assays can detect >10 microorganisms per sample but limitations are with the detection of toxins and other non nucleic acid containing prions, etc. Further, the nucleic acid based assays are to be performed in much cleaner environment as there is a possibility of DNA from laboratory and instrument contamination getting amplified and thus producing a false positive result. The immunologic assays based on the antigen and antibodies are highly specific due to the specificity of the antigen and antibody used but their specificity is low when compared with other assays. In addition to the sensitivity and specificity, the reproducibility of detection system or assay is very important as the data generated need to be reproducible under different environmental conditions. The reproducibility is decided by the stability and consistency of the reagents and assay conditions. The ideal detection system should have the capability to detect multiple threat agents as we expect the sample to contain a variety of bacteria, viruses and toxins. In the recent times, genetically modified bacterial and viral agents not covered under the list of probable agents may pose an additional threat and make the detection much more difficult. In the scenario of detection of biological warfare agents, we expect not only human clinical samples like blood, sputum, urine, stool, CSF, etc., but also suspected powdery material, food and water samples and, most importantly, the environmental air sample. The pollutants present in the air, the anticoagulants, leukocyte DNA and blood heme compounds are known to inhibit the polymerase reaction. In the case of immunoassay, the food samples with high protein and lipid content and high bacterial load in the stool samples pose a challenge in the detection. Due to this reason, a good sample preparation is needed in most of the systems that have been developed with either the nucleic acid or antibody based detection systems. The sample preparation takes hours to days depending on the standardized protocols and often cannot be performed in the field conditions. In the case of bacteria, most of the time the sample that is obtained from field condition is not viable and is unsuitable for culture conditions.[ 21 ] The sample collection procedures, handling, transportation and preparation are very important and vital for correct identification. The air monitoring requires concentration of bacteria from large air volumes and resuspension of concentrated material into appropriate liquid as most of the detection systems currently developed can use only liquid samples. Although the sample preparation and efficacy of extraction procedures determine the concentration of the agent availability for detection, in some instances, the viability of virus or bacteria has to be confirmed by conventional culture methods. The conventional detection and identification of human pathogens mostly rely on cultivation on bacteriological media and biochemical tests that take minimum 3–7 days and in some cases even more than 15 days to provide results. The culture and biochemical tests are reliable but need highly trained manpower to perform. Moreover they do not provide results on real time basis and as such cannot be used in environmental detection or in situations that need immediate results. Efforts are currently on to develop methods that are rapid, accurate and can detect the agents in field as well as in laboratories that have limited or no instrumental support. The detection technologies can be broadly classified into biochemical test based assays, antibody based assays and nucleic acid based assay. Most of these assays are at various stages of development and test on limited scale, their use in case of emergency situation need to be further evaluated on a case to case basis. Biochemical test based assays The culture of bacteria in a routine microbiological laboratory and further identification based on biochemical tests is the conventional identification method mostly followed in most of the hospital setups even today. As this method is followed for various diseases, the biological warfare agents, mainly the bacterial agents, need to be identified through this platform on a routine basis, if not during an intentional outbreak, but during natural prevalence of the disease. Most of the laboratories have standardized media for isolation for Ba. anthracis, Y. pestis, Burkholderia sp. and Brucella sp., and growth in specific media followed by few of the conventional biochemical tests can lead to the presumptive identification of these agents. Many commercial automated biochemical test platforms are today available for performing the tests after growth in bacteriological media, but these systems are mostly developed for bacterial diseases that are clinically important. These databases do contain data of important bacterial biological warfare agents and can be used in case of emergency. The tests based on substrate utilization pattern have been developed by Becton Dickinson (USA), Vitek (BioMe'rieux, France) and Microlog (Biolog, Hayward, CA USA). These systems have both visual observation and reader based formats for analysis of results and a database for comparison purposes. The main disadvantage of these systems is the need for pure culture and trained manpower to perform and for all the analyses. The BioMe'rieux API series of cards containing specific substrates have been used with good results in identification of most of the bacterial agents.[ 22 23 ] In addition to these tests, the system developed by MIDI Inc. (USA), called the microbial identification system, coverts cellular fatty acids from pure culture of bacteria to fatty acid methyl esters and uses a gas chromatograph for the separation and identification. In this method, a tedious sample preparation step in involved in which the bacterial cells need to be harvested, saponified, methylated, extracted and washed before analysis. But a chromatographic result pattern allows signature storage and database based retrieval for easy computing and analysis. Many of the databases based on the fatty acid profile have been now routinely used in laboratories for the confirmation of bacterial isolates. The major disadvantage is the need of time consuming isolation procedures as well as the need of highly calibrated instruments for reproducible performance. The advantage of these systems is that they allow simultaneous analysis of many samples and generate a highly reliable result output in the form of a fingerprint. They also allow addition of additional input based on experience.[ 24 ] Bioluminescence based detection Bioluminescence is the monitoring of luciferin and luciferase, the enzyme, interaction in the presence of ATP. The basic principle is ATP is found in all the living cells like virus and bacteria and its amount corresponds to the microbial load in air, water and in other environmental samples.[ 25 ] The bioluminescence has been widely used mainly in the quality control testing of bacterial contamination in food industry. This principle has been used in the Profile-1 hand held system developed by New Horizon Diagnostics (Columbia, MD, USA), which uses a microluminometer to read the sample bioluminescence.[ 26 ] In the Filtravette system developed by New Horizon Diagnostics, care has been taken to remove non-bacterial ATP. The main disadvantage of the bioluminescence based systems is the ATP contamination from nonmicrobial sources. Moreover, as all of them detect ATP, the results are specific and cannot be pointed toward any biowarfare agents. Since these systems are cost effective can be used for the real time air monitoring as first line defence. The continous air quality monitoring will trigger alarm for any unusual raise in the microbial load in the environment. Antigen and antibody based detection systems The antigen and antibody based immunoassays are being developed for the detection of bacterial and viral biowarfare agents.[ 27 ] The immunologic detection of antigens or antibodies based on ELISA has been developed, evaluated and currently in use for the detection of anthrax, plague, botulism, brucellosis, glanders and melioidosis. In these immunoassays, the efficacy of the detection depends on the availability of antigen or antibody in the detection matrix, the quality of the antigen that needs to be detected and the quality of the antibodies used in the detection of the antigen.[ 28 ] In short, it depends upon the quality of the antigen and antibody complex formed and the ability of the detection method to detect the formed antigen and antibody complex. Many different formats employing the ELISA principle have been developed using different substrate labels, i.e., fluorescent, chemiluminescent, electrochemiluminescent and on various platforms like solid support ELISA plates, visual dot and lateral flow formats. In addition, instruments using the immunoassay principles like biosensors, flow cytometry and micro array are under different stages of development. The disadvantage of the ELISA based systems is its ability to perform detection of only one agent at a time and sequential assays need to be performed to detect more than one agent. The specificity of the immunoassays limits the detection to minimum 10 5 CFU per test and is considered much higher than the DNA based assay. The immunoassay based on the sandwich antigen capture assay format is used in the Luminex xMAP system developed by Luminex Corp , Austin, Tx,USA, in which the polystyrene beads are coated with the specific antibodies and these beads are spectrally unique and are color coded into different sets; as each of these beads is coated with antibodies specific to an agent, it allows simultaneous detection of multiple agents in a liquid phase detection.[ 29 ] The laser in the detector detects the excitation of internal bead dyes of any capture agent and the signal intensity is proportional to the amount to agent present in the sample. This system allows to be hooked to a continuous environmental monitoring system for the detection and online monitoring. In another instrument the BV M-Series developed by BioVersis Corp. (USA), the reporter molecule uses electrochemiluminescence using paramagenetic beads as the support for the coating the capture antibodies. The electromagnetic beads are separated from the matrix by passing through a magnetic field and detected; this system has been modified as a kit for portable detection.[ 30 ] The bio-detector of Smiths Detection (USA) uses ELISA principle in a tape format and in which liquid samples are injected. In the reaction, the antigen available in the sample is mixed with biotin labeled and fluorescenin labeled antibodies. The system has a sensor that detects the signal from the biotin and fluorescenin labeled antibodies. Dissociation enhanced lanthanide fluorescence immunoassay (DELFIA) developed by Perkin Elmer Life Science (USA) is a unique system based on time resolved fluorescence.[ 31 ] This technique relies on lanthanide chelate labels which have long fluorescence decay time that allows measurements without background signals. Simply, in this system, the antibodies are labeled with lanthanide and tests are performed on standard plate ELISA format. After the antigen and antibody complex is formed, the label is disassociated from the antibody using a low pH enhancement solution and the free molecules rapidly form new stable highly fluorescent chelates that can be read by the system. The latest system that has gained popularity in the ELISA format is the lateral flow system which is rapid and cost effective when compared with the instrument-based detection systems. These tests are based on single use, disposable cartridge tests in the form immunochromatographic (ICT) line assays that generate visual lines in the membrane. In this class of assay, the detector antibodies are coated with colloidal gold or micro particles which by forming an antigen and antibody complex form a visual line or dot. These tests incorporate a procedural control that acts as a negative control and gives a confidence of testing while performing them.[ 32 ] These tests are easy to perform and are rapid than most of the immunoassays, but the sensitivity is often questioned as high false positive results are encountered. This can be used as a presumptive test and can be further followed up with a confirmative lab-based test. Lateral flows have been developed for all the biothreat agents by different agencies, but information on the use in real time detection situation and comparative evaluation with other methods is not available for recommendation. Nucleic acid based detection The nucleic acids, DNA and RNA, based detection systems is the most researched and developed detection system for the detection of biowarfare and biothreat agents. In the conventional PCR, the specific region of the genome is amplified and checked on electrophoresis for the amplification of correct size of product. In the advance quantitative real time PCR or Q PCR assays, the PCR amplification is combined with real time detection based on reporter fluorescence dyes.[ 33 34 ] In the non specific Q-PCR, the amplified DNA is detected based on DNA intercalating fluorescence dyes (SYBR green), and in specific Q-PCR, the fluorescence probes that specifically bind to a sequence are used. In the biowarfare agent detection, the specific Q-PCR assays based on probes have been developed for all the agents. In all the Q-PCR assays, software monitors the progress of the reaction, and the presence of the agent can be detected online on a monitor and the data can be transferred over a long distance for action. Today, compact, faster and sensitive real time PCR systems are available, making this method the most preferred method of agent detection. The limitations of this assay are the variation that exists with the nucleic acid based amplification techniques, the availability of starting material and its quality, inhibitory substances present in the matrices, sensitivity and specificity based issues related to quality of primers, probes, enzymes used. The advantage being the ability to detect a few cells, mostly up to 100 cells per test, makes it an attractive alternative to all the detection systems available today. The portability and stability of the reagents, use of stored power supply, etc., will make this technology more practical in the years to come. The developers based on the Q-PCR have incorporated disposable testing cartridges for various types of samples like the GeneXpert Q-PCR of Cepheid (USA) and eliminating the need to extract DNA prior to processing makes them convenient to use in field situations. The light cycler system of Roche Diagnostic, GmbH, Mannheim, Germany has assays specific to Ba. anthracis , which use a glass capillary into which extracted DNA is placed instead of a cartridge, and is commercially available. The Applied Biosystems (ABI, Foster City, USA) has several Q-PCR kits developed for anthrax and other biothreat agents, which can be used with their ABI 7500 and 7300 real time PCR systems. In case of the Bioseeq developed by Smiths Detection (USA), all the reagents are freeze dried for better stability and storage and this commercial hand held version can analyze up to six agents. Several other firms have also started marketing readymade kits for use with various real time PCR platforms for the detection of biowarfare agents, but no comparative results are available for these kits for evaluation. Sensor based detection systems The sensor based detection systems have been developed based on all the three groups that have been discussed already. The sensors have been developed for biochemical, immunologic and nucleic acid based detection systems. In the sensor based detection systems, the basic principles of all these three groups are integrated with a transducer than can transform the response in to an analyzable signal.[ 35 ] All these systems that integrate the biological component with a physical transducer are termed as biosensors. The physical transducers that produce an analyzable signal could be of electrochemical, optical, mass, thermal or of high frequency. In the electrochemical systems, the products that are formed due to an enzymatic activity associated with microbial metabolism are monitored based on conducting polymers. Array of sensors, each specific for different vapors or gases, are constructed and used to detect multiple analytes. Mostly, this technique is used to detect volatile organic compounds produced during the growth of bacteria or fungi and is termed as electronic noses. This system is very complex as most of the compounds are produced by multiple bacteria and analysis of results is difficult. But this has application in mainly detecting toxins that are highly specific and produced during the growth of bacteria. The biosensors that use antibodies are also under development; current research is oriented toward developing highly sensitive and specific antibodies, mostly monoclonal antibodies, which can be used. In addition to monoclonal antibodies, antibody fragments, recombinant antibodies or phage probes are also under development. The recombinant engineered antibodies can improve the binding kinetics to the antigen and can offer high sensitivity, specificity and stability, when compared with the conventional antibodies. The phage display libraries have improved the development of affinity probes that can be synthetically constructed, scanning across thousand of possible peptides of an agent, and decide on the best possible combination. The aptamers and peptide ligands are also alternative to antibodies. Aptamers are small DNA or RNA ligands that recognize a target by shape, not by sequence, and are mostly generated using combinatorial methods. Aptamers have been demonstrated to be useful in ricin detection based on bead based biochip sensor. Several short peptide sequences have also been tried as detection probes in the development of biosensors. Flow cytometry has been used as a detector in one of the biosensors, which uses luminex color coded beads conjugated to specific antibodies against biowarfare agents. The Autonomous pathogen detection system (APDS) developed by Lawrence Livermore National Laboratory (USA) used this system for the detection of biowarfare agents. The multicoded beads provide multiplexing capability as several agents can be looked for simultaneously; also, PCR has been successfully integrated with this system so that the positive flow cytometry result can be further confirmed using a PCR-based assay.[ 36 ] Various methods for detecting biowarfare agents using biochip technology are under development and evaluation, out of which the dielectrophoresis, for concentration of the target and further identification using an electric field driven immunoassay, is a very novel approach. In the Surface Plasmon Resonance (SPR) based biosensor, the target is directly analyzed by measuring the refractive index changes that occur when the target binds to the surface metal coated surface. The SPR biosensors that are small and field usable are being developed and tested for detection of toxins. Biochemical test based assays The culture of bacteria in a routine microbiological laboratory and further identification based on biochemical tests is the conventional identification method mostly followed in most of the hospital setups even today. As this method is followed for various diseases, the biological warfare agents, mainly the bacterial agents, need to be identified through this platform on a routine basis, if not during an intentional outbreak, but during natural prevalence of the disease. Most of the laboratories have standardized media for isolation for Ba. anthracis, Y. pestis, Burkholderia sp. and Brucella sp., and growth in specific media followed by few of the conventional biochemical tests can lead to the presumptive identification of these agents. Many commercial automated biochemical test platforms are today available for performing the tests after growth in bacteriological media, but these systems are mostly developed for bacterial diseases that are clinically important. These databases do contain data of important bacterial biological warfare agents and can be used in case of emergency. The tests based on substrate utilization pattern have been developed by Becton Dickinson (USA), Vitek (BioMe'rieux, France) and Microlog (Biolog, Hayward, CA USA). These systems have both visual observation and reader based formats for analysis of results and a database for comparison purposes. The main disadvantage of these systems is the need for pure culture and trained manpower to perform and for all the analyses. The BioMe'rieux API series of cards containing specific substrates have been used with good results in identification of most of the bacterial agents.[ 22 23 ] In addition to these tests, the system developed by MIDI Inc. (USA), called the microbial identification system, coverts cellular fatty acids from pure culture of bacteria to fatty acid methyl esters and uses a gas chromatograph for the separation and identification. In this method, a tedious sample preparation step in involved in which the bacterial cells need to be harvested, saponified, methylated, extracted and washed before analysis. But a chromatographic result pattern allows signature storage and database based retrieval for easy computing and analysis. Many of the databases based on the fatty acid profile have been now routinely used in laboratories for the confirmation of bacterial isolates. The major disadvantage is the need of time consuming isolation procedures as well as the need of highly calibrated instruments for reproducible performance. The advantage of these systems is that they allow simultaneous analysis of many samples and generate a highly reliable result output in the form of a fingerprint. They also allow addition of additional input based on experience.[ 24 ] Bioluminescence based detection Bioluminescence is the monitoring of luciferin and luciferase, the enzyme, interaction in the presence of ATP. The basic principle is ATP is found in all the living cells like virus and bacteria and its amount corresponds to the microbial load in air, water and in other environmental samples.[ 25 ] The bioluminescence has been widely used mainly in the quality control testing of bacterial contamination in food industry. This principle has been used in the Profile-1 hand held system developed by New Horizon Diagnostics (Columbia, MD, USA), which uses a microluminometer to read the sample bioluminescence.[ 26 ] In the Filtravette system developed by New Horizon Diagnostics, care has been taken to remove non-bacterial ATP. The main disadvantage of the bioluminescence based systems is the ATP contamination from nonmicrobial sources. Moreover, as all of them detect ATP, the results are specific and cannot be pointed toward any biowarfare agents. Since these systems are cost effective can be used for the real time air monitoring as first line defence. The continous air quality monitoring will trigger alarm for any unusual raise in the microbial load in the environment. Antigen and antibody based detection systems The antigen and antibody based immunoassays are being developed for the detection of bacterial and viral biowarfare agents.[ 27 ] The immunologic detection of antigens or antibodies based on ELISA has been developed, evaluated and currently in use for the detection of anthrax, plague, botulism, brucellosis, glanders and melioidosis. In these immunoassays, the efficacy of the detection depends on the availability of antigen or antibody in the detection matrix, the quality of the antigen that needs to be detected and the quality of the antibodies used in the detection of the antigen.[ 28 ] In short, it depends upon the quality of the antigen and antibody complex formed and the ability of the detection method to detect the formed antigen and antibody complex. Many different formats employing the ELISA principle have been developed using different substrate labels, i.e., fluorescent, chemiluminescent, electrochemiluminescent and on various platforms like solid support ELISA plates, visual dot and lateral flow formats. In addition, instruments using the immunoassay principles like biosensors, flow cytometry and micro array are under different stages of development. The disadvantage of the ELISA based systems is its ability to perform detection of only one agent at a time and sequential assays need to be performed to detect more than one agent. The specificity of the immunoassays limits the detection to minimum 10 5 CFU per test and is considered much higher than the DNA based assay. The immunoassay based on the sandwich antigen capture assay format is used in the Luminex xMAP system developed by Luminex Corp , Austin, Tx,USA, in which the polystyrene beads are coated with the specific antibodies and these beads are spectrally unique and are color coded into different sets; as each of these beads is coated with antibodies specific to an agent, it allows simultaneous detection of multiple agents in a liquid phase detection.[ 29 ] The laser in the detector detects the excitation of internal bead dyes of any capture agent and the signal intensity is proportional to the amount to agent present in the sample. This system allows to be hooked to a continuous environmental monitoring system for the detection and online monitoring. In another instrument the BV M-Series developed by BioVersis Corp. (USA), the reporter molecule uses electrochemiluminescence using paramagenetic beads as the support for the coating the capture antibodies. The electromagnetic beads are separated from the matrix by passing through a magnetic field and detected; this system has been modified as a kit for portable detection.[ 30 ] The bio-detector of Smiths Detection (USA) uses ELISA principle in a tape format and in which liquid samples are injected. In the reaction, the antigen available in the sample is mixed with biotin labeled and fluorescenin labeled antibodies. The system has a sensor that detects the signal from the biotin and fluorescenin labeled antibodies. Dissociation enhanced lanthanide fluorescence immunoassay (DELFIA) developed by Perkin Elmer Life Science (USA) is a unique system based on time resolved fluorescence.[ 31 ] This technique relies on lanthanide chelate labels which have long fluorescence decay time that allows measurements without background signals. Simply, in this system, the antibodies are labeled with lanthanide and tests are performed on standard plate ELISA format. After the antigen and antibody complex is formed, the label is disassociated from the antibody using a low pH enhancement solution and the free molecules rapidly form new stable highly fluorescent chelates that can be read by the system. The latest system that has gained popularity in the ELISA format is the lateral flow system which is rapid and cost effective when compared with the instrument-based detection systems. These tests are based on single use, disposable cartridge tests in the form immunochromatographic (ICT) line assays that generate visual lines in the membrane. In this class of assay, the detector antibodies are coated with colloidal gold or micro particles which by forming an antigen and antibody complex form a visual line or dot. These tests incorporate a procedural control that acts as a negative control and gives a confidence of testing while performing them.[ 32 ] These tests are easy to perform and are rapid than most of the immunoassays, but the sensitivity is often questioned as high false positive results are encountered. This can be used as a presumptive test and can be further followed up with a confirmative lab-based test. Lateral flows have been developed for all the biothreat agents by different agencies, but information on the use in real time detection situation and comparative evaluation with other methods is not available for recommendation. Nucleic acid based detection The nucleic acids, DNA and RNA, based detection systems is the most researched and developed detection system for the detection of biowarfare and biothreat agents. In the conventional PCR, the specific region of the genome is amplified and checked on electrophoresis for the amplification of correct size of product. In the advance quantitative real time PCR or Q PCR assays, the PCR amplification is combined with real time detection based on reporter fluorescence dyes.[ 33 34 ] In the non specific Q-PCR, the amplified DNA is detected based on DNA intercalating fluorescence dyes (SYBR green), and in specific Q-PCR, the fluorescence probes that specifically bind to a sequence are used. In the biowarfare agent detection, the specific Q-PCR assays based on probes have been developed for all the agents. In all the Q-PCR assays, software monitors the progress of the reaction, and the presence of the agent can be detected online on a monitor and the data can be transferred over a long distance for action. Today, compact, faster and sensitive real time PCR systems are available, making this method the most preferred method of agent detection. The limitations of this assay are the variation that exists with the nucleic acid based amplification techniques, the availability of starting material and its quality, inhibitory substances present in the matrices, sensitivity and specificity based issues related to quality of primers, probes, enzymes used. The advantage being the ability to detect a few cells, mostly up to 100 cells per test, makes it an attractive alternative to all the detection systems available today. The portability and stability of the reagents, use of stored power supply, etc., will make this technology more practical in the years to come. The developers based on the Q-PCR have incorporated disposable testing cartridges for various types of samples like the GeneXpert Q-PCR of Cepheid (USA) and eliminating the need to extract DNA prior to processing makes them convenient to use in field situations. The light cycler system of Roche Diagnostic, GmbH, Mannheim, Germany has assays specific to Ba. anthracis , which use a glass capillary into which extracted DNA is placed instead of a cartridge, and is commercially available. The Applied Biosystems (ABI, Foster City, USA) has several Q-PCR kits developed for anthrax and other biothreat agents, which can be used with their ABI 7500 and 7300 real time PCR systems. In case of the Bioseeq developed by Smiths Detection (USA), all the reagents are freeze dried for better stability and storage and this commercial hand held version can analyze up to six agents. Several other firms have also started marketing readymade kits for use with various real time PCR platforms for the detection of biowarfare agents, but no comparative results are available for these kits for evaluation. Sensor based detection systems The sensor based detection systems have been developed based on all the three groups that have been discussed already. The sensors have been developed for biochemical, immunologic and nucleic acid based detection systems. In the sensor based detection systems, the basic principles of all these three groups are integrated with a transducer than can transform the response in to an analyzable signal.[ 35 ] All these systems that integrate the biological component with a physical transducer are termed as biosensors. The physical transducers that produce an analyzable signal could be of electrochemical, optical, mass, thermal or of high frequency. In the electrochemical systems, the products that are formed due to an enzymatic activity associated with microbial metabolism are monitored based on conducting polymers. Array of sensors, each specific for different vapors or gases, are constructed and used to detect multiple analytes. Mostly, this technique is used to detect volatile organic compounds produced during the growth of bacteria or fungi and is termed as electronic noses. This system is very complex as most of the compounds are produced by multiple bacteria and analysis of results is difficult. But this has application in mainly detecting toxins that are highly specific and produced during the growth of bacteria. The biosensors that use antibodies are also under development; current research is oriented toward developing highly sensitive and specific antibodies, mostly monoclonal antibodies, which can be used. In addition to monoclonal antibodies, antibody fragments, recombinant antibodies or phage probes are also under development. The recombinant engineered antibodies can improve the binding kinetics to the antigen and can offer high sensitivity, specificity and stability, when compared with the conventional antibodies. The phage display libraries have improved the development of affinity probes that can be synthetically constructed, scanning across thousand of possible peptides of an agent, and decide on the best possible combination. The aptamers and peptide ligands are also alternative to antibodies. Aptamers are small DNA or RNA ligands that recognize a target by shape, not by sequence, and are mostly generated using combinatorial methods. Aptamers have been demonstrated to be useful in ricin detection based on bead based biochip sensor. Several short peptide sequences have also been tried as detection probes in the development of biosensors. Flow cytometry has been used as a detector in one of the biosensors, which uses luminex color coded beads conjugated to specific antibodies against biowarfare agents. The Autonomous pathogen detection system (APDS) developed by Lawrence Livermore National Laboratory (USA) used this system for the detection of biowarfare agents. The multicoded beads provide multiplexing capability as several agents can be looked for simultaneously; also, PCR has been successfully integrated with this system so that the positive flow cytometry result can be further confirmed using a PCR-based assay.[ 36 ] Various methods for detecting biowarfare agents using biochip technology are under development and evaluation, out of which the dielectrophoresis, for concentration of the target and further identification using an electric field driven immunoassay, is a very novel approach. In the Surface Plasmon Resonance (SPR) based biosensor, the target is directly analyzed by measuring the refractive index changes that occur when the target binds to the surface metal coated surface. The SPR biosensors that are small and field usable are being developed and tested for detection of toxins. Conclusion The threat of use of biological warfare agents in a terrorism-related issue or in a warfare situation is real and looming before us. The highly unpredictable nature of any event involving biological warfare agents has given rise to the need for developing rapid and accurate detection systems. The bioterrorist events are difficult to predict and prevent; in the case of a release, accurate, easy deployable detection systems are needed to minimize the damage and to prevent further spread of these agents. The intentional release of spores of Ba. anthracis in the US postal service proved the need for these detection platforms. Although many such detection systems are under development and are at various stages of evaluation, a single system to detect all the known biowarfare agents is going to be a real challenge. The complex matrices that need to be processed at the time of detection and genetically modified or uncharacterized agents that may be available in the sample have to be resolved. In this review, attempts have been made to provide basic information on biological warfare agents. Also, a comprehensive review of the current available methods for their detection has been made. The current knowledge on the agents and their detection methodologies will be highly useful in developing a highly effective response system through systematic planning.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7767301/

A Monoclonal Antibody against the C-Terminal Domain of Bacillus cereus Hemolysin II Inhibits HlyII Cytolytic Activity

Bacillus cereus is the fourth most common cause of foodborne illnesses that produces a variety of pore-forming proteins as the main pathogenic factors. B. cereus hemolysin II (HlyII), belonging to pore-forming β-barrel toxins, has a C-terminal extension of 94 amino acid residues designated as HlyIICTD. An analysis of a panel of monoclonal antibodies to the recombinant HlyIICTD protein revealed the ability of the antibody HlyIIC-20 to inhibit HlyII hemolysis. A conformational epitope recognized by HlyIIC-20 was found. by the method of peptide phage display and found that it is localized in the N-terminal part of HlyIICTD. The HlyIIC-20 interacted with a monomeric form of HlyII, thus suppressing maturation of the HlyII toxin. Protection efficiencies of various B. cereus strains against HlyII were different and depended on the epitope amino acid composition, as well as, insignificantly, on downstream amino acids. Substitution of L324P and P324L in the hemolysins ATCC14579 T and B771, respectively, determined the role of leucine localized to the epitope in suppressing the hemolysis by the antibody. Pre-incubation of HlyIIC-20 with HlyII prevented the death of mice up to an equimolar ratio. A strategy of detecting and neutralizing the toxic activity of HlyII could provide a tool for monitoring and reducing B. cereus pathogenicity. 1. Introduction Bacillus cereus is the fourth most common cause of foodborne illnesses [ 1 ]. Modern phylogeny singles out nine species in seven phylogenetic clades in the B. cereus group [ 2 ]. B. cereus species are associated with outbreaks of foodborne illnesses ( B. cereus sensu stricto and Bacillus cytotoxicus ) [ 3 ], food spoilage (psychrotolerant Bacillus weihenstephanensis and Bacillus mycoides ) [ 4 ] and anthrax disease in humans and animals ( Bacillus anthracis ) [ 5 , 6 ]. B. cereus hemolysin II (HlyII) belongs to pore-forming β-barrel toxins [ 7 , 8 , 9 , 10 ]. The hemolysin II gene is found in all clades of B. cereus sensu lato [ 11 ], but is more common among various natural isolates of Bacillus thuringiensis [ 12 ]. This places HlyII in the focus of the study, since B. thuringiensis is widely used as a biological insecticide for plant protection [ 6 ]. The toxin forms weakly anion selective pores, whose radius varies from 0.6 to 0.8 nm in both natural erythrocyte membranes and model membranes [ 13 ]. In its mature state, this toxin is the closest-known homolog of Staphylococcus aureus α-toxin with a 38% identity at the amino acid level [ 1 ]. The HlyII protein possesses a C-terminal extension of 94 amino acid residues designated as HlyIICTD (C-terminal domain) [ 7 , 14 ] previously undescribed for this class of toxins. HlyIIΔC shows an 8-fold lower hemolytic activity in affecting rabbit erythrocytes [ 14 ]; monoclonal antibodies (mAbs) against HlyIICTD proved useful in revealing the CTD role at various stages of pore formation. The formation of mAbs against pore-forming proteins is hampered by their toxicity of these proteins to animals during their immunizatin. Previously, these mAbs were produced using inactivated antigens. In this approach, however, the generated mAbs recognized proteins of an artificial structure. Some parts of such proteins can also be used as antigens. In this case, most mAbs obtained also recognize proteins with a structure different from the natural one. According to an assumption in [ 15 ], the presence of the C-terminal domain located on the surface of the mushroom-shaped part allows this part of the protein to be used as an antigen. mAbs against toxin fragments, instead of full-length proteins, are potential candidates for developing vaccine antigens [ 16 ]. In [ 17 ], we presented a panel of mAbs against this part of HlyII. Among these, the antibody HlyIIC-20 was found in the present work to be able to suppress the hemolysis of rabbit erythrocytes by HlyII. Here, we describe this mAb against HlyIICTD and determine its characteristics. 2. Results 2.1. Characteristic Features of the Monoclonal Antibody HlyIIC-20 The work [ 17 ] describes the formation of a panel of mAbs against the C-terminal domain of B. cereus HlyII, using to which the HlyIICTD is shown to be capable of independently binding erythrocyte membranes. All 24 antibodies from the formed panel have been tested for the ability to influence the hemolytic activity of HlyII. The mAb HlyIIC-20, which contains a heavy chain γ2b and a light chain κ, inhibits hemolytic activity (GenBank accession numbers MW194175 and MW194176). The antibody HlyIIC-20 recognizes HlyIICTD, both the native form of HlyII and the denatured forms of these proteins, in immunoblotting assays. The affinity constants characterizing the interaction of this antibody with HlyIICTD, intact HlyII14579 and HlyII771 were determined by indirect ELISA as described in [ 18 ]. Before the measurements and all other antibody experiments, the samples were checked for aggregates by dynamic light scattering. The analysis of the particle size distribution showed that the solution contained mainly particles corresponding to the molecular weight of the antibodies, without a significant number of high-molecular aggregates. Figure 1 shows a plot for determining the affinity constant for the interaction of HlyIIC-20 with HlyII14579. The affinity constant characterizing the interaction of HlyIICTD with HlyIIC-20 was 3.4 × 10 9 M −1 ; with HlyII14579, 2.6 × 10 9 M −1 ; with HlyII771, 1.8 × 10 9 M −1 . Based on the determined coefficients, it can be concluded that the degree of accessibility of the antigen epitope for natural hemolysins varies, and it differs from the recombinant HlyIICTD. The efficiency of recognition of the ATCC14579 T and B771 hemolysins II by the antibody HlyIIC-20 differs by 30%, while the binding efficiency of the antibody to the recombinant HlyIICTD of ATCC14579 T hemolysin is approximately 25% higher than that for the full-length protein. An excess of HlyIIC-20 did not inhibit the interaction of biotinylated HlyIICTD with erythrocytes, which indicates that the antigenic determinant on the surface of HlyIICTD recognized by this antibody is located at a site different from the site of target cell binding. 2.2. Dependence of HlyII Oligomerization on the Presence of HlyIIC-20 In the absence of erythrocyte membranes in solution, full-length HlyII is monomeric ( Figure 2 , lane 1). HlyII is characterized by the formation of stable oligomeric forms ( Figure 2 , lane 2) upon interaction with the target cell membrane [ 14 ]. Pretreatment of the full-length toxin with HlyIIC-20 prevents the formation of stable oligomeric forms ( Figure 2 , lane 3) during its interaction with the erythrocyte membrane, which means that the HlyIIC-20 epitope is located in the region of the molecule important for oligomerization of the monomeric form of the toxin into a full pore on the cell membrane and, consequently, the formation of a channel. 2.3. Identification of the Primary Sequence-Dependent Epitope Recognized by HlyIIC-20 Phage display revealed six sequences that could be epitopes recognized by HlyIIC-20 ( Figure 3 ). Analysis of the sequences in Figure 4 revealed no identical motifs, although all of them interacted with HlyIIC-20. This suggests that these epitopes are conformational, i.e., in the sequences it is necessary to look for a similar composition of amino acids, and their correct mutual arrangement occurs only when the three-dimensional (3D) structure is formed. When analyzing the composition of these sequences, we assumed that the formation of an epitope required a pair of amino acid residues with amino groups (N, Q or H) and a positively charged amino acid lysine or arginine (K, R). Figure 4 shows a conformational model of the C-domain (PDB: 6D5Z), in which amino acids N, Q and K are highlighted (see the legend to Figure 5 ). It should be noted that there are no histidines (H) in the amino acid sequence of the C-domain. An analysis of amino acid localization showed that the protein has only two places on the surface, where all the three amino acids, N, Q and K, are close together. One is at the beginning of HlyIICTD and the other in its middle part (see Figure 5 ). Phage display revealed two putative epitopes: NQKALEEQ and NGNQLK ( Figure 4 ). Figure 5 demonstrates a conformational model of the C-domain of HlyII with two regions where the amino acid residues N, Q and K are close to one another. Figure 5 B shows a 3D structure containing the amino acids that can presumably form an epitope at the beginning of the C-terminal domain; these are amino acids N320–Q327 and, in contact with them, N377 and Y396. Figure 5 D shows a 3D structure containing the amino acids that can presumably form an epitope in the middle of the C-domain; these are amino acids N350–K355 and N401, L403 in contact with them. The same areas are underlined by black lines in Figure 4 . To check which of these two putative epitopes is functional, this region of HlyII was compared with each epitope in three natural strains B771, ATCC14579 T and ATCC4342 T . The latter putative epitope was found to be identical in hemolysins of all three strains, while the former had one amino acid substitution. Hemolysin II of ATCC 14579 T and ATCC 4342 T contained an identical amino acid composition in this region, and in B771 HlyII, the amino acid residue of leucine was replaced by proline. The region located between these putative epitopes contains variable amino acid residues that are almost identical in B771 and ATCC4342 T hemolysins. To check which of the putative epitopes is true, reciprocal mutations L324P and P324L were introduced into ATCC14579 T and B771 hemolysins, respectively ( Figure 4 ). 2.4. Inhibition of HlyII in Experiments In Vitro In used natural strains, the putative conformational epitope has several amino acid substitutions both within and away from its linear region ( Figure 4 ). A set of constructs were made to verify the theoretically solved structure of the putative conformational epitope. The constructed ATCC14579 T and B771 hemolysin mutants were tested for the level of protection of rabbit erythrocytes against hemolysis by HlyIIC-20. Comparison of the curves in Figure 6 A,C shows that the level of protection against hemolysis by the mutant form of ATCC14579 T hemolysin II is significantly reduced. Thus, the presence of the amino acid L324 is essential for the suppression of hemolysis. The affinity coefficient of HlyIIC-20 with B771 HlyII differs slightly from that for ATCC14579 T . Introduction of the substitution P324L into B771 hemolysin II confirmed the role of leucine in epitope recognition, since the antibody suppressed hemolysis in mutant more efficiently than in the wild type. The presence of proline in the linear part of the epitope impairs recognition by 30%, but the inhibition level of hemolysis significantly decreases (up to 10 times). The role of leucine in the linear part of the conformational epitope in protection against hemolysis in the presence of HlyIIC-20 was confirmed by using hemolysin from the ATCC4342 T strain ( Figure 4 and Figure 6 E). The influence of the proline amino acid residue (P324) on the recognition of the epitope by HlyIIC-20 and its effect on the decrease in the level of protection by the antibody is apparently determined by the possible existence of this residue in two isoforms. The cis and trans isoforms of proline are able to alter the 3D structure of proteins that contain proline residues [ 19 ]. The ratio of the isoforms is determined by the state of the peptidyl-proline- cis - trans isomerase of the bacterial cell [ 20 ] and can additionally regulate the functional activity of bacterial proteins by changing their structure. HlyIIC-20 recognizes a conformational epitope located at the N-terminus of HlyIICTD. The presence of proline in its content in some natural hemolysins reduces both the efficiency of antibody recognition and the level of protection against hemolysis due to its effect on the protein conformation. The treatment of rabbit erythrocytes with hemolysin on ice followed by washing with cold PBS showed that HlyIIC-20 could not protect erythrocytes from hemolysis. At room temperature, the antibody was shown to inhibit the hemolytic activity. The reason is, evidently, that the described antibody binds to HlyII in a monomeric form both in the solution and on the membrane. 2.5. Inhibition of HlyII in Experiments with Mice Experiments on the inhibition of the toxic activity of HlyII (ATCC14579 T ) in vivo were carried out using white BALB/c mice. The animals were injected intravenously through the tail vein with a dose of toxin equal to LD 50 and the corresponding amount of toxin pre-incubated with excess HlyIIC-20. In the case of using HlyII (ATCC14579 T ), after 24 h observation, all mice that received an injection of the toxin incubated with all dilutions of antibodies, including that with an equimolar number of antibodies, remained alive. In the control group, 50% of animals injected with a toxin dose of LD 50 were alive, while the deviation in each individual experimental group was no more than 25% with a hemolysin dose of 200 U per animal at an intravenous injection of purified ATCC14579 T HlyII. This result evidenced the toxin-neutralizing activity of HlyIIC-20 for HlyII (ATCC14579 T ). When the HlyII toxin from B. cereus B771 was used for injection, no noticeable neutralizing activity of HlyIIC-20 was observed. This survival rate suggests the possibility of developing antidotes against HlyII based on this antibody. As seen in Figure 6 A,B, the efficiency of rabbit erythrocyte protection against hemolysis upon addition of HlyIIC-20 for hemolysins from strains ATCC14579 T and B771 differs significantly. Under conditions of the experiment, rabbit erythrocytes are protected by HlyIIC-20 from 10 U ATCC14579 T HlyII, whereas for B771 HlyII, in the presence of HlyIIC-20, they are lysed immediately upon addition of less than 1 unit of HlyII. Thus, the results obtained on experimental animals agree with those described in the previous sections. 2.1. Characteristic Features of the Monoclonal Antibody HlyIIC-20 The work [ 17 ] describes the formation of a panel of mAbs against the C-terminal domain of B. cereus HlyII, using to which the HlyIICTD is shown to be capable of independently binding erythrocyte membranes. All 24 antibodies from the formed panel have been tested for the ability to influence the hemolytic activity of HlyII. The mAb HlyIIC-20, which contains a heavy chain γ2b and a light chain κ, inhibits hemolytic activity (GenBank accession numbers MW194175 and MW194176). The antibody HlyIIC-20 recognizes HlyIICTD, both the native form of HlyII and the denatured forms of these proteins, in immunoblotting assays. The affinity constants characterizing the interaction of this antibody with HlyIICTD, intact HlyII14579 and HlyII771 were determined by indirect ELISA as described in [ 18 ]. Before the measurements and all other antibody experiments, the samples were checked for aggregates by dynamic light scattering. The analysis of the particle size distribution showed that the solution contained mainly particles corresponding to the molecular weight of the antibodies, without a significant number of high-molecular aggregates. Figure 1 shows a plot for determining the affinity constant for the interaction of HlyIIC-20 with HlyII14579. The affinity constant characterizing the interaction of HlyIICTD with HlyIIC-20 was 3.4 × 10 9 M −1 ; with HlyII14579, 2.6 × 10 9 M −1 ; with HlyII771, 1.8 × 10 9 M −1 . Based on the determined coefficients, it can be concluded that the degree of accessibility of the antigen epitope for natural hemolysins varies, and it differs from the recombinant HlyIICTD. The efficiency of recognition of the ATCC14579 T and B771 hemolysins II by the antibody HlyIIC-20 differs by 30%, while the binding efficiency of the antibody to the recombinant HlyIICTD of ATCC14579 T hemolysin is approximately 25% higher than that for the full-length protein. An excess of HlyIIC-20 did not inhibit the interaction of biotinylated HlyIICTD with erythrocytes, which indicates that the antigenic determinant on the surface of HlyIICTD recognized by this antibody is located at a site different from the site of target cell binding. 2.2. Dependence of HlyII Oligomerization on the Presence of HlyIIC-20 In the absence of erythrocyte membranes in solution, full-length HlyII is monomeric ( Figure 2 , lane 1). HlyII is characterized by the formation of stable oligomeric forms ( Figure 2 , lane 2) upon interaction with the target cell membrane [ 14 ]. Pretreatment of the full-length toxin with HlyIIC-20 prevents the formation of stable oligomeric forms ( Figure 2 , lane 3) during its interaction with the erythrocyte membrane, which means that the HlyIIC-20 epitope is located in the region of the molecule important for oligomerization of the monomeric form of the toxin into a full pore on the cell membrane and, consequently, the formation of a channel. 2.3. Identification of the Primary Sequence-Dependent Epitope Recognized by HlyIIC-20 Phage display revealed six sequences that could be epitopes recognized by HlyIIC-20 ( Figure 3 ). Analysis of the sequences in Figure 4 revealed no identical motifs, although all of them interacted with HlyIIC-20. This suggests that these epitopes are conformational, i.e., in the sequences it is necessary to look for a similar composition of amino acids, and their correct mutual arrangement occurs only when the three-dimensional (3D) structure is formed. When analyzing the composition of these sequences, we assumed that the formation of an epitope required a pair of amino acid residues with amino groups (N, Q or H) and a positively charged amino acid lysine or arginine (K, R). Figure 4 shows a conformational model of the C-domain (PDB: 6D5Z), in which amino acids N, Q and K are highlighted (see the legend to Figure 5 ). It should be noted that there are no histidines (H) in the amino acid sequence of the C-domain. An analysis of amino acid localization showed that the protein has only two places on the surface, where all the three amino acids, N, Q and K, are close together. One is at the beginning of HlyIICTD and the other in its middle part (see Figure 5 ). Phage display revealed two putative epitopes: NQKALEEQ and NGNQLK ( Figure 4 ). Figure 5 demonstrates a conformational model of the C-domain of HlyII with two regions where the amino acid residues N, Q and K are close to one another. Figure 5 B shows a 3D structure containing the amino acids that can presumably form an epitope at the beginning of the C-terminal domain; these are amino acids N320–Q327 and, in contact with them, N377 and Y396. Figure 5 D shows a 3D structure containing the amino acids that can presumably form an epitope in the middle of the C-domain; these are amino acids N350–K355 and N401, L403 in contact with them. The same areas are underlined by black lines in Figure 4 . To check which of these two putative epitopes is functional, this region of HlyII was compared with each epitope in three natural strains B771, ATCC14579 T and ATCC4342 T . The latter putative epitope was found to be identical in hemolysins of all three strains, while the former had one amino acid substitution. Hemolysin II of ATCC 14579 T and ATCC 4342 T contained an identical amino acid composition in this region, and in B771 HlyII, the amino acid residue of leucine was replaced by proline. The region located between these putative epitopes contains variable amino acid residues that are almost identical in B771 and ATCC4342 T hemolysins. To check which of the putative epitopes is true, reciprocal mutations L324P and P324L were introduced into ATCC14579 T and B771 hemolysins, respectively ( Figure 4 ). 2.4. Inhibition of HlyII in Experiments In Vitro In used natural strains, the putative conformational epitope has several amino acid substitutions both within and away from its linear region ( Figure 4 ). A set of constructs were made to verify the theoretically solved structure of the putative conformational epitope. The constructed ATCC14579 T and B771 hemolysin mutants were tested for the level of protection of rabbit erythrocytes against hemolysis by HlyIIC-20. Comparison of the curves in Figure 6 A,C shows that the level of protection against hemolysis by the mutant form of ATCC14579 T hemolysin II is significantly reduced. Thus, the presence of the amino acid L324 is essential for the suppression of hemolysis. The affinity coefficient of HlyIIC-20 with B771 HlyII differs slightly from that for ATCC14579 T . Introduction of the substitution P324L into B771 hemolysin II confirmed the role of leucine in epitope recognition, since the antibody suppressed hemolysis in mutant more efficiently than in the wild type. The presence of proline in the linear part of the epitope impairs recognition by 30%, but the inhibition level of hemolysis significantly decreases (up to 10 times). The role of leucine in the linear part of the conformational epitope in protection against hemolysis in the presence of HlyIIC-20 was confirmed by using hemolysin from the ATCC4342 T strain ( Figure 4 and Figure 6 E). The influence of the proline amino acid residue (P324) on the recognition of the epitope by HlyIIC-20 and its effect on the decrease in the level of protection by the antibody is apparently determined by the possible existence of this residue in two isoforms. The cis and trans isoforms of proline are able to alter the 3D structure of proteins that contain proline residues [ 19 ]. The ratio of the isoforms is determined by the state of the peptidyl-proline- cis - trans isomerase of the bacterial cell [ 20 ] and can additionally regulate the functional activity of bacterial proteins by changing their structure. HlyIIC-20 recognizes a conformational epitope located at the N-terminus of HlyIICTD. The presence of proline in its content in some natural hemolysins reduces both the efficiency of antibody recognition and the level of protection against hemolysis due to its effect on the protein conformation. The treatment of rabbit erythrocytes with hemolysin on ice followed by washing with cold PBS showed that HlyIIC-20 could not protect erythrocytes from hemolysis. At room temperature, the antibody was shown to inhibit the hemolytic activity. The reason is, evidently, that the described antibody binds to HlyII in a monomeric form both in the solution and on the membrane. 2.5. Inhibition of HlyII in Experiments with Mice Experiments on the inhibition of the toxic activity of HlyII (ATCC14579 T ) in vivo were carried out using white BALB/c mice. The animals were injected intravenously through the tail vein with a dose of toxin equal to LD 50 and the corresponding amount of toxin pre-incubated with excess HlyIIC-20. In the case of using HlyII (ATCC14579 T ), after 24 h observation, all mice that received an injection of the toxin incubated with all dilutions of antibodies, including that with an equimolar number of antibodies, remained alive. In the control group, 50% of animals injected with a toxin dose of LD 50 were alive, while the deviation in each individual experimental group was no more than 25% with a hemolysin dose of 200 U per animal at an intravenous injection of purified ATCC14579 T HlyII. This result evidenced the toxin-neutralizing activity of HlyIIC-20 for HlyII (ATCC14579 T ). When the HlyII toxin from B. cereus B771 was used for injection, no noticeable neutralizing activity of HlyIIC-20 was observed. This survival rate suggests the possibility of developing antidotes against HlyII based on this antibody. As seen in Figure 6 A,B, the efficiency of rabbit erythrocyte protection against hemolysis upon addition of HlyIIC-20 for hemolysins from strains ATCC14579 T and B771 differs significantly. Under conditions of the experiment, rabbit erythrocytes are protected by HlyIIC-20 from 10 U ATCC14579 T HlyII, whereas for B771 HlyII, in the presence of HlyIIC-20, they are lysed immediately upon addition of less than 1 unit of HlyII. Thus, the results obtained on experimental animals agree with those described in the previous sections. 3. Discussion Pore-forming toxins are the most important virulence bacterial factors They represent an attractive target for the development of molecules that neutralize their actions with high efficacy [ 21 , 22 ]. Monoclonal antibodies are promising as highly specific and reproducible tools for neutralizing virulent factors [ 23 ]. The mAbs can be broadly cross-reactive and recognize several pore-forming toxins [ 24 ]. Despite the significant number of described neutralizing monoclonal antibodies against bacterial toxins [ 25 , 26 ], in the fight against bacterial infections, these antibodies are not yet widespread, in contrast to other fields, such as oncology and viral infections. Research activities towards developing novel strategies for the diagnosis and suppression of pore-forming toxins allow developing the anti-virulence therapy. Unlike in other studies, in our work, a separate fragment of the toxin molecule was used to obtain antibodies, which has been shown to be of importance for the process of hemolysis. A strategy of detecting and neutralizing the toxic activity of HlyII could provide a potent tool for monitoring and reducing B. cereus pathogenicity [ 27 ]. The first involves the detection of B. cereus HlyII in food and biological fluids [ 28 ]. The latter supposes the development of approaches to anti-virulence therapy, which requires the study of all stages of pore formation: regulation of the expression of genes encoding toxin, secretion of monomeric forms from a bacterial cell into the external environment, as well as stages of pore formation during an attack by eukaryotic cells. This paper describes HlyIIC-20 from the mAb panel against HlyIICTD that recognizes a conformational epitope and is able to protect rabbit erythrocytes against hemolysis. Oligomerization of the HlyII toxin is a necessary stage in the formation of pores and determines the possibility of hemolysis of rabbit erythrocytes [ 14 ]. Based on the analysis of the effectiveness of erythrocyte protection against hemolysis by hemolysins of various B. cereus strains, it can be concluded that the efficiency of hemolysis in the presence of HlyIIC-20 depends on the primary structure of the region of conformational epitope recognized by this antibody. This work presents a comparative analysis of these regions in hemolysins from B. cereus ATCC14579 T , B771 and ATCC4342 T strains. The amino acid composition downstream of the linear region of the epitope varies in different hemolysins. In this region, the compositions of the ATCC4342 and B771 hemolysin II were almost identical and different from that of ATCC14579 T hemolysin II. However, for HlyII from the strains ATCC14579 T and ATCC4342 T , the level of hemolysis inhibition was relatively the same, while the sensitivity of HlyII from B771 to the presence of HlyIIC-20 was noticeably lower. The linear part of the epitope in B771 hemolysin II has a natural replacement of Leu324 by Pro, which apparently determines the revealed differences in the level of recognition by HlyIIC-20 and protection against hemolysis. Introduction of the reciprocal mutations L324P and P324L into ATCC14579 T and B771 hemolysins, respectively, changed their sensitivity to HlyIIC-20. The sensitivity of the mutant form of ATCC14579 T HlyII to the presence of the antibody decreased, while that of the B771 mutant increased significantly. The proline amino acid residue usually exists in proteins in two isoforms. One amino acid substitution in the epitope region reduced the recognition efficiency and inhibition of hemolysis in the presence of HlyIIC-20. The formation of mAbs based on a part of the toxin protein allows further use of these antibodies to identify toxins and to neutralize their action. 4. Conclusions An antibody capable of forming an immune complex in aqueous solution was found in a panel of monoclonal antibodies to the B. cereus recombinant HlyIICTD protein. HlyII pretreated with the antibody HlyIIC-20 inhibits the formation of oligomers and decreases the cytolytic activity of HlyII. The level of inhibition depends on the origin of HlyII. A comparison of HlyIICTD from various B. cereus strains revealed the presence of a variable region in the conformational epitope. The introduction of the reciprocal mutations L324P and P324L into this region of hemolysin II from the B. cereus ATCC14579 T and B771 strains, respectively, suggested the importance of these amino acid residues for pore maturation. Since the binding efficiency of HlyIIC-20 with respect to various hemolysins and HlyIICTD was different, and the antibody-caused suppression of hemolysis also varied, we assume a significant role of leucine located within the epitope in HlyII pore maturation at the step of oligomerization. The obtained data confirm the structure of the conformational epitope. HlyIIC-20 suppresses the HlyII-induced deaths of mice in quantities up to the equimolar pre-incubated HlyII/antibody ratio. 5. Materials and Methods 5.1. Plasmid Strains and Proteins Bacillus cereus ATCC 14579 T , Bacillus cereus ATCC 4342 T , Bacillus cereus B771 [ 29 ]. E. coli BL21 (DE3) (Novagen, Germany) was used to transform pET29b (+) (Novagen, Darmstadt, Germany), E. coli ER2738 for affinity phage selection. 5.2. Media and Solutions Medium: 2YT (16 g/L bactotryptone, 1 g/L yeast extract, 5 g/L NaCl, pH 7.0), 1.5% agar and 0.7% agar based on 2YT. Solutions: 1000 × IPTG/X-gal (1.25 g IPTG (isopropyl-β- d -thiogalactoside) and 1 g X-gal (5-bromo-4-chloro-3-indolyl-β- d -galactoside, Sigma, USA) in 25 mL DMF (dimethyl formamide, Sigma, USA), tetracycline 20 mg/mL (Sigma, USA) in 50% ethanol, phosphate buffered saline (PBS), PBST—PBS containing 0.1% Tween-20 (Sigma, USA), blocking solution of 1% gelatin (Sigma, USA) based on PBST, PEG/NaCl precipitating solution (20% (w/v) polyethylene glycol-6000, Sigma, USA, 2.5 M NaCl). Protein markers (Abcam, GB) and DNA electrophoresis markers (Fermentas, Lithuania), conjugate of streptavidin with horseradish peroxidase (Thermo Fisher Scientific, USA). 5.3. Determination of the Primary Sequence of CDR (Complementarity-Determining Region) Antibodies Hybridoma cells (10 6 ) producing antibodies that inhibit ATCC14579 T HlyII hemolysis were selected. The hydridoma cells were washed twice with PBS, supplemented with 1 mL of TRIzol (Invitrogen, Carlsbad, CA, USA) and frozen at –70 °C. RNA was isolated by the phenol-chloroform method using TRIzol reagent (Thermo Fisher Scientific, USA) according to the manufacturer's protocol (Thermo Fisher Scientific, USA) [ 30 , 31 ]. The purity and amount of RNA was assessed spectrophotometrically. Reverse transcription was performed according to [ 32 ] using SuperScript III reverse transcriptase (Invitrogen, USA) for 90 min at 42 °C. Next, the touchdown PCR with Q5 polymerase (NEB, USA), GC buffer and a proprietary set of primers (Biogen, Cambridge, MA, USA) was used. The heavy chain was sequenced using PCR primers at Evrogen (Russia). Due to the doubling of light chain sequences, the PCR product was excised from the gel and then integrated into SmaI site of pUC18 vector after blunt ends dephosphorylation using T4 ligase (NEB, Ipswich, MA, USA). The reaction was stopped by heating the mixture to 65 °C, and the mixture was used to transform chemically competent cells E. coli XL1-blue (Evrogen, Moscow, Russia). Selected clones were sequenced by Evrogen. 5.4. Production and Isolation of mAbs against HlyIICTD MAbs were isolated by affinity chromatography on protein A sepharose (Thermo Fisher Scientific, USA) [ 33 ] from hybridoma culture fluids secreting the mAbs. Then, the samples were centrifuged at 12,000× g at +4 °C. Dynamic light scattering was measured by a particle size analyzer from "Malvern Zetasizer Nano ZSP." The thermostated cuvette had a size of 0.3 × 0.3 cm and a volume of 100 μL. Light scattering was measured at an angle of 173 degrees. The data were analyzed with Malvern software. The types of heavy and light chains of immunoglobulins were determined by ELISA (Termo Fisher Scientific, USA) according to the manufacturer's instructions. 5.5. Conjugation of Antibodies with Biotin HlyIIC-20 and HlyIICTD were biotinylated using a solution of biotin N-hydroxysuccinimide ester (Sigma, St. Louis, MO, USA) in dimethyl sulfoxide at a concentration of 2.9 mM. Biotin ether was added at a 20-fold molar excess in relation to antibodies. The mixture was incubated for 4 h at room temperature. To remove the unreacted reagent, the mixture was dialyzed against PBS overnight. 5.6. Immunoblotting HlyII toxin (0.3 µM) was incubated with HlyIIC-20 (1 µM) for 1 h at 37 °C, then erythrocytes were added to a final concentration of 0.5% and incubated for 1 h at 37 °C with stirring. Control samples contained no HlyIIC-20. Electrophoretic separation of proteins was carried out in the presence of β-mercaptoethanol (Sigma, USA) as in [ 34 ]. Transfer to a nitrocellulose membrane was carried out for 15 h at a current of 20 mA in a buffer containing 25 mM Tris–HCl, 0.25 M glycine, 0.1% sodium dodecyl sulfate, 20% methanol, pH 8.3. Centers of nonspecific sorption were blocked by adding 1% ( w / v ) gelatin solution in PBST for 30 min. Then, the membrane was incubated for 2 h with biotinylated antibodies HlyIIC-20 (10 μg/mL). After incubation, the membrane was treated for 1 h with streptavidin conjugated with horseradish peroxidase diluted in PBST according to the manufacturer's recommendation (Thermo Fisher Scientific, Waltham, MA, USA). At each stage, the membrane was thoroughly washed with PBST. The membrane was stained with a solution containing 3 mM diaminobenzidine-3,3 tetrahydrochloride (Sigma-Aldrich, St. Louis, MO, USA) and 0.03% hydrogen peroxide. 5.7. Peptide Phage Display Peculiarities of the interaction of HlyIICTD with erythrocytes revealed with the use of HlyIIC-20 set the task of determining the epitope recognized by this antibody. The antigenic determinant HlyIICTD recognized by HlyIIC-20 was determined by the peptide phage display method as described in [ 35 ]. We used a library of random peptides of 12 amino acid residues (NEB, Ipswich, MA, USA) displayed on the M13KE phage [ 36 ]. The 12-dimensional phage peptide library has a repertoire of 109; peptides are exposed on the surface of bacteriophages in the composition of the minor protein pIII. Each clone in the library is represented by ~1000 copies. Selection of phages carrying peptide amino acid sequences specifically interacting with HlyIIC-20 was performed according to their ability to interact with one another. Three rounds of selection were carried out. The DNAs of the selected individual phage clones were sequenced. Using the GeneRunner program, the obtained nucleotide sequences were analyzed and the amino acid sequences of the peptides recognized by HlyIIC-20 and exposed in the pIII protein of bacteriophages were determined. After analyzing their interaction with the mAb, 10 clones were selected, in which the sequences of the peptide displayed by the phage were determined. Affinity selection of bacteriophages, cultivation of bacteriophages in semiliquid agar and isolation of bacteriophage DNA were performed as in [ 37 ]. 5.8. DNA Sequencing Analysis of the DNA sequences of the descendants of phage clones to determine the sequence of the insertions was carried out at the Evrogen (Russia). Sequencing results were processed using the Gene Runner 6.5.52 and ClustalW programs. 5.9. Interaction of Biotinylated HlyIICTD with Erythrocytes in the Presence of HlyIIC-20 The effect of the antibody HlyIIC-20 on the interaction of HlyIICTD with erythrocytes was studied in the reaction carried out in PBS containing 10% bovine serum as follows: 1.7 μM HlyIIC-20 was preincubated with 0.39 μM HlyIICTD-bio for 1 h at 37 °C, then a suspension of erythrocytes was added to a concentration of 0.025%, and the mixture was incubated for 1 h at 37 °C. Washing was carried out with 5% bovine serum albumin solution in PBS to remove a possible nonspecific interaction. A total of 100 μL of the solution was added to the wells and centrifuged for 10 min at 3000 rpm. The reaction was developed with streptavidin conjugated with horseradish peroxidase. 5.10. HlyII Neutralization Assay The hemolytic activities of wild type HlyII from various B. cereus strains (ATCC 14579 T , ATCC4342 T and B771) and mutant forms of recombinant hemolysin II (ATCC14579 T L324P, B771 and P324L) and their neutralization by HlyIIC-20 were investigated using rabbit erythrocytes [ 38 ]. Lysates of induced bacterial cells carrying recombinant plasmids with natural and mutant HlyII at a stepwise twofold dilution in a volume of 45 μL were incubated for 20 min at 37 °C in the presence of 5 μL of 6.7 μM HlyIIC-20 or PBS, then added to 50 μL of 1% rabbit erythrocytes in PBS. Analysis of the neutralization of HlyII with the antibody HlyIIC-20 after the integration of HlyII monomers into the erythrocyte membrane was performed according to the following scheme. After binding 0.5% suspension of erythrocytes to HlyII on ice, the suspension was divided into two parts and sedimented. One part was resuspended in PBS with HlyIIC-20 to a final concentration of 0.33 μM, the other part—in the antibody-free buffer. Both parts were incubated at 37 °C for 30 min. The hemolytic activity was measured as described in [ 38 ]. The 0.5% suspension of erythrocytes was bound to HlyII at room temperature in two test tubes and sedimented; one pellet was suspended in PBS with HlyIIC-20 to a final concentration of 0.67 μM, the other was suspended in PBS. The reaction mixture was kept for 5 min at room temperature and incubated at 37 °C for 30 min. Then, the hemolytic activity was measured [ 38 ]. 5.11. Animal Experiments The in vivo toxin neutralization assay was performed using female 6–8-week-old BALB/c mice about 20 g each, five experimental groups with four mice per group. The animals were used according to the protocol: "Search for toxin-neutralizing antibodies to bacterial toxins in a mouse model," registration number 674/18 of 08.10.2018, approved at a meeting of the Commission for Control over the Maintenance and Use of Laboratory Animals of the BIBCh RAS on 25 December 2018. These studies were planned for two years. Animals were obtained from the Laboratory Animal Breeding Nursery, Pushchino Branch, Institute of Bioorganic Chemistry, Russian Academy of Sciences, which has earned the international AAALACi accreditation. Toxins were applied in 100 μL PBS for intravenous challenge. Minimal lethal doses were determined in experiments using serial dilutions of toxins. For BALB/c mice, LD 50 was experimentally picked (the half-lethal dose is the average dose of a substance that causes the death of half of the members of the test group), which was 200 hemolytic units in 100 μL saline solution per animal. In an experiment for the in vivo toxin neutralization assay, groups of BALB/c mice were passively injected intravenously into the tail vein with serial twofold dilutions of HlyII. This amount of toxin was incubated with a 20-, 10-, 5- and 2-molar excess and the equimolar amount of HlyIIC-20 for 40 min at room temperature. Then, mixtures of antibodies with toxin, as well as toxin at a dose of LD 50 , were injected. The mice were observed up to 24 h and all deaths were recorded, during which time the animals were kept in their usual conditions of detention with free access to water and food. 5.1. Plasmid Strains and Proteins Bacillus cereus ATCC 14579 T , Bacillus cereus ATCC 4342 T , Bacillus cereus B771 [ 29 ]. E. coli BL21 (DE3) (Novagen, Germany) was used to transform pET29b (+) (Novagen, Darmstadt, Germany), E. coli ER2738 for affinity phage selection. 5.2. Media and Solutions Medium: 2YT (16 g/L bactotryptone, 1 g/L yeast extract, 5 g/L NaCl, pH 7.0), 1.5% agar and 0.7% agar based on 2YT. Solutions: 1000 × IPTG/X-gal (1.25 g IPTG (isopropyl-β- d -thiogalactoside) and 1 g X-gal (5-bromo-4-chloro-3-indolyl-β- d -galactoside, Sigma, USA) in 25 mL DMF (dimethyl formamide, Sigma, USA), tetracycline 20 mg/mL (Sigma, USA) in 50% ethanol, phosphate buffered saline (PBS), PBST—PBS containing 0.1% Tween-20 (Sigma, USA), blocking solution of 1% gelatin (Sigma, USA) based on PBST, PEG/NaCl precipitating solution (20% (w/v) polyethylene glycol-6000, Sigma, USA, 2.5 M NaCl). Protein markers (Abcam, GB) and DNA electrophoresis markers (Fermentas, Lithuania), conjugate of streptavidin with horseradish peroxidase (Thermo Fisher Scientific, USA). 5.3. Determination of the Primary Sequence of CDR (Complementarity-Determining Region) Antibodies Hybridoma cells (10 6 ) producing antibodies that inhibit ATCC14579 T HlyII hemolysis were selected. The hydridoma cells were washed twice with PBS, supplemented with 1 mL of TRIzol (Invitrogen, Carlsbad, CA, USA) and frozen at –70 °C. RNA was isolated by the phenol-chloroform method using TRIzol reagent (Thermo Fisher Scientific, USA) according to the manufacturer's protocol (Thermo Fisher Scientific, USA) [ 30 , 31 ]. The purity and amount of RNA was assessed spectrophotometrically. Reverse transcription was performed according to [ 32 ] using SuperScript III reverse transcriptase (Invitrogen, USA) for 90 min at 42 °C. Next, the touchdown PCR with Q5 polymerase (NEB, USA), GC buffer and a proprietary set of primers (Biogen, Cambridge, MA, USA) was used. The heavy chain was sequenced using PCR primers at Evrogen (Russia). Due to the doubling of light chain sequences, the PCR product was excised from the gel and then integrated into SmaI site of pUC18 vector after blunt ends dephosphorylation using T4 ligase (NEB, Ipswich, MA, USA). The reaction was stopped by heating the mixture to 65 °C, and the mixture was used to transform chemically competent cells E. coli XL1-blue (Evrogen, Moscow, Russia). Selected clones were sequenced by Evrogen. 5.4. Production and Isolation of mAbs against HlyIICTD MAbs were isolated by affinity chromatography on protein A sepharose (Thermo Fisher Scientific, USA) [ 33 ] from hybridoma culture fluids secreting the mAbs. Then, the samples were centrifuged at 12,000× g at +4 °C. Dynamic light scattering was measured by a particle size analyzer from "Malvern Zetasizer Nano ZSP." The thermostated cuvette had a size of 0.3 × 0.3 cm and a volume of 100 μL. Light scattering was measured at an angle of 173 degrees. The data were analyzed with Malvern software. The types of heavy and light chains of immunoglobulins were determined by ELISA (Termo Fisher Scientific, USA) according to the manufacturer's instructions. 5.5. Conjugation of Antibodies with Biotin HlyIIC-20 and HlyIICTD were biotinylated using a solution of biotin N-hydroxysuccinimide ester (Sigma, St. Louis, MO, USA) in dimethyl sulfoxide at a concentration of 2.9 mM. Biotin ether was added at a 20-fold molar excess in relation to antibodies. The mixture was incubated for 4 h at room temperature. To remove the unreacted reagent, the mixture was dialyzed against PBS overnight. 5.6. Immunoblotting HlyII toxin (0.3 µM) was incubated with HlyIIC-20 (1 µM) for 1 h at 37 °C, then erythrocytes were added to a final concentration of 0.5% and incubated for 1 h at 37 °C with stirring. Control samples contained no HlyIIC-20. Electrophoretic separation of proteins was carried out in the presence of β-mercaptoethanol (Sigma, USA) as in [ 34 ]. Transfer to a nitrocellulose membrane was carried out for 15 h at a current of 20 mA in a buffer containing 25 mM Tris–HCl, 0.25 M glycine, 0.1% sodium dodecyl sulfate, 20% methanol, pH 8.3. Centers of nonspecific sorption were blocked by adding 1% ( w / v ) gelatin solution in PBST for 30 min. Then, the membrane was incubated for 2 h with biotinylated antibodies HlyIIC-20 (10 μg/mL). After incubation, the membrane was treated for 1 h with streptavidin conjugated with horseradish peroxidase diluted in PBST according to the manufacturer's recommendation (Thermo Fisher Scientific, Waltham, MA, USA). At each stage, the membrane was thoroughly washed with PBST. The membrane was stained with a solution containing 3 mM diaminobenzidine-3,3 tetrahydrochloride (Sigma-Aldrich, St. Louis, MO, USA) and 0.03% hydrogen peroxide. 5.7. Peptide Phage Display Peculiarities of the interaction of HlyIICTD with erythrocytes revealed with the use of HlyIIC-20 set the task of determining the epitope recognized by this antibody. The antigenic determinant HlyIICTD recognized by HlyIIC-20 was determined by the peptide phage display method as described in [ 35 ]. We used a library of random peptides of 12 amino acid residues (NEB, Ipswich, MA, USA) displayed on the M13KE phage [ 36 ]. The 12-dimensional phage peptide library has a repertoire of 109; peptides are exposed on the surface of bacteriophages in the composition of the minor protein pIII. Each clone in the library is represented by ~1000 copies. Selection of phages carrying peptide amino acid sequences specifically interacting with HlyIIC-20 was performed according to their ability to interact with one another. Three rounds of selection were carried out. The DNAs of the selected individual phage clones were sequenced. Using the GeneRunner program, the obtained nucleotide sequences were analyzed and the amino acid sequences of the peptides recognized by HlyIIC-20 and exposed in the pIII protein of bacteriophages were determined. After analyzing their interaction with the mAb, 10 clones were selected, in which the sequences of the peptide displayed by the phage were determined. Affinity selection of bacteriophages, cultivation of bacteriophages in semiliquid agar and isolation of bacteriophage DNA were performed as in [ 37 ]. 5.8. DNA Sequencing Analysis of the DNA sequences of the descendants of phage clones to determine the sequence of the insertions was carried out at the Evrogen (Russia). Sequencing results were processed using the Gene Runner 6.5.52 and ClustalW programs. 5.9. Interaction of Biotinylated HlyIICTD with Erythrocytes in the Presence of HlyIIC-20 The effect of the antibody HlyIIC-20 on the interaction of HlyIICTD with erythrocytes was studied in the reaction carried out in PBS containing 10% bovine serum as follows: 1.7 μM HlyIIC-20 was preincubated with 0.39 μM HlyIICTD-bio for 1 h at 37 °C, then a suspension of erythrocytes was added to a concentration of 0.025%, and the mixture was incubated for 1 h at 37 °C. Washing was carried out with 5% bovine serum albumin solution in PBS to remove a possible nonspecific interaction. A total of 100 μL of the solution was added to the wells and centrifuged for 10 min at 3000 rpm. The reaction was developed with streptavidin conjugated with horseradish peroxidase. 5.10. HlyII Neutralization Assay The hemolytic activities of wild type HlyII from various B. cereus strains (ATCC 14579 T , ATCC4342 T and B771) and mutant forms of recombinant hemolysin II (ATCC14579 T L324P, B771 and P324L) and their neutralization by HlyIIC-20 were investigated using rabbit erythrocytes [ 38 ]. Lysates of induced bacterial cells carrying recombinant plasmids with natural and mutant HlyII at a stepwise twofold dilution in a volume of 45 μL were incubated for 20 min at 37 °C in the presence of 5 μL of 6.7 μM HlyIIC-20 or PBS, then added to 50 μL of 1% rabbit erythrocytes in PBS. Analysis of the neutralization of HlyII with the antibody HlyIIC-20 after the integration of HlyII monomers into the erythrocyte membrane was performed according to the following scheme. After binding 0.5% suspension of erythrocytes to HlyII on ice, the suspension was divided into two parts and sedimented. One part was resuspended in PBS with HlyIIC-20 to a final concentration of 0.33 μM, the other part—in the antibody-free buffer. Both parts were incubated at 37 °C for 30 min. The hemolytic activity was measured as described in [ 38 ]. The 0.5% suspension of erythrocytes was bound to HlyII at room temperature in two test tubes and sedimented; one pellet was suspended in PBS with HlyIIC-20 to a final concentration of 0.67 μM, the other was suspended in PBS. The reaction mixture was kept for 5 min at room temperature and incubated at 37 °C for 30 min. Then, the hemolytic activity was measured [ 38 ]. 5.11. Animal Experiments The in vivo toxin neutralization assay was performed using female 6–8-week-old BALB/c mice about 20 g each, five experimental groups with four mice per group. The animals were used according to the protocol: "Search for toxin-neutralizing antibodies to bacterial toxins in a mouse model," registration number 674/18 of 08.10.2018, approved at a meeting of the Commission for Control over the Maintenance and Use of Laboratory Animals of the BIBCh RAS on 25 December 2018. These studies were planned for two years. Animals were obtained from the Laboratory Animal Breeding Nursery, Pushchino Branch, Institute of Bioorganic Chemistry, Russian Academy of Sciences, which has earned the international AAALACi accreditation. Toxins were applied in 100 μL PBS for intravenous challenge. Minimal lethal doses were determined in experiments using serial dilutions of toxins. For BALB/c mice, LD 50 was experimentally picked (the half-lethal dose is the average dose of a substance that causes the death of half of the members of the test group), which was 200 hemolytic units in 100 μL saline solution per animal. In an experiment for the in vivo toxin neutralization assay, groups of BALB/c mice were passively injected intravenously into the tail vein with serial twofold dilutions of HlyII. This amount of toxin was incubated with a 20-, 10-, 5- and 2-molar excess and the equimolar amount of HlyIIC-20 for 40 min at room temperature. Then, mixtures of antibodies with toxin, as well as toxin at a dose of LD 50 , were injected. The mice were observed up to 24 h and all deaths were recorded, during which time the animals were kept in their usual conditions of detention with free access to water and food.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3186013/

The eminent need for an academic program in universities to teach nanomedicine

Nanomedicine is on the cutting edge of technology applied to medical and biological sciences. Nanodevices, nanomaterials, nanoinstruments, nanotechnologies, and nanotechniques (laboratory methods and procedures) are important for the modern practice of medicine and essential for research that could stimulate the discovery of new medical advances. Accordingly, there is an eminent need for implementing an academic program in universities to teach this indispensable and pragmatic discipline, especially in the departments of graduate studies and research in the areas of pharmacology, genetic engineering, proteomics, and molecular and cellular biology. Introduction To be in step with the future, it is necessary today for academic programs to exist that teach topics related to nanomedicine, an important and relatively recent technological advance in medicine. Such programs should be included in university departments for graduate studies and research, especially in the areas of pharmacology, genetic engineering, proteomics, and molecular and cellular biology. Later, as nanotechnology continues to develop, other areas will certainly be affected. Research in nanomedicine began with discoveries of the novel physical and chemical properties of various carbon-based materials that can only be detected in nanometer-sized structures. Nanomedicine is defined as the monitoring, repair, construction, and control of human biological systems at the cellular, molecular, and atomic levels using engineered nanodevices and nanostructures. Another definition is that provided by the National Institute of Health in its roadmap for medical research in nanomedicine, which states that nanomedicine is "an offshoot of nanotechnology, [which] refers to highly specific medical interventions at the molecular scale for curing disease or repairing damaged tissues, such as bone, muscle, or nerve". 1 Nanomedicine contains five main disciplines: analytical tools, nanoimaging, nanomaterials (medications, raw materials, and devices), novel therapeutic and drug delivery systems, as well as regulatory and toxicological issues related to clinical practice. 1 Basic nanostructured materials, engineered enzymes, and the many products of biotechnology will most probably become essential elements in some areas of medical, chemical, and biological applications. However, the full promise of nanomedicine is yet to be defined, as this will to a large extent depend on the development of new technology, such as precisely controlled or programmable medical nanomachines and nanorobots to address medical problems. Moreover, the application of nanotechniques at the cellular, molecular, and atomic levels should certainly help in the understanding of the function of organisms, as well as in the development of better techniques for imaging, diagnosing, and treating disorders at the level of molecular and cellular biology. Microscopic machines were first hypothesized by the Nobel Prize winning physicist Richard P Feynman, now considered the father of nanoscience, in 1959. His article "There is Plenty of Room at the Bottom" is considered a classic in the establishment of the important possibilities for this field. 2 The main idea of this article was the ability to reduce the size of various apparatuses and at the same time increase their effectiveness, based on the manufacture of products by reordering atoms and molecules. Of particular importance in his proposal was the reference to nanorobots, to be used in the manufacture of nanodevices, nanomaterials, and nanomachines. He also mentioned the possibility of creating a 100-fold improvement in the electron microscope. Eric Drexler, considered the father of nanotechnology, wrote the popular books, Engines of Creation (1986) and Unbounding the future (1991), in which he announced the promise and warned of the dangers of molecular and atomic manipulation. In his more recent work, Nanosystems: molecular machinery, manufacturing and, computation (1992), he predicted that nanotechnology will make the conventional means of manufacture obsolete. 3 He proposed the concept of machines with the capability of ordering and reordering atoms in order to create precise molecular structures, thus resulting in the desired product. Such machines could even have the ability of self-reproduction. Once nanomachines are available, programmable and controllable microscale robots comprised of nanoscale parts manufactured to nanometer precision will open new possibilities related to curative and reconstructive procedures in the human body at the cellular, molecular, and atomic levels (perhaps at less than 100 nm). In keeping with the adage that "the best treatments are those that intervene the least", the possibility of nanotechnology to greatly increase knowledge of cellular and atomic events could lead to improved treatments. Additionally, nanomedicine is on the cutting edge of medical practice and research, and for this reason it is an important focus of capital investment. Consequently, it is urgent that academic programs be implemented in order to teach this discipline in universities that have a medical and biological focus of study and research. An academic discipline of nanomedicine at the university level would allow students, teachers, technicians, and researchers to become familiar with: (1) nanodevices, nanostructures, nanomaterials, nanostuffs, and precise and sophisticated nanoinstruments, such as those that could be used for molecular imaging and bedside biomarker testing and monitoring ( Appendix 1 ); (2) the main categories of precise nanotechnologies in real time ( Appendix 2 ); and (3) techniques (laboratory methods and procedures) needed for furthering research ( Appendix 3 ), including detailed procedures of nanobiosensors for medical uses ( Appendix 4 ). All of these elements of nanomedicine could be applied to cancer, cardiovascular diseases, inflammation, degenerative neurological diseases, toxicity, ethics, and societal and environmental impact. 4 Perhaps one of the most important potentials for nano-medical research is the gathering of extensive information about the chemical and physical properties of nanoscale biological structures. As the catalog of the interactions between individual molecules, cells, tissues, and organs develops, researchers will gain a greater understanding of the intricate operations of molecular structures, processes, and networks inside living cells based on the use of new probing tools. Consequently, scientists will be able to build new devices for a wide range of biomedical applications, such as detecting infectious agents or metabolic imbalances with tiny sensors, generating miniature devices that search for and destroy infectious agents, and replacing "broken" machinery inside cells with new nanoscale structures. 4 Current applications The following medical advances are among the existing uses of nanomedicine. Nanomaterials: nanotechnology used for subcutaneous treatments. 5 Nanoprobes in vivo: targeted nanoprobes reveal early time point kinetics in vivo by time-resolved magnetic resonance imaging (MRI). 6 Methods for the preparation of small interfering ribonucleic acids (siRNAs): a direct, efficient method for the preparation of siRNAs containing ribo-like north pseudo-sugars, 7 and drug delivery of siRNA therapeutics. 8 Physical assessment of toxicology at nanoscale: nanodose-metrics and toxicity factor. 9 Nanomaterials for regenerative medicine: use of nanomaterials with applications for regenerative medicine. 10 Vascular toxicity: studies of vascular endothelial toxicity. 11 Nanocatalizers: molecular-level biological transporters capable of converting biomass into biofuel. Nanoparticle devices: molecular-level biological transporters that release medication in a precise and controlled manner throughout the organism, such as in target cells, for crossing biological barriers and for providing greater stability in the face of pathogenic deterioration, as well as use of nanoparticles for the treatment of cancer. 12 Future directions and challenges New opportunities will most certainly appear in relation to fighting disease as nanotechnology develops. Doctors may be able to search out and destroy the very first cancer cells before they lead to a destructive tumor in the body. Perhaps a broken part of a cell could be removed and replaced with a miniature biological machine. Pumps the size of molecules could possibly be implanted to deliver life-saving medicines precisely when and where they are needed. These scenarios are the long-term goals of the "nanomedicine roadmap". 4 With the inclusion of the academic discipline of nanomedicine in their programs, universities will be able to provide the personnel capable of accelerating the development of this field. Future possibilities include the following opportunities: Nanoneedles: the nucleus of live cells could be penetrated by nanoneedles used in conjunction with microscopy with the aim of introducing nucleic acids, proteins or other compounds inside human cells, or even doing surgery; modeling the structure-property relationships of nanoneedles. 13 Environmental nanocatalizers: biological transporters capable of greatly improving catalytic converters currently used in automobiles, thus drastically reducing the pollution emitted. Nanobiosensors: sensors capable of detecting the presence of anthrax and other biological weapons; nanosensors for probing live cells. 14 Dendrimers: artificial molecules designed to replace defective endogenous elements in the organism, induce the production or secretion of various molecules related to the immune response in case of being genetically deficient, or act as biological markers (for cancer therapy, for the detection of adherence to medication regimens, or for the evaluation of experimental results). 15 Nanodepots of insulin: microscopic deposits injected into the bloodstream that could act as an artificial pancreas, releasing insulin to counteract the effects of diabetes; similar technologies could also be employed to release dopamine into neurons as therapies for Parkinson's or Alzheimer's disease. Titanium dioxide nanotubes: microscopic tubes that could clean polluted water. Carbon nanotubes: microscopic tubes aimed at replacing defective or absent genes; use of carbon nanotubes in vivo to characterize action mechanisms at the cellular level; 16 carbon nanotubes could reactivate synaptic connections and regenerate neurons damaged by an accident. Biological nanosensors: a set of ultramicroscopic and fine silicon threads. Molecular nanopyramids: these probes could possibly be released into the human body, with the capability of detecting and destroying the first cancer cells in the brain, heart, kidney, or lungs, without damaging neighboring tissue; they might also be able to detect viral or bacterial infections in an early stage before the immune system has begun to respond. Assembler nanomachines: machines based on the rearrangement of atoms and molecules to manufacture molecular components. Smart materials: these probe-like materials could potentially detect subtle changes in body chemistry. Nanoparticles: microscopic tumor-destroying genes may possibly be transported to certain sites, directed with great precision to destroy the genetic information of tumors, with the aim of inactivating the target cells and impeding their reproduction; other genes may be transported to detect tiny alterations in proteins, as well as to scan tissue to ensure that the nanodevices are neatly congregated around the target. Nanorobots: these devices could be introduced into the human body to repair damage and combat infection and disease by acting on DNA without being detected and destroyed by the host immune system. 17 – 22 Nanobiochips: this advanced technology could work with MRI to supply an abundant amount of information about the human body. 23 Nanocomputers: computers could be developed made of material stronger than steel but with only 10% of its weight. These could possibly guide nanomachines for the purpose of examining, taking apart, and rebuilding damaged cells, tissues, and organs through the reconstruction of their molecular structures; nanoinformatics could be applied to nanomedicine; 24 new computers could search for a methodology that reproduces human organs in vitro. Conclusion The academic discipline of nanomedicine could be included in university curriculums for graduate students and researchers today in the fields of pharmacology, genetic engineering, proteomics, and molecular and cellular biology. Such programs would train personnel in order to accelerate advances of techniques, technology, and applications of nanomedicine. As such uses increase and broaden, other new programs will also become important elements of university curriculums.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9679321/

Military Culture and Cultural Competence in Public Health: U.S. Veterans and SARS-CoV-2 Vaccine Uptake

As part of the U.S. effort to encourage vaccination for SARS-CoV-2, scholars have emphasized the importance of culture and identity in vaccine uptake decisions. The culture and identity of military service are poorly understood in the context of understanding Veterans' acceptance of COVID-19 vaccines. In analyzing data from semi-structured interviews with Veterans in homeless transitional housing, this article examines their willingness to get vaccinated for COVID-19. Themes invoking military culture included (a) mandatory vaccinations in the military; (b) cynicism and mistrust toward the government; and (c) trust of and reliance on Veteran peers with shared military culture in decision-making. To further understand how military culture influences vaccine uptake and explore avenues for building culturally competent, trust-based health care interventions with Veterans, a previously published case study of Veterans volunteering in Team Rubicon (TR) disaster relief is examined. Veteran participants in TR described the experience of being in a Veteran-centric organization as an empathetic context wherein they were able to: (a) address their reintegration struggles; (b) gain a new appreciation of their Veteran status; and (c) connect with trusted peers. Given TR's credibility, Veteran-serving health care organizations could collaborate with Veteran-led organizations to expand shared efforts to address Veterans' distrust of government-promoted vaccines. Introduction and Background The U.S. public health system strongly encourages vaccination for SARS-CoV-2, the virus that causes COVID-19, as an important tool in stemming the tide of adverse impacts of this public health emergency. As part of this effort, advocates and scholars alike have emphasized concerns that historical oppression and marginalization by the medical system and government may influence how Black, Hispanic, and Indigenous populations perceive and respond to the public health system's recommendation to get vaccinated ( Schoch-Spana et al., 2021 ). Understanding these historical reasons behind such vaccine reluctance is vital to an equitable public health response to COVID-19 because Black, Hispanic, and Indigenous populations face a disproportionate burden of SARS-CoV-2 infection and death ( M. A. Garcia et al., 2020 ; Labgold et al., 2021 ; Webb Hooper et al., 2020 ). Cultural competence, the ability to respect communities' historical reasons for distrust and skepticism, is vital to an equitable public health crisis communications strategy promoting vaccination ( Lofaro & Sapat, 2022 ). An equity-focused approach to public health efforts to engage with communities of color and other subpopulations around vaccine uptake requires an empathetic approach. Such an approach involves public acknowledgment of historical abuse and other reasons for distrust of health care systems, such as ongoing health disparities ( Schoch-Spana et al., 2021 ). U.S. military Veterans also face a disproportionately greater disease burden from COVID-19 due to existing health factors ( Cardemil et al., 2021 ). They are also a group that has had a complex relationship with the U.S. government and have experienced health care inequities in the form of slow government responses to health conditions related to military service ( Elwy et al., 2021 ; Hobbs, 2008 ; Jasuja et al., 2021 ; Rein, 2021 ). Their views toward COVID-19 vaccination are not well understood, even if they have similar vaccination rates to the U.S. population, and the factors that shape their vaccination decisions are even more poorly understood ( Elwy et al, 2021 ; Jasuja et al., 2021 ; Rein, 2021 ). Within the public health community, understanding of military culture and Veteran culture, and their impact on Veterans' perception of vaccines and public health, is limited. Public health campaigns rarely incorporate principles of military-informed care into their engagement or messaging practices ( Borah et al., 2021 ; Franco et al., 2021 ; Hobbs, 2008 ). While the public health field has made significant efforts to recognize the histories of racial and ethnic communities' distrust in their efforts, Veteran engagement has not been a significant feature of the U.S. public health system's efforts to build trust with marginalized subgroups. Prior studies on Veterans' uptake of COVID-19 vaccines have not examined Veterans' military culture as a factor ( Der-Martirosian, 2013 ; Elwy et al., 2021 ; Jasuja et al., 2021 ). Efforts to vaccinate Veterans in pockets of vaccine-hesitant areas find that, for some Veterans, refusing a vaccine is a form of individual resistance to governmental power stemming from their feelings of betrayal ( Rein, 2021 ). The complex historical relationship between Veterans and the U.S. government, particularly regarding questions of institutional power and bodily autonomy, is likely to shape their response. This article examines the attitudes and beliefs of Veterans enrolled in U.S. Department of Veterans Affairs (VA) homeless programs about COVID-19 vaccines and their decisions regarding vaccinations. It seeks to understand how military culture and Veterans' group identities influence their responses to recommended COVID-19 vaccination. We seek to understand how aspects of their identity, as resistive to institutional power, impact their interpretations of such health interventions. Analyzing data from interviews with Veterans enrolled in VA-funded homeless transitional housing, we explore their responses toCOVID-19 vaccines. We unravel how military culture shapes Veterans' worldviews and how these sociocultural dimensions are reflected in their responses to COVID-19 vaccines. We analyze these Veterans' responses to vaccines and health behavior recommendations in the context of navigating their bodily autonomy, and ultimately, as forms of resistance to institutional power. Thus, we introduce the need for culturally competent health care practice informed by the culture and history of the Veterans' military experience as a more empathetic approach in efforts to promote vaccination. An empathetic approach to community engagement is foundational to person-centered interventions that recognize the need for trust-building within populations such as Veterans. Such approaches are the underpinning of humanistic psychology, which emphasizes supporting autonomous decision-making and encourages gaining the buy-in of self-directed individuals ( Joseph & Murphy, 2013 ; McLeod, 2019 ; Quinn, 2013 ). Finally, we examine the implications for public health practice. In July 2021, the VA found that 10% of its enrolled Veterans (600,000 Veterans) indicated that they did not intend to get the COVID-19 vaccine ( Slack, 2021 ), suggesting a need for more effective engagement with them to promote vaccine uptake. As of August 2021, 60% of Veterans participating in VA homeless transitional housing nationwide had received at least one vaccine dose, while 40% remained unvaccinated ( Balut et al., 2021 ). In contrast, 79% of the U.S. population has received at least one vaccine dose, while 67% had received two doses. While these Veterans who are residing in VA-funded housing may be getting vaccinated in numbers comparable to the U.S. general population, vaccine hesitancy is significant in this population, highlighting the value of understanding whether factors related to military culture may be impacting their decisions. As Veterans transition from military to civilian roles, they often navigate their social identities based on their membership in various social groups ( Thompson et al., 2017 ). Amid this post-military transition, some Veterans may find greater trust and affinity in groups that possess understanding and empathy for military culture than groups lacking that cultural awareness. One clear example is Team Rubicon (TR), whose credibility stems from its recreating of military culture in ways Veterans have autonomy over and from its status as a Veteran-run organization ( Kranke et al., 2016 ). Depending on the cohort, some Veterans, such as Vietnam-era Veterans ( Heslin et al., 2013 ), post-9/11 Veterans who comprise many TR participants ( Kranke et al., 2016 ), or Veterans experiencing homelessness ( Gin, Balut, & Dobalian, 2022 ), distrust government institutions due to their own and peers' experiences and maybe consequently cynical and resistant to government public health campaigns that are overly coercive and not sufficiently respectful of their autonomy and their military experiences. The Veteran Experience—What the Literature Tells Us Military identity and its general influence on the behaviors and perspectives of former service members remain opaque to civilians, including many health care system providers ( Borah et al., 2021 ). "Military culture," consisting of "the values, traditions, norms, and perceptions that govern how members of the armed forces think, communicate, and interact with one another and civilians" ( Coll et al., 2011 , p. 489), is a distinct subculture within American civilian society that continues to be significant for some Veterans after separating from the military ( Exum et al., 2011 ; Meyer et al., 2022 ; Weiss & Coll, 2011 ; Wool, 2015 ). For example, military culture may shape their willingness to engage with services such as mental health care, due to concerns about the stigma of "weakness" going against military values of toughness ( Castro et al., 2015 ; Weiss & Coll, 2011 ). Becoming a Veteran involves joining the military, serving, and transitioning from the military into civilian life. This process of socialization may hold insight into how belonging to this cultural subgroup "matters" when it comes to how Veterans perceive and respond to information and health recommendations. We build on the work of Coll et al. (2011) , and additional literature regarding the unique, experience-based mindset of former members of the U.S. military. Becoming a "Veteran" can be broken down into four processes. These are not intended to include all psychological processes involved with joining the U.S. military, which are complex and diverse, nor are they intended to be generalizable to all Veterans. Entering the Military: A New Relationship With the Government First, giving up one's existing civilian identity and joining the military involves entering a long-term, complex relationship with the government ( Hobbs, 2008 ; Jasuja et al., 2021 ; Rein, 2021 ) and accepting the U.S. government's authority over oneself and one's own body. In some cases, this involves forgoing bodily autonomy and informed consent over medical and other decisions. In particular, military members are not afforded the same option of informed consent that civilians enjoy, as established in the United States through the 1979 Belmont Report, over their acceptance of investigational drugs and vaccines ( Black, 2007 ; Rettig, 1999 ; U.S. Government Accountability Office [GAO], 2002 ): A service member "freely agrees when joining the military to relinquish autonomy" to the interests of the unit or mission. Furthermore, they are aware "as they approach actual fighting, their autonomy dramatically decreases." ( Rettig, 1999 : p. 50) Individual informed consent was not required in the military's anthrax vaccination program ( Black, 2007 ; GAO, 2002 ) during post-9/11 military campaigns in Iraq and Afghanistan nor for investigational drugs such as pyridostigmine bromide and botulinum toxin administered during the 1991 Gulf War in Iraq ( Rettig, 1999 ). Military service also includes the historical legacy of exposures to environmental hazards for which compensation for service-connected disability was previously denied by the U.S. government. Currently, the VA is studying the long-term health effects of such exposures as Agent Orange in Vietnam, and respiratory illnesses from burn pits while serving in Iraq and Afghanistan ( U.S. Department of Veterans Affairs, 2022 ), and has subsequently recognized the health effects of these exposures as service-connected disabilities. As of April 2022, respiratory illnesses connected to exposure to airborne hazards in burn pits in the Middle East were eligible for presumed service-connected disabilities ( U.S. Department of Veterans Affairs, 2022 ). Acculturation: Adopting Military Cultural Values as One's Own Second, joining the military, individuals often take on a new identity ( Coll et al., 2011 , p. 489). Military cultural values such as "service, competence, sacrifice, mission, and collectivism" ( Meyer et al., 2022 ) are deeply ingrained within the identity of military members. Military members forgo privileges taken for granted by most U.S. civilians, subsuming their individual autonomy to an institutional authority, whose structure of rewards and punishments is designed to foster conformity and collectivism. For example, despite military policy ostensibly allowing lesbian and gay service members to openly serve, previous research ( McNamara et al., 2021 ) has suggested that these individuals may distrust the policy and still feel pressured to hide their sexual orientation due to their perception that military culture implicitly does not allow such differences and they could be punished as a result. Military culture has also been found to be at odds with non-normative gender and sexuality identities ( Meyer et al., 2022 ). Military Deployment (or Service Without Deployment): Forming Lifelong Bonds of Trust Third, combat-related deployment forms powerful, lifelong bonds of brotherhood and effects drastic psychological changes, many life-transforming ( Castro et al., 2015 ). For some, deployment combines the stress of being in a foreign land for "endless hours in a miserable work environment" with occasional moments of existential "terror and destruction of human life" ( Coll et al., 2011 , p. 491). The intensity of the experience and the bonds forged therein further reinforce the perception of necessary sacrifice and trusting only others who share this culture. These conditions can cause Veterans to continue having mistrust toward those who are not in-group members, a form of psychological hypervigilance wherein they adopt the narrative that "if you've never been there, you wouldn't understand" ( H. A. Garcia, 2017 ). Even many Veterans who have never experienced foreign deployment in a theater of war go through this inculcation into the military culture and establish strong cohesive bonds with other team members, making these military-forged relationships highly salient ( Coll et al., 2011 ). Long after separation, these bonds can continue to be meaningful, as Veterans often evaluate the trustworthiness of information and people through the lens of shared Veteran identity; that is, they may view information from fellow Veterans with greater trust than information from non-Veterans ( Alenkin, 2015 ). Separation and Navigating Culture and Identity Transition Fourth, military separation is often accompanied by a sense of disillusionment. Some Veterans return with a loss of trust in the U.S. government ( Coll et al., 2011 , p. 488). After being psychologically embedded within the military culture for years, separation typically engenders a type of "culture shock" and disorientation ( Coll et al., 2011 , p. 488). "No one leaves unchanged" is often the description given to the experience of separation after military service. After the socialization and experience of military service, fully transitioning to being a civilian may be very challenging, if not impossible for some Veterans. Separation confers on Veterans a level of autonomy they gave up when they joined the military; upon separating Veterans are now free to make many decisions over their body, including health care choices as well as food, hairstyle, and daily attire. However, this can also be overwhelming and for some feel like a letdown. The disillusionment and loss of raison d'etre may cause Veterans to question their own self-worth and feel that they are simultaneously isolated and only able to connect with others who share their military identity. Veterans transitioning from military to civilian roles must negotiate their social identity based on their membership in various social groups and the value that they attach to those memberships ( Thompson et al., 2017 ). For some, the process can involve difficulties in negotiating between their perceived military versus civilian group allegiances, leading to isolation and mental health challenges ( Castro et al., 2015 ; Coll et al., 2011 ; Kranke et al., 2016 ), as has been established in interviews with Veteran volunteers active in Team Rubicon ( Kranke et al., 2016 ). Veterans struggling with such reintegration challenges are particularly distrustful of public health initiatives wherein the voices of Veterans are largely absent ( Franco et al., 2021 ) due to gaps between military and civilian experiences that exacerbate Veterans' perception of a divide that can only often be bridged by Veteran peers ( Hobbs, 2008 ). For some Veterans, mistrust of civilians, whom they perceive to be incapable of comprehending the scale of their sacrifice and lacking similar loyalty and commitment, remains after separation ( H. A. Garcia, 2017 ). Additional detail on the separation process, as discussed by Veterans participating in TR ( Kranke et al., 2016 ), is provided in the discussion section. Veteran's Military Culture, Health Decision-Making, and the COVID-19 Vaccine This article focuses on Veterans' military culture to identify how the lived experience of being acculturated into the military, followed by separation, leads some group members to view themselves as separate and apart from mainstream U.S. society, with their own unique subculture. For some members, this status permeates many aspects of their lives, especially their relationship with and perception of the U.S. government and government health care such as the VA. The effects of Veteran experiences and culture on their health decision-making have been established in existing literature, particularly in work in the field of Veteran mental health ( Borah et al., 2021 ; Coll et al., 2011 ; Hobbs, 2008 ). In examining the impact of this subculture on Veterans' narratives around their COVID-19 vaccine decisions, we argue that a more empathetic, culturally competent approach to engaging with this group would be more effective in building trust, both for the COVID pandemic and for future health campaigns and health emergencies. Past research has identified differences in health-seeking behavior among Veterans, which may suggest attitudinal factors affecting their decisions. Borah et al. (2021) identify some possible adverse consequences of communication breakdowns that may result from Veterans encountering health care that is not informed by military culture. When some Veterans feel that their concerns are not adequately heard or addressed, they may agree to comply with provider recommendations and then later fail to follow through due to their conditioning of obedience to authority although they may not buy into recommended health behaviors ( Borah et al., 2021 ). Heslin et al. (2013) found that Veterans' levels of trust in local health departments were markedly different from those of non-Veterans, especially among Vietnam-era Veterans expressing cynicism toward the government. Kranke et al. (2016) suggest that shared cultural values of self-reliance or "a bootstrap mentality" instilled during military service can hinder their willingness to seek or accept recommended mental health services. Mental health providers are therefore advised to better have cultural competence in military culture to better empathize with Veterans and thus "get through to them" more effectively within the therapeutic relationship ( Borah et al., 2021 ; Castro et al., 2015 ; Coll et al., 2011 ; Meyer et al., 2022 ). However, this body of work does not fully speak to how Veterans' military culture might shape perceptions toward COVID-19 vaccination. In this article, we argue that the role of Veteran culture has been neglected in both research on Veterans' vaccine behavior and efforts to encourage vaccine uptake among Veterans. Using data from interviews with Veterans in VA homeless programs, this article seeks to better understand how Veterans, especially those who have experienced challenges with maintaining stable housing, respond to, and interpret the COVID-19 vaccines. It aims to thoroughly unpack the role of Veteran experience and culture in an effort to better understand how campaigns to encourage vaccine uptake can gain greater credibility within the Veteran population. Entering the Military: A New Relationship With the Government First, giving up one's existing civilian identity and joining the military involves entering a long-term, complex relationship with the government ( Hobbs, 2008 ; Jasuja et al., 2021 ; Rein, 2021 ) and accepting the U.S. government's authority over oneself and one's own body. In some cases, this involves forgoing bodily autonomy and informed consent over medical and other decisions. In particular, military members are not afforded the same option of informed consent that civilians enjoy, as established in the United States through the 1979 Belmont Report, over their acceptance of investigational drugs and vaccines ( Black, 2007 ; Rettig, 1999 ; U.S. Government Accountability Office [GAO], 2002 ): A service member "freely agrees when joining the military to relinquish autonomy" to the interests of the unit or mission. Furthermore, they are aware "as they approach actual fighting, their autonomy dramatically decreases." ( Rettig, 1999 : p. 50) Individual informed consent was not required in the military's anthrax vaccination program ( Black, 2007 ; GAO, 2002 ) during post-9/11 military campaigns in Iraq and Afghanistan nor for investigational drugs such as pyridostigmine bromide and botulinum toxin administered during the 1991 Gulf War in Iraq ( Rettig, 1999 ). Military service also includes the historical legacy of exposures to environmental hazards for which compensation for service-connected disability was previously denied by the U.S. government. Currently, the VA is studying the long-term health effects of such exposures as Agent Orange in Vietnam, and respiratory illnesses from burn pits while serving in Iraq and Afghanistan ( U.S. Department of Veterans Affairs, 2022 ), and has subsequently recognized the health effects of these exposures as service-connected disabilities. As of April 2022, respiratory illnesses connected to exposure to airborne hazards in burn pits in the Middle East were eligible for presumed service-connected disabilities ( U.S. Department of Veterans Affairs, 2022 ). Acculturation: Adopting Military Cultural Values as One's Own Second, joining the military, individuals often take on a new identity ( Coll et al., 2011 , p. 489). Military cultural values such as "service, competence, sacrifice, mission, and collectivism" ( Meyer et al., 2022 ) are deeply ingrained within the identity of military members. Military members forgo privileges taken for granted by most U.S. civilians, subsuming their individual autonomy to an institutional authority, whose structure of rewards and punishments is designed to foster conformity and collectivism. For example, despite military policy ostensibly allowing lesbian and gay service members to openly serve, previous research ( McNamara et al., 2021 ) has suggested that these individuals may distrust the policy and still feel pressured to hide their sexual orientation due to their perception that military culture implicitly does not allow such differences and they could be punished as a result. Military culture has also been found to be at odds with non-normative gender and sexuality identities ( Meyer et al., 2022 ). Military Deployment (or Service Without Deployment): Forming Lifelong Bonds of Trust Third, combat-related deployment forms powerful, lifelong bonds of brotherhood and effects drastic psychological changes, many life-transforming ( Castro et al., 2015 ). For some, deployment combines the stress of being in a foreign land for "endless hours in a miserable work environment" with occasional moments of existential "terror and destruction of human life" ( Coll et al., 2011 , p. 491). The intensity of the experience and the bonds forged therein further reinforce the perception of necessary sacrifice and trusting only others who share this culture. These conditions can cause Veterans to continue having mistrust toward those who are not in-group members, a form of psychological hypervigilance wherein they adopt the narrative that "if you've never been there, you wouldn't understand" ( H. A. Garcia, 2017 ). Even many Veterans who have never experienced foreign deployment in a theater of war go through this inculcation into the military culture and establish strong cohesive bonds with other team members, making these military-forged relationships highly salient ( Coll et al., 2011 ). Long after separation, these bonds can continue to be meaningful, as Veterans often evaluate the trustworthiness of information and people through the lens of shared Veteran identity; that is, they may view information from fellow Veterans with greater trust than information from non-Veterans ( Alenkin, 2015 ). Separation and Navigating Culture and Identity Transition Fourth, military separation is often accompanied by a sense of disillusionment. Some Veterans return with a loss of trust in the U.S. government ( Coll et al., 2011 , p. 488). After being psychologically embedded within the military culture for years, separation typically engenders a type of "culture shock" and disorientation ( Coll et al., 2011 , p. 488). "No one leaves unchanged" is often the description given to the experience of separation after military service. After the socialization and experience of military service, fully transitioning to being a civilian may be very challenging, if not impossible for some Veterans. Separation confers on Veterans a level of autonomy they gave up when they joined the military; upon separating Veterans are now free to make many decisions over their body, including health care choices as well as food, hairstyle, and daily attire. However, this can also be overwhelming and for some feel like a letdown. The disillusionment and loss of raison d'etre may cause Veterans to question their own self-worth and feel that they are simultaneously isolated and only able to connect with others who share their military identity. Veterans transitioning from military to civilian roles must negotiate their social identity based on their membership in various social groups and the value that they attach to those memberships ( Thompson et al., 2017 ). For some, the process can involve difficulties in negotiating between their perceived military versus civilian group allegiances, leading to isolation and mental health challenges ( Castro et al., 2015 ; Coll et al., 2011 ; Kranke et al., 2016 ), as has been established in interviews with Veteran volunteers active in Team Rubicon ( Kranke et al., 2016 ). Veterans struggling with such reintegration challenges are particularly distrustful of public health initiatives wherein the voices of Veterans are largely absent ( Franco et al., 2021 ) due to gaps between military and civilian experiences that exacerbate Veterans' perception of a divide that can only often be bridged by Veteran peers ( Hobbs, 2008 ). For some Veterans, mistrust of civilians, whom they perceive to be incapable of comprehending the scale of their sacrifice and lacking similar loyalty and commitment, remains after separation ( H. A. Garcia, 2017 ). Additional detail on the separation process, as discussed by Veterans participating in TR ( Kranke et al., 2016 ), is provided in the discussion section. Veteran's Military Culture, Health Decision-Making, and the COVID-19 Vaccine This article focuses on Veterans' military culture to identify how the lived experience of being acculturated into the military, followed by separation, leads some group members to view themselves as separate and apart from mainstream U.S. society, with their own unique subculture. For some members, this status permeates many aspects of their lives, especially their relationship with and perception of the U.S. government and government health care such as the VA. The effects of Veteran experiences and culture on their health decision-making have been established in existing literature, particularly in work in the field of Veteran mental health ( Borah et al., 2021 ; Coll et al., 2011 ; Hobbs, 2008 ). In examining the impact of this subculture on Veterans' narratives around their COVID-19 vaccine decisions, we argue that a more empathetic, culturally competent approach to engaging with this group would be more effective in building trust, both for the COVID pandemic and for future health campaigns and health emergencies. Past research has identified differences in health-seeking behavior among Veterans, which may suggest attitudinal factors affecting their decisions. Borah et al. (2021) identify some possible adverse consequences of communication breakdowns that may result from Veterans encountering health care that is not informed by military culture. When some Veterans feel that their concerns are not adequately heard or addressed, they may agree to comply with provider recommendations and then later fail to follow through due to their conditioning of obedience to authority although they may not buy into recommended health behaviors ( Borah et al., 2021 ). Heslin et al. (2013) found that Veterans' levels of trust in local health departments were markedly different from those of non-Veterans, especially among Vietnam-era Veterans expressing cynicism toward the government. Kranke et al. (2016) suggest that shared cultural values of self-reliance or "a bootstrap mentality" instilled during military service can hinder their willingness to seek or accept recommended mental health services. Mental health providers are therefore advised to better have cultural competence in military culture to better empathize with Veterans and thus "get through to them" more effectively within the therapeutic relationship ( Borah et al., 2021 ; Castro et al., 2015 ; Coll et al., 2011 ; Meyer et al., 2022 ). However, this body of work does not fully speak to how Veterans' military culture might shape perceptions toward COVID-19 vaccination. In this article, we argue that the role of Veteran culture has been neglected in both research on Veterans' vaccine behavior and efforts to encourage vaccine uptake among Veterans. Using data from interviews with Veterans in VA homeless programs, this article seeks to better understand how Veterans, especially those who have experienced challenges with maintaining stable housing, respond to, and interpret the COVID-19 vaccines. It aims to thoroughly unpack the role of Veteran experience and culture in an effort to better understand how campaigns to encourage vaccine uptake can gain greater credibility within the Veteran population. Approach and Method This article presents data and findings from a study focused on vaccination attitudes among Veterans enrolled in VA homeless transitional housing. This study was reviewed and approved by the Institutional Review Board at the VA Greater Los Angeles Health Care System. Data were drawn from semi-structured telephone interviews conducted between January and April 2021 with 20 Veterans residing in the VA's Grant and Per Diem (GPD) program, which funds nonprofit organizations providing transitional housing to Veterans experiencing homelessness. Only Veterans participating in VA GPD programs were eligible to be recruited for this study. Interviews were also conducted with staff leaders at seven GPD organizations and VA staff members who assisted with recruitment by inviting Veteran residents to sign up to participate in the study. Further details regarding the recruitment methods, interview guide, and data collection and analyses are available elsewhere ( Gin, Balut, & Dobalian, 2022 ). Digital audio recordings were transcribed verbatim, and transcripts were analyzed thematically. Themes were reported based on substantive significance ( Patton, 2002 , 2014 ), or how the themes inform the existing literature. As there is a paucity of research on homeless Veterans' vaccine attitudes, we first conducted open coding to identify emerging themes. Data corresponding to these themes were sorted by shared content using the "sort and sift method" ( Maietta et al., 2021 ) pulling together interview passages with shared meaning and adopting a phenomenological approach to interpreting Veterans' narratives regarding their life experiences. We coded Veterans' reactions to questions regarding their willingness to get vaccinated; the codes reported in this article regarding veterans' military identities were emergent codes identified through inductive coding, allowing patterns in the codes to emerge. Two researchers coded a single interview to cross-check for interrater reliability. Researchers then coded the remaining interviews independently and reviewed coded text, using the constant comparative method ( Boeije, 2002 ), to sort the codes by shared findings (e.g., "mandatory military vaccines" and "distrust of government."). The resulting group codes were compared, contrasted, and sorted into themes. When differences in the analysis occurred, the individuals on the research team provided justification for their coding scheme, and the theme was adjusted so that it encompassed the additional components identified. From this, a master codebook was created and used for the analysis of the final interviews. To ensure that our interpretation and understanding of Veterans' lived reality accurately represents Veterans' perspectives, we solicited input from a VA Veteran Peer Specialist to help in triangulating our findings and contextualizing our data. This Veteran, who is identified in the acknowledgements, served in the U.S. Marine Corps from 1997 to 2001. Peer Specialists are Veterans "with mental health experience—actively engaged in his or her own recovery—trained and certified to help other Veterans" ( U.S. Department of Veterans Affairs, 2018 ). As important allies for Veterans, they bring the added advantage of familiarity with the experiences of other Veterans, as well as their own. We also examine our identities in the research process: two authors identify as "military-connected"; by military-connected, we mean having a service member or veteran that is a close family relation. None of the authors has personally served in the U.S. military. However, all four authors have over a decade of experience each working on Veterans' issues in both our research and practice, and we are all committed to understanding the lived realities of Veterans, particularly those who experience challenges in their return to civilian life. In conducting and interpreting interviews, the authors sought to present empathetic understandings of Veterans' experiential realities from their vantage points and perspectives ( Meyer et al., 2022 ), in keeping with phenomenological research and practice traditions in qualitative research methods and humanistic psychology, respectively. Findings and Analysis All 20 Veterans interviewed were living in VA GPD facilities, were male, and were between 29 and 65 years old, with most in their 40s and 50s. Eleven were White, five were people of color, representing Black, Native American, Hispanic, and Biracial identities, and four declined to state their racial/ethnic group. Veterans were not asked any questions about the military, their service, nor their identities as Veterans. Rather, these topics emerged as they shared their stories. Veteran culture themes were spontaneously mentioned by 11 of 20 Veterans, alluding to its salience. Three themes emerged: (1) mandatory military vaccination and bodily autonomy; (2) distrust and cynicism toward the government; and (3) reliance on Veteran peers in vaccine decisions. Mandatory Military Vaccination and Bodily Autonomy Veterans who served in the post-9/11conflicts mentioned mandatory vaccines such as the anthrax vaccination program and the influenza vaccines. One Veteran, who was eager to get the COVID vaccine, rationalized it in terms that mirrored his military thinking: when somebody offers me a vaccine, I'm like all right I'll get it. I'm prepared. . . I'm used to being a guinea pig. I was in the Army active for three and a half years, and then I did nine and a half years total, the reserve and active. . . You know people sticking needles in me, telling me, all right you need to take this, and I'm like okay. (MA4) His willingness to accept vaccines was partly based on his experience with mandatory vaccines in the military, that is, desensitization. However, his narrative is also tinged with cynicism, conveying skepticism about complying with taking medicine that he barely understands and portraying himself as a passive subject, complying with having needles stuck into him. His perception of the experimental aspects of the anthrax vaccine that he received continues: Basically, what they did is they gave everybody this vaccine to test it out. Because that's what they do. They give it to prisoners and military members to see what exactly is gonna happen over a period of time because they're government property. (MA4) Veterans' quotes about mandatory vaccination in the military involved processing what it means to receive these vaccines then, and the implications of getting the COVID-19 vaccine now. Another post 9/11 era Veteran explains his views on getting vaccines: every Veteran I've talked to about this is, the way in which the military treats vaccines is you get desensitized to the idea of the government putting things in your body. I've had anthrax, Hep A, Hep B. . .—things I can't even pronounce, pumped into me. And I'm still alive. I just don't worry about vaccines. Nor does any Veteran, that I know. (IA1) Rather than "taking medicines," which would involve active participation, he emphasizes that he cannot even pronounce the names of the medicines that are being "pumped into" his body. His cynicism about the passive nature of his identity when getting vaccines underscores the lack of informed consent involved. This example frames the vaccination process in terms that evoke a lack of free agency and choice in vaccination and would be considered an example of obedience without evidence of buy-in ( Borah et al., 2021 ). Distrust and Cynicism Toward the Government Both Veterans who got the vaccine and those who refused expressed strong distrust and cynicism toward the government, but these sentiments were almost universal among vaccine refusers. The same post 9/11 era Veteran whose rationale in getting vaccinated was quoted above said elsewhere in his interview: I don't trust the information they give me. Because I understand their track record. And I was in the military too long to trust what anybody said to me. . . They give me a piece of information and like okay I take it at face value. And then I go, and I find out that it's wrong, and now I'm not pleased but that's what governments do, they lie to people to make them feel better. . . (MA4) Despite accepting the vaccine, he expresses an unequivocal worldview of cynicism and distrust, suggesting that Veterans with such beliefs are unlikely to view government authorities as trusted information sources. He later cynically states that he does not trust the VA: The VA was made to mend the broken toy soldiers that the military made. And for a very long time they were not good at their job. They're getting better. The amount of care that they're offering now because of the Gulf War and (post 9/11 era) Veterans. . . So the VA is forced to become better at their job. . . that's why I don't trust the VA. You know the military they give you a piece of gear and it's made by the lowest bidder. But they're getting better, and I know a lot of people in the system that are trying to do good. (MA4) The reference to "broken toy soldiers" alludes to the loss of autonomy and sense of betrayal that often occurs when Veterans reflect on the experience of giving up their individual identities to join the military, only to enter the VA in need of care often perceived as substandard, bureaucratic, and insufficiently nimble and responsive to their needs ( Franco et al., 2021 ). The historical legacy of exposure to environmental and toxic health hazards in the military was another source of mistrust that Veterans mentioned. A Veteran in his 50s who was opposed to vaccination mentioned that he was enrolled in a VA study on the long-term health effects of being stationed at Camp Lejeune, a Marine base where drinking water was found contaminated with toxic chemicals for more than 30 years: So, I don't really trust the government when it comes to saying everything is all right when it comes to medicine for patients. I'm also in the study of the Camp Lejeune. . . And they're . . . still refusing to step up . . . and they're not taking care of people that were exposed to it. (MA1) He also mentioned the exposure to Agent Orange during the Vietnam War as a reason for his mistrust and cynicism toward information from government institutions about health and safety: FDA is saying oh, they're good, but we've heard those things before. With Agent Orange guys from Vietnam, you know, 50 years before the government said yeah, there was a problem, and by that time, most of those guys had died terrible deaths. (MA1) Interestingly, although the Veteran (MA1) served a decade after the Vietnam War, the historical legacy of Agent Orange was salient to him, possibly because of his status as a Veteran enrolled in the study on long-term health effects of environmental exposure. Another Veteran (MA4) from the post-9/11era mentioned burn pits in the Middle East as another military environmental exposure that contributed to his cynicism and distrust of the government: There were consequences . . . Just like the heavy metals we all ingested from the burn pits when we were deployed. Raised the rates of cancer. Stillbirth with children. (MA4) Burn pits, Agent Orange, and Camp Lejeune are all hazardous exposures whose long-term health impacts the VA is following for possible linkages to cancer or death (U.S. Department of Veterans Affairs Public Health website, April 26, 2022). Moreover, these are all exposures for which the U.S. government significantly delayed approving the payment of service-connected disability benefits due to the lengthy scientific review needed to establish a connection between military toxic exposures and cancer diagnoses ( Shane, 2021 ). These narratives, all of which involved the VA's system of awarding benefits based on documented service-connected harm, undoubtedly contribute to government mistrust and cynicism among Veterans, particularly Veterans enrolled in VA Homeless Programs who may have had the experience of having to apply for VA benefits. Belief in government conspiracy theories related to COVID-19 and vaccines revealed another dimension of Veterans' distrust: their relationship with government power. None of the Veterans explicitly mentioned power, but one GPD staff director and one VA provider touched on it in their discussion of Veterans' mistrust of the VA. The GPD staff director, who was a post 9/11 era Veteran himself, noted that his generation of Veterans do not trust the VA: my generation of Veterans and younger have a cultural distrust of the VA. That's something I work very hard to try and chip away at. But there's a cultural mistrust of the VA, so they're not going into the VA to be counted . . . Which is unfounded. But it's perpetuated. So it's very institutionalized among Vets that they don't trust the VA. (NJ) This quote suggests that many of his peer generation of Veterans refuse to go into the VA to receive services and "be counted" as an act of resistance. Refusing may be a form of asserting one's autonomy against feeling constantly surveilled during one's service. This suggests that Veterans experiencing homelessness who do not use the VA may be especially resistant to a government authority, and possibly vaccination, due to even higher distrust levels. The power struggle aspect of Veterans' resistance to vaccination due to distrust of the government is amplified in the following VA staff member's discussion of why the VA homeless health clinic tells Veterans that they are not required to get vaccinated to receive care: when there's not a lot of pressure to get the vaccine, like 'if you want it we'll give it you, if you don't that's fine, we're still going to provide you services,' kind of reassuring them of that, that takes some of the power struggle out of it. Because there's a lot of people that think like 'the government's not going to force me to do anything,' and the minute we said 'no we're not going to require you to get the vaccine, if you don't get it that's your choice, you get to make that decision,' and so things calm down pretty quickly (ND) The prospect of being forcibly compelled to accept medical treatment against one's will has, for some Veterans, special meaning, and may be connected to other aspects of loss of autonomy during their service. Government entities, particularly the VA, may be viewed with suspicion by some Veterans and thus may not always be effective direct messengers. However, many Veterans who developed strong bonds with peers during their service trust their Veteran peers. Reliance on Veteran Peers Looking to others who are either Veterans or military-connected in making decisions about their own vaccine acceptance was mentioned by both Veterans and GPD staff interviewed. Others who share the Veteran identity tended to be trusted over health authorities or civilians. Some Veterans mentioned that talking to Veteran peers who were vaccinated helped persuade them, while one respondent noted that his GPD counselor was the son of a Vietnam Veteran and thus trusted his advice on the vaccine. The GPD staff director, who was a post 9/1 era Veteran, noted that his generation of younger Veterans would be more likely to get vaccinated if they heard the message from other Veterans, particularly of their generation, due to their distrust of the VA: a Veteran will always listen to another Veteran. There will always be . . . an inherent trust . . . the majority of Veterans trust each other. . . because there's that sense of you know we have each other's back no matter what. . . (NJ) This quote exemplifies the rationale undergirding Veterans' tendency to rely on peers with military backgrounds in making personal health decisions: Others with a shared identity and group membership are trusted, because of the military cultural norm that service members will always protect and look out for each other. Prior research identifies Veterans' tendency to rely on Veteran peers in decision-making on health behaviors ( Alenkin, 2015 ). This ability to trust other Veterans for information is especially vital for individuals who, as described, may not trust the VA as a source of health information. Mandatory Military Vaccination and Bodily Autonomy Veterans who served in the post-9/11conflicts mentioned mandatory vaccines such as the anthrax vaccination program and the influenza vaccines. One Veteran, who was eager to get the COVID vaccine, rationalized it in terms that mirrored his military thinking: when somebody offers me a vaccine, I'm like all right I'll get it. I'm prepared. . . I'm used to being a guinea pig. I was in the Army active for three and a half years, and then I did nine and a half years total, the reserve and active. . . You know people sticking needles in me, telling me, all right you need to take this, and I'm like okay. (MA4) His willingness to accept vaccines was partly based on his experience with mandatory vaccines in the military, that is, desensitization. However, his narrative is also tinged with cynicism, conveying skepticism about complying with taking medicine that he barely understands and portraying himself as a passive subject, complying with having needles stuck into him. His perception of the experimental aspects of the anthrax vaccine that he received continues: Basically, what they did is they gave everybody this vaccine to test it out. Because that's what they do. They give it to prisoners and military members to see what exactly is gonna happen over a period of time because they're government property. (MA4) Veterans' quotes about mandatory vaccination in the military involved processing what it means to receive these vaccines then, and the implications of getting the COVID-19 vaccine now. Another post 9/11 era Veteran explains his views on getting vaccines: every Veteran I've talked to about this is, the way in which the military treats vaccines is you get desensitized to the idea of the government putting things in your body. I've had anthrax, Hep A, Hep B. . .—things I can't even pronounce, pumped into me. And I'm still alive. I just don't worry about vaccines. Nor does any Veteran, that I know. (IA1) Rather than "taking medicines," which would involve active participation, he emphasizes that he cannot even pronounce the names of the medicines that are being "pumped into" his body. His cynicism about the passive nature of his identity when getting vaccines underscores the lack of informed consent involved. This example frames the vaccination process in terms that evoke a lack of free agency and choice in vaccination and would be considered an example of obedience without evidence of buy-in ( Borah et al., 2021 ). Distrust and Cynicism Toward the Government Both Veterans who got the vaccine and those who refused expressed strong distrust and cynicism toward the government, but these sentiments were almost universal among vaccine refusers. The same post 9/11 era Veteran whose rationale in getting vaccinated was quoted above said elsewhere in his interview: I don't trust the information they give me. Because I understand their track record. And I was in the military too long to trust what anybody said to me. . . They give me a piece of information and like okay I take it at face value. And then I go, and I find out that it's wrong, and now I'm not pleased but that's what governments do, they lie to people to make them feel better. . . (MA4) Despite accepting the vaccine, he expresses an unequivocal worldview of cynicism and distrust, suggesting that Veterans with such beliefs are unlikely to view government authorities as trusted information sources. He later cynically states that he does not trust the VA: The VA was made to mend the broken toy soldiers that the military made. And for a very long time they were not good at their job. They're getting better. The amount of care that they're offering now because of the Gulf War and (post 9/11 era) Veterans. . . So the VA is forced to become better at their job. . . that's why I don't trust the VA. You know the military they give you a piece of gear and it's made by the lowest bidder. But they're getting better, and I know a lot of people in the system that are trying to do good. (MA4) The reference to "broken toy soldiers" alludes to the loss of autonomy and sense of betrayal that often occurs when Veterans reflect on the experience of giving up their individual identities to join the military, only to enter the VA in need of care often perceived as substandard, bureaucratic, and insufficiently nimble and responsive to their needs ( Franco et al., 2021 ). The historical legacy of exposure to environmental and toxic health hazards in the military was another source of mistrust that Veterans mentioned. A Veteran in his 50s who was opposed to vaccination mentioned that he was enrolled in a VA study on the long-term health effects of being stationed at Camp Lejeune, a Marine base where drinking water was found contaminated with toxic chemicals for more than 30 years: So, I don't really trust the government when it comes to saying everything is all right when it comes to medicine for patients. I'm also in the study of the Camp Lejeune. . . And they're . . . still refusing to step up . . . and they're not taking care of people that were exposed to it. (MA1) He also mentioned the exposure to Agent Orange during the Vietnam War as a reason for his mistrust and cynicism toward information from government institutions about health and safety: FDA is saying oh, they're good, but we've heard those things before. With Agent Orange guys from Vietnam, you know, 50 years before the government said yeah, there was a problem, and by that time, most of those guys had died terrible deaths. (MA1) Interestingly, although the Veteran (MA1) served a decade after the Vietnam War, the historical legacy of Agent Orange was salient to him, possibly because of his status as a Veteran enrolled in the study on long-term health effects of environmental exposure. Another Veteran (MA4) from the post-9/11era mentioned burn pits in the Middle East as another military environmental exposure that contributed to his cynicism and distrust of the government: There were consequences . . . Just like the heavy metals we all ingested from the burn pits when we were deployed. Raised the rates of cancer. Stillbirth with children. (MA4) Burn pits, Agent Orange, and Camp Lejeune are all hazardous exposures whose long-term health impacts the VA is following for possible linkages to cancer or death (U.S. Department of Veterans Affairs Public Health website, April 26, 2022). Moreover, these are all exposures for which the U.S. government significantly delayed approving the payment of service-connected disability benefits due to the lengthy scientific review needed to establish a connection between military toxic exposures and cancer diagnoses ( Shane, 2021 ). These narratives, all of which involved the VA's system of awarding benefits based on documented service-connected harm, undoubtedly contribute to government mistrust and cynicism among Veterans, particularly Veterans enrolled in VA Homeless Programs who may have had the experience of having to apply for VA benefits. Belief in government conspiracy theories related to COVID-19 and vaccines revealed another dimension of Veterans' distrust: their relationship with government power. None of the Veterans explicitly mentioned power, but one GPD staff director and one VA provider touched on it in their discussion of Veterans' mistrust of the VA. The GPD staff director, who was a post 9/11 era Veteran himself, noted that his generation of Veterans do not trust the VA: my generation of Veterans and younger have a cultural distrust of the VA. That's something I work very hard to try and chip away at. But there's a cultural mistrust of the VA, so they're not going into the VA to be counted . . . Which is unfounded. But it's perpetuated. So it's very institutionalized among Vets that they don't trust the VA. (NJ) This quote suggests that many of his peer generation of Veterans refuse to go into the VA to receive services and "be counted" as an act of resistance. Refusing may be a form of asserting one's autonomy against feeling constantly surveilled during one's service. This suggests that Veterans experiencing homelessness who do not use the VA may be especially resistant to a government authority, and possibly vaccination, due to even higher distrust levels. The power struggle aspect of Veterans' resistance to vaccination due to distrust of the government is amplified in the following VA staff member's discussion of why the VA homeless health clinic tells Veterans that they are not required to get vaccinated to receive care: when there's not a lot of pressure to get the vaccine, like 'if you want it we'll give it you, if you don't that's fine, we're still going to provide you services,' kind of reassuring them of that, that takes some of the power struggle out of it. Because there's a lot of people that think like 'the government's not going to force me to do anything,' and the minute we said 'no we're not going to require you to get the vaccine, if you don't get it that's your choice, you get to make that decision,' and so things calm down pretty quickly (ND) The prospect of being forcibly compelled to accept medical treatment against one's will has, for some Veterans, special meaning, and may be connected to other aspects of loss of autonomy during their service. Government entities, particularly the VA, may be viewed with suspicion by some Veterans and thus may not always be effective direct messengers. However, many Veterans who developed strong bonds with peers during their service trust their Veteran peers. Reliance on Veteran Peers Looking to others who are either Veterans or military-connected in making decisions about their own vaccine acceptance was mentioned by both Veterans and GPD staff interviewed. Others who share the Veteran identity tended to be trusted over health authorities or civilians. Some Veterans mentioned that talking to Veteran peers who were vaccinated helped persuade them, while one respondent noted that his GPD counselor was the son of a Vietnam Veteran and thus trusted his advice on the vaccine. The GPD staff director, who was a post 9/1 era Veteran, noted that his generation of younger Veterans would be more likely to get vaccinated if they heard the message from other Veterans, particularly of their generation, due to their distrust of the VA: a Veteran will always listen to another Veteran. There will always be . . . an inherent trust . . . the majority of Veterans trust each other. . . because there's that sense of you know we have each other's back no matter what. . . (NJ) This quote exemplifies the rationale undergirding Veterans' tendency to rely on peers with military backgrounds in making personal health decisions: Others with a shared identity and group membership are trusted, because of the military cultural norm that service members will always protect and look out for each other. Prior research identifies Veterans' tendency to rely on Veteran peers in decision-making on health behaviors ( Alenkin, 2015 ). This ability to trust other Veterans for information is especially vital for individuals who, as described, may not trust the VA as a source of health information. Discussion and Implications For many Veterans, the influence of military culture continues after their return to civilian life. Military culture, and the effects of military separation, shapes how many Veterans perceive and respond to COVID-19 vaccination. These data suggest that for many Veterans with high levels of mistrust, decisions regarding the uptake of COVID-19 vaccines are inextricably bound up with their complex relationship with both the U.S. government ( Jasuja et al., 2021 ; Meyer et al., 2022 ; Rein, 2021 ) and their attitudes toward civilian life in general. Because nearly all Veterans have experienced mandatory vaccinations in the military, the experience of joining and later separating from the military involves shifting definitions of bodily autonomy they must navigate to varying degrees. Some respondents view vaccines promoted by government health care institutions from the same perspective they may have developed while in the military—that of a passive subject accepting that "government knows best," while others may be determined to reclaim their autonomy by questioning or refusing vaccinations. Moreover, Veterans who have high levels of civilian mistrust and face reintegration challenges are likely to transfer that distrust to vaccination campaigns, especially if vaccines are being promoted by a government entity that has earned a reputational distrust among Veterans. This is likely true even if that entity, in the case of the VA, has improved its quality of care and responsiveness to service-connected health conditions such as posttraumatic stress disorder (PTSD) and Agent Orange, as many Veterans often rely on peers rather than official government notices in health care decision-making ( Alenkin, 2015 ). Much of this distrust revolves around power and autonomy; hence vaccine promotion interventions that seem too "heavy-handed" to some, such as mandates, may appear coercive to Veterans. When evaluating whether to get vaccinated, Veterans often rely on peers as trusted sources to help with decisions. This group identity aspect of Veterans only trusting other Veterans derives from their military experience, where they are inculcated into a culture where trusting only in-group members becomes a survival strategy ( H. A. Garcia, 2017 ). A culturally competent approach to engaging with Veterans around health care questions such as COVID-19 vaccination would not only incorporate such aspects of Veteran culture but also engage Veterans in the process of creating Veteran vaccine outreach campaigns. When designing health care intervention and messaging efforts aimed at Veterans, the phrase "nothing about us without us" may be instructive, given the significant trust and culture gaps between Veterans and civilians. Public health campaigns designed without their involvement and input may not resonate, perpetuating the potential disconnect ( Borah et al., 2021 ; Franco et al., 2021 ) and exacerbating prospective inequities in marginalized groups such as unhoused Veterans. Given many Veterans' trust barriers with civilians and government health care systems, how might a perspective informed by person-centered interventions and empathy toward the lived experiences of subpopulations suggest that outreach efforts regarding health behaviors to Veterans be designed? Team Rubicon (TR), a nonprofit disaster relief organization of 130,000 mostly Veteran members ( Team Rubicon website, n.d. ), offers some possible answers that point to best practices for engagement. Team Rubicon As part of the Biden Administration's COVID-19 Community Corps, Team Rubicon (TR) launched a 2021 awareness campaign to facilitate vaccine uptake among vaccine-reluctant Veterans ( Callaway, 2021 ; Team Rubicon, 2021 ; U.S. Department of Health & Human Services, 2021 ). As of August 2021, TR has distributed more than 1.6 vaccines to Veterans in nearly 100 cities ( Slack, 2021 ). They use the slogan: a "call to arms-yours"—immediately invoking military identity in their vaccination push. Even before the COVID-19 pandemic, TR was well-known among Veterans because of its efforts to harness the unique skills and talents of recently separated military Veterans as disaster relief volunteers. In so doing, TR addresses the crisis of identity that many recently separated Veterans experience. TR's founders and leaders, themselves recently separated post-9/11 era Veterans, had credibility among Veterans because they empathize with their identity and experiences. Initially founded to reduce the high prevalence of Veteran suicides, TR sought to address post-9/11 Veterans' post-separation struggles with the feelings of loss of purpose and belonging that often drive their difficulties with reintegration and contribute to their feelings of civilian mistrust ( H. A. Garcia, 2017 ; Kranke et al., 2016 ; Lawrence et al., 2019 ; Matthieu et al., 2018 , 2021 ). Thus, TR embodies the very challenges of addressing Veterans' struggles with trust by (a) enabling them to form bonds of brotherhood with Veteran peers who are both trusted and often able to relate to reintegration challenges and (b) valuing their military identity by having them apply skills learned during military service in service of a collective raison d'etre benefiting others, usually civilians—a feeling often lost during separation. A recent study on TR ( Kranke et al., 2016 , 2017 ) revealed the ways in which TR connected with Veterans' identities and struggles with reintegration, thus addressing their mistrust and solidifying TR as a credible entity among Veterans. These interviews with Veterans illuminate how an empathetic understanding of Veterans' institutionalized distrust and disconnect with civilian society can be applied to engagement in public health efforts such as encouraging vaccine uptake: 1. Loss of identity and sense of isolation upon military separation. Veterans discussed how separation from the military left them feeling like the civilian life they returned to was underwhelming in comparison: I believed we had a real purpose that was just and worth fighting for. . . Life here isn't as compelling as life there. . . It's hard to come home and be motivated and play the same game when the game is one-tenth of the stakes you're used to playing ( Kranke et al., 2016 , p. 77) after you've belonged to something as, you know, as powerful and as meaningful and respectful as the United States Marine Corps, and then you're looking at going and getting a job at Best Buy. You know, it's not the fact that it's the work at Best Buy. It's just like that's—that's you now? ( Kranke et al., 2016 , p. 78) These two quotes exemplify Veterans' sense of disillusionment at the loss of purpose realized upon returning home, separating their identity from that of civilians. 2. Enabling them to form bonds of brotherhood with Veteran peers. The disconnect and loss of identity fuels a sense of mistrust among civilians, including non-Veteran health care providers, whom they perceive as incapable of understanding the worldviews of Veterans because they have not gone through the same experiences as Veterans ( Coll et al., 2011 ; H. A. Garcia, 2017 ). Given this loss of identity, some Veterans may feel disconnected from civilian friends and family and trust and relate better to Veteran peers. TR enables them to be with Veteran peers and recreate the powerful bonds of brotherhood first developed in the military ( Castro et al., 2015 ): it's the interpersonal relationships that develop between us, and it's a bond that you can't explain unless you are a Veteran. And the trust that you know, you may not have ever met this person before, but you know he has your back. ( Kranke et al., 2016 , p. 78) The trust of Veteran peers, even those they have never met, above all others, due to shared military culture of "having each other's backs" is repeatedly invoked by both TR Veterans. This bond cannot ever be fully understood unless someone shares the Veteran identity, underscoring the level of in-group culture of trust that TR evokes while convening Veterans. 3. Re-establishing the value of military culture and identity through the application of military service skills in a collective mission to benefit others. TR's trust-building in Veterans goes beyond identity and emotions. Its disaster relief mission aligns with the skill set that individuals previously deployed with the U.S. military have homed in their service, enabling them to recover their sense of meaning: a lot of us are wounded combat Vets. . . We have a set of skills Uncle Sam paid for and gave us that we have an obligation to give back to our community. ( Kranke et al., 2016 , p.78) This empowerment narrative is core to Veterans' identity and self-esteem. It is an important component of Veterans' military culture, which discourages appearing weak or dependent and values collective sacrifice and contribution to the greater good ( Castro et al., 2015 ; Coll et al., 2011 ; Meyer et al., 2022 ). The desire to identify as a contributor, not a victim, and to de-emphasize one's sacrifice for country is part of the "combat Veteran paradox" ( Castro et al., 2015 ). This enables Veterans to reclaim their own agency, which is essential for building trust. 4. Normalizing Veterans' experience as a core aspect of separating from the military. If military culture, and the process of military separation, consists of a core set of common processes that are normal and likely expected aspects of the Veteran journey, then these shared experiences, which set Veterans apart from civilians, can be transformed from distrust and cynicism into a basis for commonality. The need to normalize these aspects of Veteran identity has been discussed extensively in the military mental health literature ( Castro et al., 2015 ; Coll et al., 2011 ; H. A. Garcia, 2017 ) and is illustrated here: [Team Rubicon lets] "us see that . . . is a very common experience among vets. So, it's not us. You are not individually damaged; you are reacting predictably and rationally to a set of experiences that you've been put in. ( Kranke et al., 2016 , p. 79) Normalizing Veterans' feelings misunderstood and disconnected from civilians upon separation ( H. A. Garcia, 2017 ) helps Veterans address their mistrust by helping them find acceptance and reduce their self-perceived stigma ( Kranke et al., 2016 ). Through these processes of creating empathetic space for Veterans to express their shared reality, TR effectively connects with their life experiences. Thus, Veterans begin forming a basis for trust, if only in other Veterans at first. TR's focus on building these empathetic relationships lends them a degree of authenticity, cultural competence, and trust in the eyes of Veterans. Thus, organizations like TR, which are led by Veterans and focus on addressing their unique experiences and cultural perspectives, are ideal candidates to engage with Veterans on COVID-19 vaccination. In TR's vaccine awareness campaign, they proudly display the images of men and women from iconic past trust campaigns enlisting arms of every kind: the tatted, the toned, the tanned, and the sun-deprived ( www.teamrubiconusa.org ), with the slogan: "Rolling up sleeves can help us defeat the virus." TR, as a Veteran-run organization, illustrates how entities that exemplify credibility within the Veteran community can serve as a possible approach to offer culturally competent outreach to Veterans. Veteran engagement initiatives, such as the Milwaukee VA's Veteran-run Dryhootch initiative, can leverage the power of Veterans' trust in their peers to offer a culturally resonant perspective on vaccination. For example, COVID-19 vaccination promotion campaigns could have Veterans, through social media and video testimony, serve as spokespeople to amplify the message of encouraging Veteran audiences to get vaccinated. As Franco et al. (2021) note, the absence of the voice of Veterans in health care initiatives is a pervasive barrier to more effective public health responses. Acknowledgment of historical and current sources of Veteran mistrust and disconnect from health care providers that do not provide military-informed care or care that is not informed by Veteran-specific health disparities is essential to pre-empting Veterans' sense that they are either invisible or not valued. Other avenues of conveying such empathetic, culturally competent health behavior advocacy to Veterans include VA Veteran peer support specialists, non-profit housing providers such as GPD organizations, and VA health care providers who work in specialty clinics such as the Homeless Patient Aligned Care Team ( Balut et al., 2021 ; Gin, Balut, Alenkin, & Dobalian, 2022 ). VA service providers who understand that there is a "power struggle" component to some Veterans' vaccination attitudes can be helpful by enabling Veterans to feel "listened to," as one VA provider phrased it. Given the power of those with an insiders' perspective into Veteran experience and knowledge, Peer Specialists in the VA and other health care organizations could be specifically trained to build trust in COVID-19 vaccines. Having trained Veteran peer specialists to engage in public health outreach to other Veterans would be consonant with the promotora community health outreach worker model of non-judgmental negotiation between cultures to engage communities with health information ( Deitrick et al., 2010 ; Lujan, 2009 ). Currently, Peer Specialists working with Veterans in VA homeless programs may individually opt to initiate conversations to encourage Veterans to get vaccinated, but this practice is not formally institutionalized within VA. Limitations This study's understanding of the effect of Veterans' military culture on vaccine attitudes and behavior may be limited by the fact that researchers did not initially ask Veterans about the role of military culture in their vaccination decisions. This theme was identified inductively, as more than half of the Veterans interviewed brought up their military service as a core factor in their personal identities and their thinking about vaccines. Consequently, information about respondents' branch of service, era, rank, and file, or prior experience with military vaccinations, which may have offered valuable insight into respondents' military experiences, was not collected. This calls for future research to enhance understanding about how Veterans experiencing reintegration challenges view health care efforts to promote COVID-19 vaccination. Furthermore, this research was confined to a population of Veterans experiencing homelessness who are enrolled in VA housing services. Expanding such a study to other Veteran groups, including post-9/11-era Veterans—a demographic that participates in TR in significant numbers, Peer Support Specialists, or others with experience with PTSD or other mental health challenges, could possibly offer additional insight into the relationship between Veterans' post-separation culture and COVID-19 vaccine attitudes. Generalizability and transferability of these results are also limited. This study evaluated the perspectives of Veterans in VA-funded GPD programs, data that is triangulated through interviews with staff in these nonprofit programs. Given the trust issues inherent in this population, we cannot ascertain whether unhoused Veterans who are not participating in VA homeless programs could be persuaded by even the most culturally competent Veteran community engagement initiatives. However, GPD-enrolled Veterans had higher vaccination rates than homeless Veterans who were not enrolled in GPD housing ( Balut et al., 2021 ) suggesting that the Veterans interviewed may be more open to vaccination than other unhoused Veterans. Furthermore, we cannot be certain whether Veterans who have other reintegration-related mental health experiences such as PTSD, would respond similarly if asked about vaccination or other recommended health behaviors. The TR case study suggests, however, that these findings may be transferable to Veteran populations with experiences other than homelessness. Team Rubicon As part of the Biden Administration's COVID-19 Community Corps, Team Rubicon (TR) launched a 2021 awareness campaign to facilitate vaccine uptake among vaccine-reluctant Veterans ( Callaway, 2021 ; Team Rubicon, 2021 ; U.S. Department of Health & Human Services, 2021 ). As of August 2021, TR has distributed more than 1.6 vaccines to Veterans in nearly 100 cities ( Slack, 2021 ). They use the slogan: a "call to arms-yours"—immediately invoking military identity in their vaccination push. Even before the COVID-19 pandemic, TR was well-known among Veterans because of its efforts to harness the unique skills and talents of recently separated military Veterans as disaster relief volunteers. In so doing, TR addresses the crisis of identity that many recently separated Veterans experience. TR's founders and leaders, themselves recently separated post-9/11 era Veterans, had credibility among Veterans because they empathize with their identity and experiences. Initially founded to reduce the high prevalence of Veteran suicides, TR sought to address post-9/11 Veterans' post-separation struggles with the feelings of loss of purpose and belonging that often drive their difficulties with reintegration and contribute to their feelings of civilian mistrust ( H. A. Garcia, 2017 ; Kranke et al., 2016 ; Lawrence et al., 2019 ; Matthieu et al., 2018 , 2021 ). Thus, TR embodies the very challenges of addressing Veterans' struggles with trust by (a) enabling them to form bonds of brotherhood with Veteran peers who are both trusted and often able to relate to reintegration challenges and (b) valuing their military identity by having them apply skills learned during military service in service of a collective raison d'etre benefiting others, usually civilians—a feeling often lost during separation. A recent study on TR ( Kranke et al., 2016 , 2017 ) revealed the ways in which TR connected with Veterans' identities and struggles with reintegration, thus addressing their mistrust and solidifying TR as a credible entity among Veterans. These interviews with Veterans illuminate how an empathetic understanding of Veterans' institutionalized distrust and disconnect with civilian society can be applied to engagement in public health efforts such as encouraging vaccine uptake: 1. Loss of identity and sense of isolation upon military separation. Veterans discussed how separation from the military left them feeling like the civilian life they returned to was underwhelming in comparison: I believed we had a real purpose that was just and worth fighting for. . . Life here isn't as compelling as life there. . . It's hard to come home and be motivated and play the same game when the game is one-tenth of the stakes you're used to playing ( Kranke et al., 2016 , p. 77) after you've belonged to something as, you know, as powerful and as meaningful and respectful as the United States Marine Corps, and then you're looking at going and getting a job at Best Buy. You know, it's not the fact that it's the work at Best Buy. It's just like that's—that's you now? ( Kranke et al., 2016 , p. 78) These two quotes exemplify Veterans' sense of disillusionment at the loss of purpose realized upon returning home, separating their identity from that of civilians. 2. Enabling them to form bonds of brotherhood with Veteran peers. The disconnect and loss of identity fuels a sense of mistrust among civilians, including non-Veteran health care providers, whom they perceive as incapable of understanding the worldviews of Veterans because they have not gone through the same experiences as Veterans ( Coll et al., 2011 ; H. A. Garcia, 2017 ). Given this loss of identity, some Veterans may feel disconnected from civilian friends and family and trust and relate better to Veteran peers. TR enables them to be with Veteran peers and recreate the powerful bonds of brotherhood first developed in the military ( Castro et al., 2015 ): it's the interpersonal relationships that develop between us, and it's a bond that you can't explain unless you are a Veteran. And the trust that you know, you may not have ever met this person before, but you know he has your back. ( Kranke et al., 2016 , p. 78) The trust of Veteran peers, even those they have never met, above all others, due to shared military culture of "having each other's backs" is repeatedly invoked by both TR Veterans. This bond cannot ever be fully understood unless someone shares the Veteran identity, underscoring the level of in-group culture of trust that TR evokes while convening Veterans. 3. Re-establishing the value of military culture and identity through the application of military service skills in a collective mission to benefit others. TR's trust-building in Veterans goes beyond identity and emotions. Its disaster relief mission aligns with the skill set that individuals previously deployed with the U.S. military have homed in their service, enabling them to recover their sense of meaning: a lot of us are wounded combat Vets. . . We have a set of skills Uncle Sam paid for and gave us that we have an obligation to give back to our community. ( Kranke et al., 2016 , p.78) This empowerment narrative is core to Veterans' identity and self-esteem. It is an important component of Veterans' military culture, which discourages appearing weak or dependent and values collective sacrifice and contribution to the greater good ( Castro et al., 2015 ; Coll et al., 2011 ; Meyer et al., 2022 ). The desire to identify as a contributor, not a victim, and to de-emphasize one's sacrifice for country is part of the "combat Veteran paradox" ( Castro et al., 2015 ). This enables Veterans to reclaim their own agency, which is essential for building trust. 4. Normalizing Veterans' experience as a core aspect of separating from the military. If military culture, and the process of military separation, consists of a core set of common processes that are normal and likely expected aspects of the Veteran journey, then these shared experiences, which set Veterans apart from civilians, can be transformed from distrust and cynicism into a basis for commonality. The need to normalize these aspects of Veteran identity has been discussed extensively in the military mental health literature ( Castro et al., 2015 ; Coll et al., 2011 ; H. A. Garcia, 2017 ) and is illustrated here: [Team Rubicon lets] "us see that . . . is a very common experience among vets. So, it's not us. You are not individually damaged; you are reacting predictably and rationally to a set of experiences that you've been put in. ( Kranke et al., 2016 , p. 79) Normalizing Veterans' feelings misunderstood and disconnected from civilians upon separation ( H. A. Garcia, 2017 ) helps Veterans address their mistrust by helping them find acceptance and reduce their self-perceived stigma ( Kranke et al., 2016 ). Through these processes of creating empathetic space for Veterans to express their shared reality, TR effectively connects with their life experiences. Thus, Veterans begin forming a basis for trust, if only in other Veterans at first. TR's focus on building these empathetic relationships lends them a degree of authenticity, cultural competence, and trust in the eyes of Veterans. Thus, organizations like TR, which are led by Veterans and focus on addressing their unique experiences and cultural perspectives, are ideal candidates to engage with Veterans on COVID-19 vaccination. In TR's vaccine awareness campaign, they proudly display the images of men and women from iconic past trust campaigns enlisting arms of every kind: the tatted, the toned, the tanned, and the sun-deprived ( www.teamrubiconusa.org ), with the slogan: "Rolling up sleeves can help us defeat the virus." TR, as a Veteran-run organization, illustrates how entities that exemplify credibility within the Veteran community can serve as a possible approach to offer culturally competent outreach to Veterans. Veteran engagement initiatives, such as the Milwaukee VA's Veteran-run Dryhootch initiative, can leverage the power of Veterans' trust in their peers to offer a culturally resonant perspective on vaccination. For example, COVID-19 vaccination promotion campaigns could have Veterans, through social media and video testimony, serve as spokespeople to amplify the message of encouraging Veteran audiences to get vaccinated. As Franco et al. (2021) note, the absence of the voice of Veterans in health care initiatives is a pervasive barrier to more effective public health responses. Acknowledgment of historical and current sources of Veteran mistrust and disconnect from health care providers that do not provide military-informed care or care that is not informed by Veteran-specific health disparities is essential to pre-empting Veterans' sense that they are either invisible or not valued. Other avenues of conveying such empathetic, culturally competent health behavior advocacy to Veterans include VA Veteran peer support specialists, non-profit housing providers such as GPD organizations, and VA health care providers who work in specialty clinics such as the Homeless Patient Aligned Care Team ( Balut et al., 2021 ; Gin, Balut, Alenkin, & Dobalian, 2022 ). VA service providers who understand that there is a "power struggle" component to some Veterans' vaccination attitudes can be helpful by enabling Veterans to feel "listened to," as one VA provider phrased it. Given the power of those with an insiders' perspective into Veteran experience and knowledge, Peer Specialists in the VA and other health care organizations could be specifically trained to build trust in COVID-19 vaccines. Having trained Veteran peer specialists to engage in public health outreach to other Veterans would be consonant with the promotora community health outreach worker model of non-judgmental negotiation between cultures to engage communities with health information ( Deitrick et al., 2010 ; Lujan, 2009 ). Currently, Peer Specialists working with Veterans in VA homeless programs may individually opt to initiate conversations to encourage Veterans to get vaccinated, but this practice is not formally institutionalized within VA. Limitations This study's understanding of the effect of Veterans' military culture on vaccine attitudes and behavior may be limited by the fact that researchers did not initially ask Veterans about the role of military culture in their vaccination decisions. This theme was identified inductively, as more than half of the Veterans interviewed brought up their military service as a core factor in their personal identities and their thinking about vaccines. Consequently, information about respondents' branch of service, era, rank, and file, or prior experience with military vaccinations, which may have offered valuable insight into respondents' military experiences, was not collected. This calls for future research to enhance understanding about how Veterans experiencing reintegration challenges view health care efforts to promote COVID-19 vaccination. Furthermore, this research was confined to a population of Veterans experiencing homelessness who are enrolled in VA housing services. Expanding such a study to other Veteran groups, including post-9/11-era Veterans—a demographic that participates in TR in significant numbers, Peer Support Specialists, or others with experience with PTSD or other mental health challenges, could possibly offer additional insight into the relationship between Veterans' post-separation culture and COVID-19 vaccine attitudes. Generalizability and transferability of these results are also limited. This study evaluated the perspectives of Veterans in VA-funded GPD programs, data that is triangulated through interviews with staff in these nonprofit programs. Given the trust issues inherent in this population, we cannot ascertain whether unhoused Veterans who are not participating in VA homeless programs could be persuaded by even the most culturally competent Veteran community engagement initiatives. However, GPD-enrolled Veterans had higher vaccination rates than homeless Veterans who were not enrolled in GPD housing ( Balut et al., 2021 ) suggesting that the Veterans interviewed may be more open to vaccination than other unhoused Veterans. Furthermore, we cannot be certain whether Veterans who have other reintegration-related mental health experiences such as PTSD, would respond similarly if asked about vaccination or other recommended health behaviors. The TR case study suggests, however, that these findings may be transferable to Veteran populations with experiences other than homelessness. Conclusion Being a Veteran means one's military identity never fully goes away. Veterans' identities, defined as the ways in which those separated from military experience filter temporal experiences of the social world, are liminal. Liminal identities Veterans inhabit complicate seamless transitions to civilian life. Identity is reconstructed not as fully civilian or entirely military. This was the expressed reality for both Veterans experiencing homelessness and recently discharged post-9/11 Veterans who turned to TR for a needed sense of shared community. Shared military identity is likely to be more salient for Veterans residing in VA-funded homeless transitional housing, or who go on TR missions because these are Veterans who have chosen to recreate experiences of communal living with other Veterans. Understanding the experience of being a Veteran underscores the importance of having trusted messengers serve as spokespeople to encourage Veterans to get vaccinated for COVID-19. While no cultural group has a single message that resonates with all members because culture is inherently diverse and diffuse, the discussion and data presented here sufficiently illustrate that Veterans have shared cultural elements stemming from common life experiences to warrant the beginning of a conversation around what a culturally competent Veteran engagement campaign would be needed to build their trust, either for the current pandemic or for a future public health emergency. Public health institutions and others seeking to increase vaccination uptake have been criticized for approaching people with health vulnerabilities as "objects of intervention." Such a tactic is likely to meet with resistance from Veterans who already have a cynical view toward those who treat them as passive patients without a sense of agency. Amplifying the voice of the Veteran in vaccination campaigns would enable them to retain a sense of group autonomy in this important decision. A more nuanced empathetic approach that meets Veterans where they are, without judgment, and honoring the realities of their service and reintegration efforts, is needed.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3433916/

Restricting Fermentative Potential by Proteome Remodeling

Pathogenesis hinges on successful colonization of the gastrointestinal (GI) tract by pathogenic facultative anaerobes. The GI tract is a carbohydrate-limited environment with varying oxygen availability and oxidoreduction potential (ORP). How pathogenic bacteria are able to adapt and grow in these varying conditions remains a key fundamental question. Here, we designed a system biology-inspired approach to pinpoint the key regulators allowing Bacillus cereus to survive and grow efficiently under low ORP anoxic conditions mimicking those encountered in the intestinal lumen. We assessed the proteome components using high throughput nanoLC-MS/MS techniques, reconstituted the main metabolic circuits, constructed Δ ohrA and Δ ohrR mutants, and analyzed the impacts of ohrA and ohrR disruptions by a novel round of shotgun proteomics. Our study revealed that OhrR and OhrA are crucial to the successful adaptation of B. cereus to the GI tract environment. Specifically, we showed that B. cereus restricts its fermentative growth under low ORP anaerobiosis and sustains efficient aerobic respiratory metabolism, motility, and stress response via OhrRA-dependent proteome remodeling. Finally, our results introduced a new adaptive strategy where facultative anaerobes prefer to restrict their fermentative potential for a long term benefit.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7330614/

Ethno-medicinal and bio-cultural importance of aloes from south and east of the Great Rift Valley floristic regions of Ethiopia

There are 46 Aloe species identified from Ethiopia out of which 67.3% are endemics but comprehensive data on their ethno-medicinal and bio-cultural values are lacking. Interview, focus group discussion (FGD), and guided field walks were conducted with 210 respondents (152 men and 58 women). Relative frequency of citation (RFC s ), informants' consensus factor (F ic ), use value (UV s ), relative importance index (RI s ), and cultural value index (CV) were analyzed. Non-parametric Kruskal Wallis and Wilcoxon tests were performed using R software. Twenty-three Aloe species were recorded in the study areas with 196 use-reports and 2158 citations, grouped into six major use categories (N UC = 6). Medicinal use categories accounted for 149 use-reports (76%) with 1607 citations. The species with the highest numbers of use-reports were Aloe megalacantha subsp. alticola , A. trichosantha subsp. longiflora and A. calidophila of which 87, 75 and 61.1% respectively were medicinal uses. Aloe calidophila has highest values in all indices UV (11.72), RFC (0.68), RI (0.89), and CV (6.2). Among Aloe parts, leaf exudate accounted for 111 use-reports (49.1%) of which 92.9% were used for medicinal purposes. Aloe retrospiciens and A. ruspoliana were reported poisonous to carnivores. F ic values of the six major use categories ranged from 0.86 to 0.22. Elderly people (>60) had more knowledge than 25–40 and 41–60 age groups (Kruskal-Wallis chi-squared = 12.17, df = 3, p = 0.006), which is significant difference in depth of ethno-medicinal knowledge. Men had more knowledge of medicinal uses than women (Wilcoxon test, p = 0.002) significantly different, while women were knowledgeable than men for cultural uses like, cosmetic (Wilcoxon test, p = 0.06), not significantly different. The ways in which aloes are used and valued have implications for their future medicinal utility, which instigate detailed phytochemical and pharmacological studies. 1 Introduction The genus Aloe L. belongs to family Asphodelaceae [ 1 ], which has 560 accepted species and 21 infraspecific taxa [ 2 ] is renowned for its use in herbal medicine throughout its native range in Africa South of the Sahara, the Arabian Peninsula, Madagascar and the Mascarene Islands [ 3 ]. The dried latex extracted from the leaves has been used medicinally in Africa, Asia, Europe and the Middle East for hundreds of years [ 4 , 5 , 6 , 7 , 8 ]. Aloes has been used for the treatment of wounds and skin complaints, malaria, microbial infections, and complaints of the digestive system [ 9 , 10 , 11 , 12 , 13 ]. In addition, commercial preparations containing Aloe species include laxative drugs, health drinks and tonics, after shaving gel, mouthwash and toothpaste, hair tonic and shampoo, and skin-moistening gel [ 14 , 15 , 16 ]. Ethiopia, a likely centre of diversity for Aloe , has 46 identified and documented species with three subspecies, of which 67.3% are endemic. Several of these are highly threatened [ 17 , 18 ]. Three local centres of endemism are recognised in Ethiopia, each with characteristic endemic species ( Table 1 ) [ 18 , 19 ]. Six species of Aloe in Ethiopia were classified on the IUCN Red List [ 20 ] as endangered ( A. harlana and A. yavellana ), near threatened ( A. tewoldei and A. pubescens ), and vulnerable ( A. retrospiciens and A. rugosifolia ), while many of the remaining species are still data-deficient [ 21 , 22 , 23 , 24 ]. Despite conservation concerns, aloes have been recognised for their economic potential in Ethiopia [ 19 ], particularly for livelihood security, economic development and enhancing biodiversity conservation on marginal lands [ 25 ]. For example, Aloe debrana leaf mesophyll is used in a thickening agent [ 26 ] and for treating sisal fibre for packing Ethiopian export coffee (e.g. www.gseventiplc.com ). In the Borena of Oromia region in Southeast Ethiopia, the local community has been cultivating A. calidophila to collect aloe leaf exudate and gel for small-scale aloe-soap manufacturing [ 27 ]. Data collated from the literature indicated that the medicinal uses of the genus Aloe comprised 74% of the use records [ 28 ] in which the use of exudate was common. Table 1 The three local centres of Aloe endemism recognised in Ethiopia. Table 1 Center of endemism Floristic regions Number of endemic species List of endemic species and infraspecific taxa Northern and central highlands, north and west of the great rift valley SU, KF, IL, WG, GJ, GD, WU, TU 15 Aloe adigratana, A. ankoberensis, A. benishangulana, A. camperi, A. clarkei, A. debrana, A. elegans, A. monticola, A. percrassa, A. pulcherrima, A. schelpei, A. sinana, A. steudneri, A. trigonantha , and A. weloensis Eastern highlands and lowlands AF, HA 7 Aloe bertemariae , A. harlana, A. mcloughlinii, A. megalacantha subsp. alticola, A. pirottae ; A. pubescens ; and A. trichosantha subsp. longiflora Southern highlands, lowlands and rift valley AR, BA, SD, GG 11 Aloe elkerriana, A. friisii, A. ghibensis, A. gilbertii subsp. gilbertii, A. gilbertii subsp. megalacanthoides, A. jacksonii, A. kefaensis, A. otallensis, A. tewoldei, A. welmelensis, and A. yavellana SU = Shew; KF = Kefa; IL = Ilubabur; WG = Welega; GJ = Gojam; GD = Gonder; WU = Welo; TU = Tigray; AF = Afar; HA = Harargae; AR = Arsi; BA = Bale; SD = Sidamo; GG = Gamo Gofa. Despite the limited reports on the uses and conservation concerns for aloes of Ethiopia, their ethnomedicinal and biocultural values, and the impact of these values on sustainable use have not previously been assessed systematically. It is expected that people will be motivated to conserve resources that are most important to them, in contrast to resources perceived as less useful [ 29 , 30 , 31 ]. In this regard, further effort to improve the perception of local community towards the resource is needed. This can be achieved by indicating the invisible but potential values of the genus Aloe for effective conservation of the genus Aloe . In addition, it has been hypothesized that the local communities in all the study areas make use of Aloe species arbitrarily for similar purpose and irrespective of different cultural/ethinic communities on the use of Aloe species. Therefore, we comprehensively investigated ethnomedicinal values, biocultural importance, and the emic perception of the wild population status of Aloe species in the East and South of the Great Rift Valley floristic region of Ethiopia. 2 Materials and methods 2.1 Study areas This study focused on the eastern, southeastern, and southern floristic regions of Ethiopia called Hararge (HA), Bale (BA), Sidamo (SD), Arsi (AR), and Afar (AF) floristic regions, which are stated as east and south of the Great Rift Valley floristic regions in this study ( Figure 1 ). Figure 1 The study areas, stated as east and south of the Great Rift Valley floristic regions in this study (Site numbers indicated Aloe species: 1 = A. calidophila; 2 = A. citrina; 3 = A. gilbertii subsp. gilbertii; 4 = A. harlana; 5 = A. lateritia; 6 = A. macrocarpa; 7 = A. mcloughlinii; 8 = A. megalacantha subsp. alticola; 9 = A. megalacantha subsp. megalacantha; 10 = A. otallensis; 11 = A. pirottae; 12 = A. pubescens; 13 = A. retrospiciens; 14 = A. rivae; 15 = A. rugosifolia; 16 = A. ruspoliana; 17 = A. secundiflora; 18 = A. tewoldei; 19 = A. trichsantha subsp. longiflora; 20 = A. welmelensis; 21 = A. yavellana; 22 = AHU51; 23 = AHU53). Figure 1 2.2 Data collection Comprehensive data was collected on the ethno-medicinal and bio-cultural values, population trends of species in the natural habitat from the emic perspective, and associated indigenous knowledge of Aloe species in these floristic regions. Data was collected in the different seasons from 2017 (March to April and October to November) to 2018 (January to February and June to July). A total of 210 respondents (152 men and 58 women) from four cultural communities (Oromo, Somali, Afar, and Harari) participated. Informants were either randomly chosen (RI) or were systematically selected traditional healers (TH/key informants) who have in-depth traditional knowledge concerning multi-utility of Aloe species of their locality. Age, gender, and occupation ( Table 2 ) were considered, as suggested by Martin [ 32 ] and Caruso et al , [ 33 ]. Table 2 Demographic characteristics of informants representing ethnic communities in east and south of the Great Rift Valley Floristic regions, Ethiopia. Table 2 Total Informant Oromo (N = 140) Somali (N = 32) Afar (N = 24) Harari (N = 14) Men 97 (69.3%) 25 (78.1%) 19 (79.2%) 11 (78.6%) Women 43 (30.7%) 7 (21.9%) 5 (20.8%) 3 (21.4%) RI Men 60 16 14 8 Women 33 5 4 2 TH/key informants Men 37 9 5 3 Women 10 2 1 1 Age categories ∗ 25–40 50 (32.8 ± 5.3) 9 (31.1 ± 5.5) 4 (30.3 ± 4.0) 4 (32.6 ± 4.1) 41–60 65 (51.7 ± 5.8) 20 (50.3 ± 6.8) 15 (51.9 ± 5.9) 8 (51.1 ± 6.0) above 60 25 (70.1 ± 6.4) 3 (76.0 ± 5.7) 5 (74.5 ± 6.6) 2 (72.0 ± 5.2) Occupation Farmers 104 0 0 6 Agro-pastoralist 12 4 2 0 Pastoralist 18 26 22 0 Others 6 2 0 8 ∗ Number of informants = N (mean age ±standard deviation) for each age categories; RI = random informants, TH = traditional healers. Ethnobotanical data collection procedures were approved by ethical committee for human involvement and use of lab animals in school of animal and range sciences of Haramaya University. Semi-structured interviews, focus group discussion (FGD), practical observation sessions, and guided field walk in Aloe localities were conducted after Oral Prior Informed Consent (PIC) was sought from every respondent. Most interviews were conducted in the field in order to avoid the risk of confusing identity of Aloe species. By repeated inquiries at least two times with the same informants the validity and reliability of recorded information was confirmed [ 32 , 33 , 34 ]. The respondents were asked to freely list all possible uses of each Aloe species, each time a plant was mentioned as "used" was considered as one "use-report", and repeated mention of same use-report by different informants was taken as "use mention or number of citation". Data recorded were: vernacular names of Aloe species, local uses, parts used, ingredients added during the use formulations (if any), and locally marketable aloe products produced. For medicinal use-reports, the ailments treated, preparation procedures, method of administration, and antidote (if any) were recorded. In addition, population trends (noticeably increasing, increasing, noticeably decreasing, decreasing, stable, and not sure/uncertain) from emic perspective were recorded (Appendix II). However, vague use-reports, from which it was difficult to distinguish the specific values were avoided, e.g. "used for livestock disease treatment"; "it has cultural importance"; etc. Finally, the use-reports recorded were categorized into major use categories (UC) of level 1 and sub-categories of level 2 using the Economic Botany Data Standard [ 35 ]. Aloe specimens were identified using taxonomic keys in the Flora of Ethiopia and Eritrea [ 17 ], through visual comparisons with authenticated plant specimens kept at the National Herbarium (ETH) of Addis Ababa University and at Herbarium of Haramaya University, and authenticated by Prof. Sebsebe Demissew (Professor of Plant Systematics and Biodiversity). Voucher specimens of all species with all herbarium sheet data were deposited at both herbaria. Voucher numbers of Haramaya University were used in this study. 2.3 Data analysis The data were organized and cleaned in an Excel spreadsheet (Microsoft office 2016), to be suitable for both qualitative and quantitative analysis. Qualitative data were analyzed following [ 32 , 34 , 36 ] like, free listing, part used, use-reports and use mentions. Quantitative data were analyzed using the corresponding formulas as follows: 2.3.1 Frequency of citation for use-report (FC UR ) Frequency of citation for use-report (FC UR ) is the percentage of informants who mentioned each use-report of particular Aloe species, which has been calculated using the formula: (1) FC UR ( % ) = n i / N × 100 where n i is the number of informants who cites each use-report per species and N is total number of informants. 2.3.2 Relative frequency of citation (RFC s ) Relative frequency of citation was calculated by dividing the frequency of citation (FC) of a species by total number of informants (N) involved in the five floristic regions that makes N to varies or the summation of use-report (∑UR i ) of all the informants interviewed for a species divided by N [ 37 ]. (2) RFC s = FC s N = ∑ i = i 1 i N UR i / N 2.3.3 Informant consensus factor (F ic ) The informant consensus factor (Fic) of each Aloe species is the proportion of informants who independently reported its use against a particular use category calculated using the formula [ 38 , 39 ]: (3) Fic = n ur − nt / n ur − 1 where, 'n ur ' is the "number of use-reports" in each use category and 'nt' is the "number of taxa used". 2.3.4 Use value (UV) Use value was used to demonstrate the relative importance of each Aloe species known locally, which can be calculated according to Albuquerque et al. formula [ 40 ]: (4) UV s = ∑ U i / N s where, UV s refers to the use value of Aloe species 's', U i to the number of different uses mentioned by each informant i per specific Aloe species, and N is the total number of informants interviewed for Aloe species 's'. 2.3.5 Relative importance index (RI) This index takes into account the number of major use-categories only and calculated as follows [ 37 ]: (5) RI s = RFCs ( max ) + RNUs ( max ) 2 where, RFC s (max) is the relative frequency of citation over the maximum. It is obtained by dividing FC s by the maximum value in all Aloe species of the study [RFC s(max) = FCs/max (FC)], and RNUs(max) is the relative number of use-categories over the maximum, obtained dividing the number of uses of the species NU s = u = u NC u = u 1 ∑ UR u by the maximum value in all Aloe species of the survey RNs(max) = NUs/max (NU). 2.3.6 Cultural value index (CV) This index estimates the cultural significance of each Aloe species, which combines the three variables, informant (i), a species (s), and use-category (u), which is calculated using the following formula [ 41 ]: (6) CV s = [ NU s NC ] [ FC s N ] [ ∑ U = U i uNC ∑ i = i 1 i N UR u 1 N ] The first factor is the relationship between the numbers of different uses reported for ethnospecies (each Aloe species ) and total number of use-categories. The second factor is the relative frequency of citation of a species. The third factor is the sum of all the UR for a species, i.e., the sum of number of participants who mentioned each use of a species, divided by N. To test if there was any correlation between age of the informant and their knowledge on use of aloes (number of use-reports), the nonparametric Kruskal Wallis Test was performed. If there was a significant difference between the informant's gender and knowledge about use of aloes, the non-parametric pair wise Wilcoxon test was performed using R software version 3.3.4. for Windows using multicompview and R companion packages. P-values of less than 0.05 were taken as statistically significant difference. 2.1 Study areas This study focused on the eastern, southeastern, and southern floristic regions of Ethiopia called Hararge (HA), Bale (BA), Sidamo (SD), Arsi (AR), and Afar (AF) floristic regions, which are stated as east and south of the Great Rift Valley floristic regions in this study ( Figure 1 ). Figure 1 The study areas, stated as east and south of the Great Rift Valley floristic regions in this study (Site numbers indicated Aloe species: 1 = A. calidophila; 2 = A. citrina; 3 = A. gilbertii subsp. gilbertii; 4 = A. harlana; 5 = A. lateritia; 6 = A. macrocarpa; 7 = A. mcloughlinii; 8 = A. megalacantha subsp. alticola; 9 = A. megalacantha subsp. megalacantha; 10 = A. otallensis; 11 = A. pirottae; 12 = A. pubescens; 13 = A. retrospiciens; 14 = A. rivae; 15 = A. rugosifolia; 16 = A. ruspoliana; 17 = A. secundiflora; 18 = A. tewoldei; 19 = A. trichsantha subsp. longiflora; 20 = A. welmelensis; 21 = A. yavellana; 22 = AHU51; 23 = AHU53). Figure 1 2.2 Data collection Comprehensive data was collected on the ethno-medicinal and bio-cultural values, population trends of species in the natural habitat from the emic perspective, and associated indigenous knowledge of Aloe species in these floristic regions. Data was collected in the different seasons from 2017 (March to April and October to November) to 2018 (January to February and June to July). A total of 210 respondents (152 men and 58 women) from four cultural communities (Oromo, Somali, Afar, and Harari) participated. Informants were either randomly chosen (RI) or were systematically selected traditional healers (TH/key informants) who have in-depth traditional knowledge concerning multi-utility of Aloe species of their locality. Age, gender, and occupation ( Table 2 ) were considered, as suggested by Martin [ 32 ] and Caruso et al , [ 33 ]. Table 2 Demographic characteristics of informants representing ethnic communities in east and south of the Great Rift Valley Floristic regions, Ethiopia. Table 2 Total Informant Oromo (N = 140) Somali (N = 32) Afar (N = 24) Harari (N = 14) Men 97 (69.3%) 25 (78.1%) 19 (79.2%) 11 (78.6%) Women 43 (30.7%) 7 (21.9%) 5 (20.8%) 3 (21.4%) RI Men 60 16 14 8 Women 33 5 4 2 TH/key informants Men 37 9 5 3 Women 10 2 1 1 Age categories ∗ 25–40 50 (32.8 ± 5.3) 9 (31.1 ± 5.5) 4 (30.3 ± 4.0) 4 (32.6 ± 4.1) 41–60 65 (51.7 ± 5.8) 20 (50.3 ± 6.8) 15 (51.9 ± 5.9) 8 (51.1 ± 6.0) above 60 25 (70.1 ± 6.4) 3 (76.0 ± 5.7) 5 (74.5 ± 6.6) 2 (72.0 ± 5.2) Occupation Farmers 104 0 0 6 Agro-pastoralist 12 4 2 0 Pastoralist 18 26 22 0 Others 6 2 0 8 ∗ Number of informants = N (mean age ±standard deviation) for each age categories; RI = random informants, TH = traditional healers. Ethnobotanical data collection procedures were approved by ethical committee for human involvement and use of lab animals in school of animal and range sciences of Haramaya University. Semi-structured interviews, focus group discussion (FGD), practical observation sessions, and guided field walk in Aloe localities were conducted after Oral Prior Informed Consent (PIC) was sought from every respondent. Most interviews were conducted in the field in order to avoid the risk of confusing identity of Aloe species. By repeated inquiries at least two times with the same informants the validity and reliability of recorded information was confirmed [ 32 , 33 , 34 ]. The respondents were asked to freely list all possible uses of each Aloe species, each time a plant was mentioned as "used" was considered as one "use-report", and repeated mention of same use-report by different informants was taken as "use mention or number of citation". Data recorded were: vernacular names of Aloe species, local uses, parts used, ingredients added during the use formulations (if any), and locally marketable aloe products produced. For medicinal use-reports, the ailments treated, preparation procedures, method of administration, and antidote (if any) were recorded. In addition, population trends (noticeably increasing, increasing, noticeably decreasing, decreasing, stable, and not sure/uncertain) from emic perspective were recorded (Appendix II). However, vague use-reports, from which it was difficult to distinguish the specific values were avoided, e.g. "used for livestock disease treatment"; "it has cultural importance"; etc. Finally, the use-reports recorded were categorized into major use categories (UC) of level 1 and sub-categories of level 2 using the Economic Botany Data Standard [ 35 ]. Aloe specimens were identified using taxonomic keys in the Flora of Ethiopia and Eritrea [ 17 ], through visual comparisons with authenticated plant specimens kept at the National Herbarium (ETH) of Addis Ababa University and at Herbarium of Haramaya University, and authenticated by Prof. Sebsebe Demissew (Professor of Plant Systematics and Biodiversity). Voucher specimens of all species with all herbarium sheet data were deposited at both herbaria. Voucher numbers of Haramaya University were used in this study. 2.3 Data analysis The data were organized and cleaned in an Excel spreadsheet (Microsoft office 2016), to be suitable for both qualitative and quantitative analysis. Qualitative data were analyzed following [ 32 , 34 , 36 ] like, free listing, part used, use-reports and use mentions. Quantitative data were analyzed using the corresponding formulas as follows: 2.3.1 Frequency of citation for use-report (FC UR ) Frequency of citation for use-report (FC UR ) is the percentage of informants who mentioned each use-report of particular Aloe species, which has been calculated using the formula: (1) FC UR ( % ) = n i / N × 100 where n i is the number of informants who cites each use-report per species and N is total number of informants. 2.3.2 Relative frequency of citation (RFC s ) Relative frequency of citation was calculated by dividing the frequency of citation (FC) of a species by total number of informants (N) involved in the five floristic regions that makes N to varies or the summation of use-report (∑UR i ) of all the informants interviewed for a species divided by N [ 37 ]. (2) RFC s = FC s N = ∑ i = i 1 i N UR i / N 2.3.3 Informant consensus factor (F ic ) The informant consensus factor (Fic) of each Aloe species is the proportion of informants who independently reported its use against a particular use category calculated using the formula [ 38 , 39 ]: (3) Fic = n ur − nt / n ur − 1 where, 'n ur ' is the "number of use-reports" in each use category and 'nt' is the "number of taxa used". 2.3.4 Use value (UV) Use value was used to demonstrate the relative importance of each Aloe species known locally, which can be calculated according to Albuquerque et al. formula [ 40 ]: (4) UV s = ∑ U i / N s where, UV s refers to the use value of Aloe species 's', U i to the number of different uses mentioned by each informant i per specific Aloe species, and N is the total number of informants interviewed for Aloe species 's'. 2.3.5 Relative importance index (RI) This index takes into account the number of major use-categories only and calculated as follows [ 37 ]: (5) RI s = RFCs ( max ) + RNUs ( max ) 2 where, RFC s (max) is the relative frequency of citation over the maximum. It is obtained by dividing FC s by the maximum value in all Aloe species of the study [RFC s(max) = FCs/max (FC)], and RNUs(max) is the relative number of use-categories over the maximum, obtained dividing the number of uses of the species NU s = u = u NC u = u 1 ∑ UR u by the maximum value in all Aloe species of the survey RNs(max) = NUs/max (NU). 2.3.6 Cultural value index (CV) This index estimates the cultural significance of each Aloe species, which combines the three variables, informant (i), a species (s), and use-category (u), which is calculated using the following formula [ 41 ]: (6) CV s = [ NU s NC ] [ FC s N ] [ ∑ U = U i uNC ∑ i = i 1 i N UR u 1 N ] The first factor is the relationship between the numbers of different uses reported for ethnospecies (each Aloe species ) and total number of use-categories. The second factor is the relative frequency of citation of a species. The third factor is the sum of all the UR for a species, i.e., the sum of number of participants who mentioned each use of a species, divided by N. To test if there was any correlation between age of the informant and their knowledge on use of aloes (number of use-reports), the nonparametric Kruskal Wallis Test was performed. If there was a significant difference between the informant's gender and knowledge about use of aloes, the non-parametric pair wise Wilcoxon test was performed using R software version 3.3.4. for Windows using multicompview and R companion packages. P-values of less than 0.05 were taken as statistically significant difference. 2.3.1 Frequency of citation for use-report (FC UR ) Frequency of citation for use-report (FC UR ) is the percentage of informants who mentioned each use-report of particular Aloe species, which has been calculated using the formula: (1) FC UR ( % ) = n i / N × 100 where n i is the number of informants who cites each use-report per species and N is total number of informants. 2.3.2 Relative frequency of citation (RFC s ) Relative frequency of citation was calculated by dividing the frequency of citation (FC) of a species by total number of informants (N) involved in the five floristic regions that makes N to varies or the summation of use-report (∑UR i ) of all the informants interviewed for a species divided by N [ 37 ]. (2) RFC s = FC s N = ∑ i = i 1 i N UR i / N 2.3.3 Informant consensus factor (F ic ) The informant consensus factor (Fic) of each Aloe species is the proportion of informants who independently reported its use against a particular use category calculated using the formula [ 38 , 39 ]: (3) Fic = n ur − nt / n ur − 1 where, 'n ur ' is the "number of use-reports" in each use category and 'nt' is the "number of taxa used". 2.3.4 Use value (UV) Use value was used to demonstrate the relative importance of each Aloe species known locally, which can be calculated according to Albuquerque et al. formula [ 40 ]: (4) UV s = ∑ U i / N s where, UV s refers to the use value of Aloe species 's', U i to the number of different uses mentioned by each informant i per specific Aloe species, and N is the total number of informants interviewed for Aloe species 's'. 2.3.5 Relative importance index (RI) This index takes into account the number of major use-categories only and calculated as follows [ 37 ]: (5) RI s = RFCs ( max ) + RNUs ( max ) 2 where, RFC s (max) is the relative frequency of citation over the maximum. It is obtained by dividing FC s by the maximum value in all Aloe species of the study [RFC s(max) = FCs/max (FC)], and RNUs(max) is the relative number of use-categories over the maximum, obtained dividing the number of uses of the species NU s = u = u NC u = u 1 ∑ UR u by the maximum value in all Aloe species of the survey RNs(max) = NUs/max (NU). 2.3.6 Cultural value index (CV) This index estimates the cultural significance of each Aloe species, which combines the three variables, informant (i), a species (s), and use-category (u), which is calculated using the following formula [ 41 ]: (6) CV s = [ NU s NC ] [ FC s N ] [ ∑ U = U i uNC ∑ i = i 1 i N UR u 1 N ] The first factor is the relationship between the numbers of different uses reported for ethnospecies (each Aloe species ) and total number of use-categories. The second factor is the relative frequency of citation of a species. The third factor is the sum of all the UR for a species, i.e., the sum of number of participants who mentioned each use of a species, divided by N. To test if there was any correlation between age of the informant and their knowledge on use of aloes (number of use-reports), the nonparametric Kruskal Wallis Test was performed. If there was a significant difference between the informant's gender and knowledge about use of aloes, the non-parametric pair wise Wilcoxon test was performed using R software version 3.3.4. for Windows using multicompview and R companion packages. P-values of less than 0.05 were taken as statistically significant difference. 3 Results 3.1 Ethno-medicinal and bio-cultural values of aloes with frequency of citation A total of 23 Aloe species (Appendix I) were recorded in the study areas, of which 21 are found in the Flora of Ethiopia and Eritrea [ 17 ] and two (recorded as unknown, voucher number AHU51 and AHU53) could not be identified to species level. Among the Aloe species reported in this study, 11 species (52%) are endemic and near endemic. The total number of use-reports was 196, were categorized into six major use categories (N UC = 6) with 2158 citations (use mentions) from the 23 Aloe species ( Table 3 ) by 210 respondents ( Table 1 ). The major use categories (UC) are medicines (Md), social uses (SU), materials (Mt), environmental uses (EU), vertebrate poisons (VP), and food (Fd). The medicinal use category accounted 149 use-reports (76%) with 1607 citations. The highest number of use-reports was recorded for Aloe megalacantha subsp. alticola with 23 use-reports, of which about 87 % were medicinal uses for humans and livestock. The next most frequently cited species was A. trichosantha subsp. longiflora with 20 use-reports followed with A. calidophila with 18 use-reports, in which 75% and 61.1% were medicinal (human and veterinary) uses, respectively ( Table 3 ). Table 3 List of Aloe species, major use categories, and frequency of citation per use reports (FC). E = Exudate; F = Flower; G = Gel; I = Inflorescence; L = Leaf; LP = Live plant; P = Pedicel; R = Root. Table 3 Scientific name & Voucher No. Major use category Sub-category Use report Part used Use description FC% Aloe calidophila Social uses Memorial Memorial LP Planting on graveyard 94.44 Reynolds Weaning Weaning child from breastfeeding E Apply to the nipple/breast 88.89 AHU103 Magic Belief to increase livestock herd size LP Planting at gate of the traditional cattle shelter 61.11 Cosmetics Skin softening G Apply on skin as a balm 22.22 Medicines Endocrine system Bile duct problem/Jaundice R Pulverize with honey & drink 83.33 P Chew & swallow the fluid 27.78 Infections and infestations Sexually transmitted infections/STI L Smoke-bathe the genitals 77.78 Malaria E Fresh exudate taken orally 44.44 R Pulverize with water & drink filtrate 44.44 Eye infection E Drop in infected eye 44.44 E Put on the head & apply externally around infected eye 27.78 Gonorrhea L Smoke-bathe the genitals 38.89 Repel flies from infected eye E Apply externally all around the eye 27.78 Skin and Wound healing E Apply externally 72.22 subcutaneous tissue Small swelling on skin locally called bocha L Warm fresh leaf & keep on small swellings 61.11 Wound of livestock due to carnivore/hyena attack L Fresh leaf crushed and tied on wound & smoke-bathed 55.56 Musculo-skeletal system Bone pain L Warm well & keep on painful part repeatedly 61.11 Bone pain of cattles L Warm well & keep on painful part repeatedly 61.11 Materials Domestic utensils Repel flies from wounds E Apply externally 38.89 Soap making Soap making G Used in small-scale soep production with ingredients 22.22 Food Metabolic system Water source G Fresh gel eaten as source of water in extremely hot areas 33.33 Aloe citrina Carter & Brandham Medicines Skin and subcutaneous tissue Wound healing E Apply externally 75.00 Leg and hand swelling G Tie onto swollen part 62.50 AHU123 Wound healing in livestock E Apply externally 75.00 Infections and infestations Eye infection E Drop in infected eye 75.00 Malaria E Taken orally 50.00 Digestive system Abdominal disorder E Taken orally 75.00 ∗ Aloe gilbertii Reynolds ex Sebsebe & Brandham subsp. gilbertii Medicines Endocrine system Bile duct problem/Jaundice E Taken orally 91.67 Digestive system Colon cleaner E Taken orally 75.00 Gastric L Fresh young leaf pulverized & filtrate take orally 58.33 Stomach disorder of cattles L Fresh young leaf pulverized & filtrate given orally 41.67 Skin and subcutaneous tissue Wound healing E Apply externally 75.00 AHU102 Infections and infestations Eye infection E Drop in infected eye 58.33 Malaria L & E Concocted & taken orally 25.00 Environment al uses Barrier Boundary marker LP Planting 75.00 Soil improver Soil conservation LP Planting on terracing 66.67 Social uses Memorial Memorial LP Planting on graveyard 58.33 ∗ , ¤ Aloe harlana Reynolds Medicines Skin and subcutaneous tissue Wound E Apply on wound 83.33 AHU117 Skin inflammation G & E Used as ointment 27.78 Snake and spider bites Snakebite L Pulverized with water & filtrate taken orally 72.22 Infections and infestations Hair fungus G Apply on hair or wash with fresh gel daily 61.11 Skin fungus G & E Used as ointment 50.00 Digestive system Bloated stomach of calltes locally called belelo L Fresh leaf pulverized & filtrate given orally 77.78 Colon cleaner E Powder (locally called SIBRI) taken orally in water 27.78 Endocrine system Liver swelling L Pulverized and filtrate taken orally 22.22 Spleen swelling/Splenomegaly L Pulverized and filtrate taken orally 22.22 Materials Domestic utensils Repel honeybees during honey harvest L Smoked while harvesting the honey to prevent bee stings 38.89 Aloe lateritia Engler Social uses Cosmetics Hair wash G Make shampoo for hair wash 66.67 AHU125 Soften hard skin G Scrape the gel & apply on skin 41.67 Medicines Skin and subcutaneous tissue Skin infection G & E Apply on skin 58.33 Infections and infestations Eye infection E Drop in the eye 58.33 Aloe macrocarpa Social uses Cosmetics Emollient G Scrape the gel & apply on skin 65.63 Todaro Medicines Infections and Skin diseases/fungal G & E Apply on skin 59.38 AHU19 infestations Hair fungus G Apply on hair or wash with fresh gel 50.00 Eye infection E Drop in eye 40.63 Infections and infestations Malaria E Collect & drink 12.50 Skin and subcutaneous tissue Fire burn G Apply immediately during an accident 37.50 Wound healing E Apply on wound as a cream 28.13 Wound healing of livestock E Apply on wound as a cream 31.25 Reproductive system and sex health Sexual impotency R Pulverized, mix with fresh butter and use as ointment & smoke-bathe the penis 9.38 ∗ Aloe mcloughlinii Medicines Skin and Wound healing E Powdered and applied on wound 62.50 Chris. subcutaneous tissue Wound healing/goats E Apply on infected part 50.00 AHU161 Infections and Eye infections E Drop in infected eye 50.00 infestations Antiparasite E Powdered solution taken orally 29.17 Snake and spider bites Snakebite S & E Concocted and filtrate taken orally 29.17 Digestive system Laxative E Collect & drink 20.83 ∗ Aloe megalacantha Baker subs. megalacantha Medicines Skin and subcutaneous tissue Skin infection G & E Concocted & tied onto the skin 90.91 AHU24 Infections and infestations Eye infection E Drop in infected ear 86.36 Digestive system Colon cleaner/locally called sibrii E Crystallized & juice made is taken orally in the morning 81.82 Stomach ulcer E Powder/ sibrii in water taken oral 54.55 Blood and Cardiovascular system Blood pressure E Powder/ sibrii of exudate in water solution taken oral 31.82 Musculo-skeletal system Leg and back pain L Cross-section leaf slices warmed & tied on foot while a bit hot 13.64 Environment Barrier Boundary marker LP Planting 81.82 al uses Boundaries Fence support LP Planting 68.18 Soil improver Soil conservation LP Planting on terracing 54.55 Aloe megalacantha Baker subs. Medicines Digestive system Colon cleaner/locally called sibrii E Collect fresh exudate, crystallized & Juice made taken orally 85.71 alticola Stomach disorder in cattles L Pulverized & filtrate taken orally 57.14 AHU162 Infections and Eye infection E Drop in infected ear 67.86 infestations E Apply external around the eye 28.57 Ear infection E Drop in infected eye 32.14 Tonsillitis E Drop on throat 25.00 Itching eye E Apply externally around the eye 21.43 Foot and mouth disease S Smoke-bathe infected part 21.43 Endocrine system Bile duct problem locally called hadhoftuu E Taken oral 60.71 Diabetics E Powder in water solution taken orally 53.57 Skin and subcutaneous tissue Skin infection S Smoke-bathe infected part with wet stem 57.14 Goats skin wound L Crush & rub on affected part 50.00 Camel skin wound L Crush & rub on affected part 42.86 Skin infection G & E Concocted & tied onto infected part 39.29 Itching skin of goats L Crush & rub on affected part 32.14 General Ailments with unspecific Symptoms Cold problem L Pulverized & massage the body 42.86 Cold problem locally called qorraa S Smoke-bathe body with fresh leaf until well sweating locally called qayyaa 28.57 Body pain feeling L Warm well & keep on the painful part repeatedly 28.57 Weak body feeling E Drops in water & drink 21.43 Musculo-skeletal system Knee pain due to cold G Softly massage the knee 42.86 E Apply on the knee 28.57 Blood and Cardiovascular system Clean the blood E Drops in water, mix & drink 21.43 Materials Domestic utensils Honey harvesting S Smoke near the bee hive while harvesting 75.00 Social uses Cosmetics Smoke bath for women S Smoke-bathe the genital part with dried and wet stem 64.29 Tattoo Colouring hand and leg/women S Smoke bath of hand and leg with wet stem 28.57 ∗ Aloe otallensis Baker Social uses Weaning Weaning child from breastfeeding E Apply to the nipple/breast 100.00 AHU107 Medicines Endocrine system Bile duct problem called hadhoftu L Pulverized & filtrate given oral 75.00 Skin and subcutaneous tissue Wound due to carnivore/hyena attack L Burned and blackish ash mixed with butter used as ointment 75.00 Wound healing E Apply external 66.67 Hand/leg swelling L Warm & put on swollen part 41.67 Digestive system Colon cleaner E Powder in water solution taken orally 58.33 Chicken disease/diarrhea E Given orally 75.00 ∗ Aloe pirottae Medicines Skin and Wound healing E Apply externally 87.50 Berger subcutaneous tissue Tropical ulcer G & E Concocted & used external as ointment 70.83 AHU13 Infections and Eye infections E Drop in infected eye 87.50 infestations Malaria E Powder of exudate in water solution taken orally 54.17 Gonorrhea E Mixed with honey and taken oral 25.00 Antiparasite E Taken orally 20.83 Endocrine system Bile duct problem E Taken orally 58.33 Gallstone G & E Taken orally 20.83 Snake and spider bites Snakebite E Taken orally 37.50 Digestive system Colon cleaner/SIBRI E Powder of exudate in water solution taken orally 33.33 Musculo-skeletal system Muscular pain G Boiled & soft massage painful part 16.67 Social uses Weaning Weaning child from breastfeeding E Apply to the nipple/breast 75.00 Materials Domestic utensils Insect repellent L Smoking around the area to stifle insects 62.50 Mosquito repellent L Smoke of dried leaves to stifle mosquitoes 41.67 ∗ , ◊ Aloe pubescens Reynolds Medicines Digestive system Colon cleaner E Powder of exudate locally called SIBRI in water solution taken orally 86.36 AHU06 Soften alimentary canal L Young leaf eaten 59.09 Gastric G Fresh gel eaten 54.55 Stomachache/kurtet R & F Concocted & filterate taken orally 18.18 Endocrine system Bile duct problem locally called Hadhoftu E Taken orally 59.09 Skin and subcutaneous tissue Wound E Apllied on dermal wound 54.55 Endocrine system Liver disease E Taken orally 31.82 Infections and infestations Anthrax L, F & R Concocted & given orally 13.64 Sicial uses Memorial Graveyard LP Planting on graveyards 72.73 Environment Boundaries Fence support LP Planting 68.18 al uses Soil improver Soil conservation LP Planting on mountain slopes along the terrace 36.36 Food Food Food G Fresh gel scraped and eaten 31.82 ∗∗ Aloe retrospiciens Reynolds & Bally Vertebrate poisons Poison Poison carnivore E Concentrated exudate hide in meat to feed hyena 58.33 AHU160 Poison Kill goats if eaten in dry season, due to starvation L If eaten in the drought season can kill goats 41.67 Poison Poison rats E Dried exudate and apply in rat feed 25.00 Aloe rivae Baker Social uses Cosmetics Body & hair wash G Make shampoo for washing 91.67 AHU115 Medicines Snake and spider bites Snakebite E Taken orally 75.00 Skin and subcutaneous tissue Wound healing E Apply externally 66.67 Cancer Breast cancer E Taken orally 25.00 Materials Domestic utensils Repel flies E Apply externally on wound 41.67 ∗∗ Aloe rugosifolia Gilbert & Sebsebe Medicines Musculo-skeletal system Bone pain L Warmth well & keep on painful part repeatedly 77.78 AHU113 Skin and subcutaneous tissue Wound healing E Apply externally 72.22 Wound of livestock E Apply externally 61.11 Small swelling on skin locally called bocha , on hand and legs L Warm & keep on small swellings 38.89 Infections and infestations Eye infection E Drop in infected eye 44.44 Aloe ruspoliana Baker Vertebrate poisons Carnivor prevention Deter carnivore from night shelter of livestock LP Planting around livestock night shelter 87.50 AHU121 Rodent control Deter rats due to bad smell LP Planting around rats nests 50.00 Medicines Skin and subcutaneous tissue Itching skin on goat locally called chito L Warm & rub the skin while warmer 37.50 Aloe secundiflora Medicines Skin and subcutaneous tissue Skin infections E External use on skin 88.89 Engler Wound healing E Drop on wound and bandage 83.33 AHU106 Skin infection of livestock L & E Prepared for external use on skin 77.78 Wound on livestock skin L & E Prepared & applied to wound 61.11 Musculo-skeletal system Inflammation in muscles/Rheumatism G & E Mix & tie on inflamed part 38.89 Infections and infestations Ectoparasite L & E Concocted for external use on skin 50.00 Malaria E Taken orally 33.33 Diarrhea E Taken orally 22.22 ∗ , ◊ Aloe tewoldei Gilbert & Sebsebe Medicines Musculo-skeletal system Bone fracture G Bandage the gel around to soften the part before traditional fracture medication 87.50 AHU120 General Ailments with Unspecific Symptoms Cold problem L Pulverized & filterate taken orally 87.50 Digestive system Stomach disorder L Pulverized & filterate taken orally 87.50 Skin and subcutaneous tissue Itching skin on goat locally called chito L Put in fire until very soft then rub on skin 62.50 Social uses Poison Crop pest control E Powder used in crop storage 50.00 ∗ Aloe trichosantha subsp. longiflora Gilbert & Sebsebe Medicines Digestive system Laxative/Purgative, colon cleaner, Constipation E Powdered & drink in the form of solution or swallow with banana 50.98 Skin and subcutaneous tissue Infection on skin or wound/Antidermatosic E Apply on infected skin 33.33 Fire burn G Applied externally 9.80 AHU05 Skin hardening/emollient G & E Make juicy & apply on skin 4.90 Snake and spider bites Snakebites antidote L & E Concocted with water & taken oral 23.53 Endocrine system Bile duct problem L Pulverized kept for 12 h and drink filtrate/young leaf 22.55 Infections and Tonsillitis E Drops on the throat 11.76 infestations Malaria E Taken orally 10.78 L Pulverize & filtrate taken orally 1.96 Eye infection E Drop in infected eye 6.86 Skin and subcutaneous tissue After male circumcision L Smoke bathed after circumisision to prevent potential infection 14.71 Wound on livestock skin E Apply externally on wound 14.71 General Ailments with Unspecific Symptoms Pain due to cold G Massage the pain part softly 10.78 Sensory system Improve poor sight E A drop in eyes 0.98 Reproductive system and sex health Infertility of man and woman G Wash the body and genitalia 0.98 Pregnancy, birth and puerperial Delayed placenta in cattles L & R Concoction given orally 5.88 Social uses Weaning Weaning child from breastfeeding E Apply to the nipple/breast 50.00 Magic Increase herd size of livestock and camel L Smoke-bathe milking utensils 1.96 Illuminant Lighting bonfire I Lighting bonfire/torch with dried sticks used in Christian holidays 6.86 Materials Domestic utensils Repel flies from wounds E Apply on and around the wound 18.63 Food Metabolic system Relief dehydration in extreme hot condition G Make it free from exudate & eaten 5.88 ∗ Aloe welmelensis Sebsebe & Nordal Medicines Skin and subcutaneous tissue Wound healing E Apply externally 75.00 AHU124 Wound healing for cattles E Apply externally 75.00 ∗ , ¤ Aloe yavellana Social uses Weaning Weaning child from breastfeeding E Apply to the niple/breast 94.44 Reynolds Environment al uses Soil improver Soil conservation LP Planting on terracing 88.89 AHU116 Boundaries Fence support LP Planting 61.11 Medicines Skin and subcutaneous tissue Wound healing E Apply externally 72.22 Skin and subcutaneous tissue Wound healing of livestock E Apply externally 72.22 Ectoparasite of livestock L & E Concocted for external use on skin 50.00 Snake and spider bites Snake poison E Taken orally 44.44 Materials Domestic utensils Mosquito repellent L Smoking around to stifle mosquito 66.67 ∗ Highly spotted aloe Medicines Snake and spider Snake poison E Drink very soon after snakebite 63.33 AHU53 bites Spider poison E Drink very soon after spiderbite 56.67 Skin and subcutaneous tissue Wound healing E Powder apply on wound 56.67 Digestive system Diarrhea in cattles L Pulverized & filtrate given orally 46.67 Materials Pest control Crop pest E Powdered & applied in traditional crop storage 36.67 Unidentified Medicines Infections and Hair fungus G Apply on hair 83.33 AHU51 infestations Tonsillitis E Drop in the throat 61.11 Endocrine system Diabetics E Powder in solution & drink daily morning 72.22 Social uses Cosmetics Skin softening G Apply on skin 66.67 ∗ Endemic. ∗ Narrowly endemic. ¤ Endangered. ◊ Near threatened. ∗∗ Vulnerable. The highest frequency of citation was 100% recorded for Aloe otallensis leaf exudate used for weaning children from breast-feeding followed with 94.4% for A. yavellana leaf exudate used to treat jaundice, 91.6% for A. rivae leaf gel used for body and hair wash, and 90.9% for A. megalacantha subsp. alticola exudate to treat skin infections ( Table 3 ). The most frequently cited use of Aloe species in the study areas was for medicines (human and veterinary): 149 use-reports (76%) from 22 Aloe species, with a total number of 1607 citations (use mentions) were recorded. The least use-report was for food use-category: three use-reports (1.5%) from three Aloe species with 19 citations ( Figure 2 ). Figure 2 The number of use-reports and number of Aloe species used in each major use category. Figure 2 A total of nine different plant parts were used in the diverse bio-cultural uses ( Figure 3 ) within the six major use categories. Leaf exudate was the most frequently sought part, accounting for 111 use-reports (49.1%) in which about 89 (80.2%) were for human medicinal formulations and 12 (11.7%) were for livestock medicinal formulations. That means, 92.9% of exudate were used for medicinal purposes and the remainder in formulations were in social uses and vertebrate poisons categories. In addition, the entire leaf for 47 use-reports (20.8%), which also include the use of exudate as part of entire leaf. This would suggest that the use of exudate exceeds 69% of the plant parts used for medicinal purpose. The fewest use-reports were reported from formulations made from the inflorescence and pedicel, with just a single use-report, which is 0.4% each ( Figure 3 ). Figure 3 Number of uses and percentage of each part used for the respective treatment purposes. Figure 3 The local communities practiced nine types of preparation methods for medicinal applications, out of these 42% are prepared in the form of pure exudate collection from the fresh leaf to be used for different application followed by pulverization (11.5%) and concoction (10.8%) ( Figure 4 ). Figure 4 Number of preparation in each type of preparation method for medicinal uses. Figure 4 3.2 Use-reports among cultural communities The distribution of use-reports among the cultural communities in the study areas showed 161 use-reports for the Oromo community (N = 140), followed with 17 for the Somali community (N = 32), 9 for the Afar community (N = 24), and 9 use-reports for the Harari community (N = 14). In most of the study areas, the medicinal uses of Aloe species were found more popular among the Oromo community, which accounted for 125 medicinal use-reports, in which 77.6% of the total use-report of this community. The fewest uses were reported from the Afar and Harari communities with 9 use-reports each. Among the bio-cultural values, environmental use-reports like boundary marking, soil conservation, and living fence support were documented only from the Oromo community, from four Aloe species. A unique use-report "poisonous to carnivores" was reported by the Somali and Oromo communities for two Aloe species called Aloe retrospiciens Reynolds & Bally and A. ruspoliana Baker, respectively. The Oromo community used 22 out of the 23 species documented in this study (the exception was A. retrospiciens ). The Somali community used four species ( A. retrospiciens, A . megalacantha subs. megalacantha, A. mcloughlinii , and A. pirottaei ) followed by the Afar community who use two species, A. retrospiciens and A . trichosantha subsp. longiflora and the Harari community who also use two species, A. trichosantha subsp. longiflora and A. macrocarpa. 3.3 Ethno-medicinal and bio-cultural knowledge among gender and age categories The level of knowledge of Aloe species diversity and use-reports are noticeably lower in the younger age categories compared to elderly informants. For example, out of the total citations (1336) for human medicinal uses, 654 were by elderly people (above 60 years, N = 35) followed with 398 use mention by adults (age 41–60, N = 108) and 284 by young informants (age 25–40, N = 67). The highest number of use-report in all major use categories such as medicines (Md), social uses (SU), materials (Mt), environmental uses (EU), vertebrate poisons (VP), and food (Fd) were reported by elderly people ( Figure 5 ). In the medicines category, 21 species were reported by elderly people (above 60 years, N = 35) followed by 16 species by adults (age 41–60, N = 108) and eight species by young informants (age 25–40, N = 67) ( Figure 6 ). Figure 5 Number of use-reports in each major use category for each age group (medicines (Md), social uses (SU), materials (Mt), environmental uses (EU), vertebrate poisons (VP), and food (Fd)). Figure 5 Figure 6 The number of Aloe species reported in each major use category for each age group (Md), social uses (SU), materials (Mt), environmental uses (EU), vertebrate poisons (VP), and food (Fd)). Figure 6 The depth of comprehensive bio-cultural knowledge among different age groups indicated that elderly people (above 60 years) had a much deeper knowledge than the two age group ranging from 25 to 40 and 41–60 (Kruskal-Wallis chi-squared = 12.17, df = 3, p = 0.006∗). There is a significant difference in the depth of ethno-medicinal knowledge between the 25–40 age group and the above 60 age group (p = 0.0004∗∗), and a significant difference between 41-60 year olds and the above 60 age group (p = 0.02∗) (p 0.05). 3.4 Informant consensus factor and value indexes Informant consensus factors (F ic ) of the seven major use categories ranged from 0.86 to 0.22. The highest F ic value 0.86 was for medicines with 22 species and 11 use sub-categories, followed by 0.63 for environmental uses with four species and three sub-categories, and least F ic value was 0.22 for materials with eight species and two sub-categories ( Table 4 ). Table 4 Major use categories with the corresponding sub-categories, number of Aloe species used and informant consensus factor values. Table 4 Major use categories with sub-categories No. of sub-category N. of use reports No. of citation No. of species F ic Medicines : Blood and cardiovascular system, cancer, digestive system, infections and infestations, general ailments with unspecific symptoms, endocrine system, snake and spider bites, pregnancy, birth and puerperial, reproductive system and sex health, sensory system, skin and subcutaneous tissue, & musculo-skeletal system 12 146 (76%) 1607 22 0.86 Social uses : Weaning, illuminant, magic, memorial, tattoo & cosmetics 7 20 (10.2%) 276 13 0.37 Materials: Domestic utensils & soap making 2 10 (5.1%) 107 8 0.22 Environmental uses : Soil improver, boundaries, & barrier 3 9 (4.6%) 112 4 0.63 Vertebrate Poisons : Poison, carnivore prevention, rodent control, & pest control 6 5 (2.6%) 37 3 0.50 Food: Metabolic system & edible 2 3 (1.5%) 19 2 0.50 Aloe species were compared based on the values of importance metrics. Some species like Aloe calidophila, A. megalacantha subs. alticola, A. pirottae, and A. gilbertii subsp. gilbertii showed higher values across the indices due to be mentioned by a higher number of informants, relatively higher number of use-categories, and larger number of use-reports to elucidate species that were of high importance in all indices ( Table 5 ). Table 5 Relative Importance metrics of Aloe species used in the Great Rift Valley floristic regions of Ethiopia. Table 5 Species UC UR UV RFC RI CV Aloe calidophila 4 18 11.72 0.68 0.89 6.20 Aloe megalacantha subsp. alticola 3 23 10.89 0.52 0.88 5.66 Aloe pirottae 3 14 6.92 0.49 0.66 2.06 Aloe gilbertii subsp. gilbertii 3 10 6.00 0.60 0.66 1.55 Aloe pubescens 4 12 5.86 0.49 0.62 1.49 Aloe megalacantha subsp. megalacantha 2 9 5.63 0.63 0.66 1.39 Aloe yavellana 4 8 5.00 0.63 0.63 1.10 Aloe harlana 2 10 4.83 0.48 0.57 1.01 Aloe secundiflora 1 8 4.55 0.57 0.59 0.90 Aloe trichosantha subsp. longiflora 4 20 3.14 0.24 0.61 0.64 Aloe otallensis 2 7 4.92 0.61 0.60 0.91 Aloe macrocarpa 2 9 3.44 0.55 0.60 0.74 Aloe rivae 3 5 3.91 0.58 0.53 0.49 Aloe citrina 1 6 4.12 0.51 0.51 0.55 Aloe rugosifolia 1 5 2.94 0.39 0.39 0.25 Unidentified (AHU51) 2 4 2.83 0.61 0.54 0.30 Unidentified (AHU53) 2 5 2.60 0.52 0.49 0.28 Aloe tewoldei 2 5 2.50 0.33 0.34 0.18 Aloe mcloughlinii 1 6 2.42 0.40 0.43 0.25 Aloe lateritia 2 4 2.25 0.56 0.50 0.22 Aloe welmelensis 1 2 1.50 0.38 0.32 0.12 Aloe retrospiciens 1 3 1.25 0.42 0.37 0.07 Aloe ruspoliana 2 3 1.17 0.28 0.27 0.05 UC = number of major use categories per species; UR = number of use-reports per species; UV = use value of a species; RFC = relative frequency of citation for species; RI = relative importance index; CV = cultural value index of species, which considered the three factors i.e. s (species), i (informants), and u (uses). 3.1 Ethno-medicinal and bio-cultural values of aloes with frequency of citation A total of 23 Aloe species (Appendix I) were recorded in the study areas, of which 21 are found in the Flora of Ethiopia and Eritrea [ 17 ] and two (recorded as unknown, voucher number AHU51 and AHU53) could not be identified to species level. Among the Aloe species reported in this study, 11 species (52%) are endemic and near endemic. The total number of use-reports was 196, were categorized into six major use categories (N UC = 6) with 2158 citations (use mentions) from the 23 Aloe species ( Table 3 ) by 210 respondents ( Table 1 ). The major use categories (UC) are medicines (Md), social uses (SU), materials (Mt), environmental uses (EU), vertebrate poisons (VP), and food (Fd). The medicinal use category accounted 149 use-reports (76%) with 1607 citations. The highest number of use-reports was recorded for Aloe megalacantha subsp. alticola with 23 use-reports, of which about 87 % were medicinal uses for humans and livestock. The next most frequently cited species was A. trichosantha subsp. longiflora with 20 use-reports followed with A. calidophila with 18 use-reports, in which 75% and 61.1% were medicinal (human and veterinary) uses, respectively ( Table 3 ). Table 3 List of Aloe species, major use categories, and frequency of citation per use reports (FC). E = Exudate; F = Flower; G = Gel; I = Inflorescence; L = Leaf; LP = Live plant; P = Pedicel; R = Root. Table 3 Scientific name & Voucher No. Major use category Sub-category Use report Part used Use description FC% Aloe calidophila Social uses Memorial Memorial LP Planting on graveyard 94.44 Reynolds Weaning Weaning child from breastfeeding E Apply to the nipple/breast 88.89 AHU103 Magic Belief to increase livestock herd size LP Planting at gate of the traditional cattle shelter 61.11 Cosmetics Skin softening G Apply on skin as a balm 22.22 Medicines Endocrine system Bile duct problem/Jaundice R Pulverize with honey & drink 83.33 P Chew & swallow the fluid 27.78 Infections and infestations Sexually transmitted infections/STI L Smoke-bathe the genitals 77.78 Malaria E Fresh exudate taken orally 44.44 R Pulverize with water & drink filtrate 44.44 Eye infection E Drop in infected eye 44.44 E Put on the head & apply externally around infected eye 27.78 Gonorrhea L Smoke-bathe the genitals 38.89 Repel flies from infected eye E Apply externally all around the eye 27.78 Skin and Wound healing E Apply externally 72.22 subcutaneous tissue Small swelling on skin locally called bocha L Warm fresh leaf & keep on small swellings 61.11 Wound of livestock due to carnivore/hyena attack L Fresh leaf crushed and tied on wound & smoke-bathed 55.56 Musculo-skeletal system Bone pain L Warm well & keep on painful part repeatedly 61.11 Bone pain of cattles L Warm well & keep on painful part repeatedly 61.11 Materials Domestic utensils Repel flies from wounds E Apply externally 38.89 Soap making Soap making G Used in small-scale soep production with ingredients 22.22 Food Metabolic system Water source G Fresh gel eaten as source of water in extremely hot areas 33.33 Aloe citrina Carter & Brandham Medicines Skin and subcutaneous tissue Wound healing E Apply externally 75.00 Leg and hand swelling G Tie onto swollen part 62.50 AHU123 Wound healing in livestock E Apply externally 75.00 Infections and infestations Eye infection E Drop in infected eye 75.00 Malaria E Taken orally 50.00 Digestive system Abdominal disorder E Taken orally 75.00 ∗ Aloe gilbertii Reynolds ex Sebsebe & Brandham subsp. gilbertii Medicines Endocrine system Bile duct problem/Jaundice E Taken orally 91.67 Digestive system Colon cleaner E Taken orally 75.00 Gastric L Fresh young leaf pulverized & filtrate take orally 58.33 Stomach disorder of cattles L Fresh young leaf pulverized & filtrate given orally 41.67 Skin and subcutaneous tissue Wound healing E Apply externally 75.00 AHU102 Infections and infestations Eye infection E Drop in infected eye 58.33 Malaria L & E Concocted & taken orally 25.00 Environment al uses Barrier Boundary marker LP Planting 75.00 Soil improver Soil conservation LP Planting on terracing 66.67 Social uses Memorial Memorial LP Planting on graveyard 58.33 ∗ , ¤ Aloe harlana Reynolds Medicines Skin and subcutaneous tissue Wound E Apply on wound 83.33 AHU117 Skin inflammation G & E Used as ointment 27.78 Snake and spider bites Snakebite L Pulverized with water & filtrate taken orally 72.22 Infections and infestations Hair fungus G Apply on hair or wash with fresh gel daily 61.11 Skin fungus G & E Used as ointment 50.00 Digestive system Bloated stomach of calltes locally called belelo L Fresh leaf pulverized & filtrate given orally 77.78 Colon cleaner E Powder (locally called SIBRI) taken orally in water 27.78 Endocrine system Liver swelling L Pulverized and filtrate taken orally 22.22 Spleen swelling/Splenomegaly L Pulverized and filtrate taken orally 22.22 Materials Domestic utensils Repel honeybees during honey harvest L Smoked while harvesting the honey to prevent bee stings 38.89 Aloe lateritia Engler Social uses Cosmetics Hair wash G Make shampoo for hair wash 66.67 AHU125 Soften hard skin G Scrape the gel & apply on skin 41.67 Medicines Skin and subcutaneous tissue Skin infection G & E Apply on skin 58.33 Infections and infestations Eye infection E Drop in the eye 58.33 Aloe macrocarpa Social uses Cosmetics Emollient G Scrape the gel & apply on skin 65.63 Todaro Medicines Infections and Skin diseases/fungal G & E Apply on skin 59.38 AHU19 infestations Hair fungus G Apply on hair or wash with fresh gel 50.00 Eye infection E Drop in eye 40.63 Infections and infestations Malaria E Collect & drink 12.50 Skin and subcutaneous tissue Fire burn G Apply immediately during an accident 37.50 Wound healing E Apply on wound as a cream 28.13 Wound healing of livestock E Apply on wound as a cream 31.25 Reproductive system and sex health Sexual impotency R Pulverized, mix with fresh butter and use as ointment & smoke-bathe the penis 9.38 ∗ Aloe mcloughlinii Medicines Skin and Wound healing E Powdered and applied on wound 62.50 Chris. subcutaneous tissue Wound healing/goats E Apply on infected part 50.00 AHU161 Infections and Eye infections E Drop in infected eye 50.00 infestations Antiparasite E Powdered solution taken orally 29.17 Snake and spider bites Snakebite S & E Concocted and filtrate taken orally 29.17 Digestive system Laxative E Collect & drink 20.83 ∗ Aloe megalacantha Baker subs. megalacantha Medicines Skin and subcutaneous tissue Skin infection G & E Concocted & tied onto the skin 90.91 AHU24 Infections and infestations Eye infection E Drop in infected ear 86.36 Digestive system Colon cleaner/locally called sibrii E Crystallized & juice made is taken orally in the morning 81.82 Stomach ulcer E Powder/ sibrii in water taken oral 54.55 Blood and Cardiovascular system Blood pressure E Powder/ sibrii of exudate in water solution taken oral 31.82 Musculo-skeletal system Leg and back pain L Cross-section leaf slices warmed & tied on foot while a bit hot 13.64 Environment Barrier Boundary marker LP Planting 81.82 al uses Boundaries Fence support LP Planting 68.18 Soil improver Soil conservation LP Planting on terracing 54.55 Aloe megalacantha Baker subs. Medicines Digestive system Colon cleaner/locally called sibrii E Collect fresh exudate, crystallized & Juice made taken orally 85.71 alticola Stomach disorder in cattles L Pulverized & filtrate taken orally 57.14 AHU162 Infections and Eye infection E Drop in infected ear 67.86 infestations E Apply external around the eye 28.57 Ear infection E Drop in infected eye 32.14 Tonsillitis E Drop on throat 25.00 Itching eye E Apply externally around the eye 21.43 Foot and mouth disease S Smoke-bathe infected part 21.43 Endocrine system Bile duct problem locally called hadhoftuu E Taken oral 60.71 Diabetics E Powder in water solution taken orally 53.57 Skin and subcutaneous tissue Skin infection S Smoke-bathe infected part with wet stem 57.14 Goats skin wound L Crush & rub on affected part 50.00 Camel skin wound L Crush & rub on affected part 42.86 Skin infection G & E Concocted & tied onto infected part 39.29 Itching skin of goats L Crush & rub on affected part 32.14 General Ailments with unspecific Symptoms Cold problem L Pulverized & massage the body 42.86 Cold problem locally called qorraa S Smoke-bathe body with fresh leaf until well sweating locally called qayyaa 28.57 Body pain feeling L Warm well & keep on the painful part repeatedly 28.57 Weak body feeling E Drops in water & drink 21.43 Musculo-skeletal system Knee pain due to cold G Softly massage the knee 42.86 E Apply on the knee 28.57 Blood and Cardiovascular system Clean the blood E Drops in water, mix & drink 21.43 Materials Domestic utensils Honey harvesting S Smoke near the bee hive while harvesting 75.00 Social uses Cosmetics Smoke bath for women S Smoke-bathe the genital part with dried and wet stem 64.29 Tattoo Colouring hand and leg/women S Smoke bath of hand and leg with wet stem 28.57 ∗ Aloe otallensis Baker Social uses Weaning Weaning child from breastfeeding E Apply to the nipple/breast 100.00 AHU107 Medicines Endocrine system Bile duct problem called hadhoftu L Pulverized & filtrate given oral 75.00 Skin and subcutaneous tissue Wound due to carnivore/hyena attack L Burned and blackish ash mixed with butter used as ointment 75.00 Wound healing E Apply external 66.67 Hand/leg swelling L Warm & put on swollen part 41.67 Digestive system Colon cleaner E Powder in water solution taken orally 58.33 Chicken disease/diarrhea E Given orally 75.00 ∗ Aloe pirottae Medicines Skin and Wound healing E Apply externally 87.50 Berger subcutaneous tissue Tropical ulcer G & E Concocted & used external as ointment 70.83 AHU13 Infections and Eye infections E Drop in infected eye 87.50 infestations Malaria E Powder of exudate in water solution taken orally 54.17 Gonorrhea E Mixed with honey and taken oral 25.00 Antiparasite E Taken orally 20.83 Endocrine system Bile duct problem E Taken orally 58.33 Gallstone G & E Taken orally 20.83 Snake and spider bites Snakebite E Taken orally 37.50 Digestive system Colon cleaner/SIBRI E Powder of exudate in water solution taken orally 33.33 Musculo-skeletal system Muscular pain G Boiled & soft massage painful part 16.67 Social uses Weaning Weaning child from breastfeeding E Apply to the nipple/breast 75.00 Materials Domestic utensils Insect repellent L Smoking around the area to stifle insects 62.50 Mosquito repellent L Smoke of dried leaves to stifle mosquitoes 41.67 ∗ , ◊ Aloe pubescens Reynolds Medicines Digestive system Colon cleaner E Powder of exudate locally called SIBRI in water solution taken orally 86.36 AHU06 Soften alimentary canal L Young leaf eaten 59.09 Gastric G Fresh gel eaten 54.55 Stomachache/kurtet R & F Concocted & filterate taken orally 18.18 Endocrine system Bile duct problem locally called Hadhoftu E Taken orally 59.09 Skin and subcutaneous tissue Wound E Apllied on dermal wound 54.55 Endocrine system Liver disease E Taken orally 31.82 Infections and infestations Anthrax L, F & R Concocted & given orally 13.64 Sicial uses Memorial Graveyard LP Planting on graveyards 72.73 Environment Boundaries Fence support LP Planting 68.18 al uses Soil improver Soil conservation LP Planting on mountain slopes along the terrace 36.36 Food Food Food G Fresh gel scraped and eaten 31.82 ∗∗ Aloe retrospiciens Reynolds & Bally Vertebrate poisons Poison Poison carnivore E Concentrated exudate hide in meat to feed hyena 58.33 AHU160 Poison Kill goats if eaten in dry season, due to starvation L If eaten in the drought season can kill goats 41.67 Poison Poison rats E Dried exudate and apply in rat feed 25.00 Aloe rivae Baker Social uses Cosmetics Body & hair wash G Make shampoo for washing 91.67 AHU115 Medicines Snake and spider bites Snakebite E Taken orally 75.00 Skin and subcutaneous tissue Wound healing E Apply externally 66.67 Cancer Breast cancer E Taken orally 25.00 Materials Domestic utensils Repel flies E Apply externally on wound 41.67 ∗∗ Aloe rugosifolia Gilbert & Sebsebe Medicines Musculo-skeletal system Bone pain L Warmth well & keep on painful part repeatedly 77.78 AHU113 Skin and subcutaneous tissue Wound healing E Apply externally 72.22 Wound of livestock E Apply externally 61.11 Small swelling on skin locally called bocha , on hand and legs L Warm & keep on small swellings 38.89 Infections and infestations Eye infection E Drop in infected eye 44.44 Aloe ruspoliana Baker Vertebrate poisons Carnivor prevention Deter carnivore from night shelter of livestock LP Planting around livestock night shelter 87.50 AHU121 Rodent control Deter rats due to bad smell LP Planting around rats nests 50.00 Medicines Skin and subcutaneous tissue Itching skin on goat locally called chito L Warm & rub the skin while warmer 37.50 Aloe secundiflora Medicines Skin and subcutaneous tissue Skin infections E External use on skin 88.89 Engler Wound healing E Drop on wound and bandage 83.33 AHU106 Skin infection of livestock L & E Prepared for external use on skin 77.78 Wound on livestock skin L & E Prepared & applied to wound 61.11 Musculo-skeletal system Inflammation in muscles/Rheumatism G & E Mix & tie on inflamed part 38.89 Infections and infestations Ectoparasite L & E Concocted for external use on skin 50.00 Malaria E Taken orally 33.33 Diarrhea E Taken orally 22.22 ∗ , ◊ Aloe tewoldei Gilbert & Sebsebe Medicines Musculo-skeletal system Bone fracture G Bandage the gel around to soften the part before traditional fracture medication 87.50 AHU120 General Ailments with Unspecific Symptoms Cold problem L Pulverized & filterate taken orally 87.50 Digestive system Stomach disorder L Pulverized & filterate taken orally 87.50 Skin and subcutaneous tissue Itching skin on goat locally called chito L Put in fire until very soft then rub on skin 62.50 Social uses Poison Crop pest control E Powder used in crop storage 50.00 ∗ Aloe trichosantha subsp. longiflora Gilbert & Sebsebe Medicines Digestive system Laxative/Purgative, colon cleaner, Constipation E Powdered & drink in the form of solution or swallow with banana 50.98 Skin and subcutaneous tissue Infection on skin or wound/Antidermatosic E Apply on infected skin 33.33 Fire burn G Applied externally 9.80 AHU05 Skin hardening/emollient G & E Make juicy & apply on skin 4.90 Snake and spider bites Snakebites antidote L & E Concocted with water & taken oral 23.53 Endocrine system Bile duct problem L Pulverized kept for 12 h and drink filtrate/young leaf 22.55 Infections and Tonsillitis E Drops on the throat 11.76 infestations Malaria E Taken orally 10.78 L Pulverize & filtrate taken orally 1.96 Eye infection E Drop in infected eye 6.86 Skin and subcutaneous tissue After male circumcision L Smoke bathed after circumisision to prevent potential infection 14.71 Wound on livestock skin E Apply externally on wound 14.71 General Ailments with Unspecific Symptoms Pain due to cold G Massage the pain part softly 10.78 Sensory system Improve poor sight E A drop in eyes 0.98 Reproductive system and sex health Infertility of man and woman G Wash the body and genitalia 0.98 Pregnancy, birth and puerperial Delayed placenta in cattles L & R Concoction given orally 5.88 Social uses Weaning Weaning child from breastfeeding E Apply to the nipple/breast 50.00 Magic Increase herd size of livestock and camel L Smoke-bathe milking utensils 1.96 Illuminant Lighting bonfire I Lighting bonfire/torch with dried sticks used in Christian holidays 6.86 Materials Domestic utensils Repel flies from wounds E Apply on and around the wound 18.63 Food Metabolic system Relief dehydration in extreme hot condition G Make it free from exudate & eaten 5.88 ∗ Aloe welmelensis Sebsebe & Nordal Medicines Skin and subcutaneous tissue Wound healing E Apply externally 75.00 AHU124 Wound healing for cattles E Apply externally 75.00 ∗ , ¤ Aloe yavellana Social uses Weaning Weaning child from breastfeeding E Apply to the niple/breast 94.44 Reynolds Environment al uses Soil improver Soil conservation LP Planting on terracing 88.89 AHU116 Boundaries Fence support LP Planting 61.11 Medicines Skin and subcutaneous tissue Wound healing E Apply externally 72.22 Skin and subcutaneous tissue Wound healing of livestock E Apply externally 72.22 Ectoparasite of livestock L & E Concocted for external use on skin 50.00 Snake and spider bites Snake poison E Taken orally 44.44 Materials Domestic utensils Mosquito repellent L Smoking around to stifle mosquito 66.67 ∗ Highly spotted aloe Medicines Snake and spider Snake poison E Drink very soon after snakebite 63.33 AHU53 bites Spider poison E Drink very soon after spiderbite 56.67 Skin and subcutaneous tissue Wound healing E Powder apply on wound 56.67 Digestive system Diarrhea in cattles L Pulverized & filtrate given orally 46.67 Materials Pest control Crop pest E Powdered & applied in traditional crop storage 36.67 Unidentified Medicines Infections and Hair fungus G Apply on hair 83.33 AHU51 infestations Tonsillitis E Drop in the throat 61.11 Endocrine system Diabetics E Powder in solution & drink daily morning 72.22 Social uses Cosmetics Skin softening G Apply on skin 66.67 ∗ Endemic. ∗ Narrowly endemic. ¤ Endangered. ◊ Near threatened. ∗∗ Vulnerable. The highest frequency of citation was 100% recorded for Aloe otallensis leaf exudate used for weaning children from breast-feeding followed with 94.4% for A. yavellana leaf exudate used to treat jaundice, 91.6% for A. rivae leaf gel used for body and hair wash, and 90.9% for A. megalacantha subsp. alticola exudate to treat skin infections ( Table 3 ). The most frequently cited use of Aloe species in the study areas was for medicines (human and veterinary): 149 use-reports (76%) from 22 Aloe species, with a total number of 1607 citations (use mentions) were recorded. The least use-report was for food use-category: three use-reports (1.5%) from three Aloe species with 19 citations ( Figure 2 ). Figure 2 The number of use-reports and number of Aloe species used in each major use category. Figure 2 A total of nine different plant parts were used in the diverse bio-cultural uses ( Figure 3 ) within the six major use categories. Leaf exudate was the most frequently sought part, accounting for 111 use-reports (49.1%) in which about 89 (80.2%) were for human medicinal formulations and 12 (11.7%) were for livestock medicinal formulations. That means, 92.9% of exudate were used for medicinal purposes and the remainder in formulations were in social uses and vertebrate poisons categories. In addition, the entire leaf for 47 use-reports (20.8%), which also include the use of exudate as part of entire leaf. This would suggest that the use of exudate exceeds 69% of the plant parts used for medicinal purpose. The fewest use-reports were reported from formulations made from the inflorescence and pedicel, with just a single use-report, which is 0.4% each ( Figure 3 ). Figure 3 Number of uses and percentage of each part used for the respective treatment purposes. Figure 3 The local communities practiced nine types of preparation methods for medicinal applications, out of these 42% are prepared in the form of pure exudate collection from the fresh leaf to be used for different application followed by pulverization (11.5%) and concoction (10.8%) ( Figure 4 ). Figure 4 Number of preparation in each type of preparation method for medicinal uses. Figure 4 3.2 Use-reports among cultural communities The distribution of use-reports among the cultural communities in the study areas showed 161 use-reports for the Oromo community (N = 140), followed with 17 for the Somali community (N = 32), 9 for the Afar community (N = 24), and 9 use-reports for the Harari community (N = 14). In most of the study areas, the medicinal uses of Aloe species were found more popular among the Oromo community, which accounted for 125 medicinal use-reports, in which 77.6% of the total use-report of this community. The fewest uses were reported from the Afar and Harari communities with 9 use-reports each. Among the bio-cultural values, environmental use-reports like boundary marking, soil conservation, and living fence support were documented only from the Oromo community, from four Aloe species. A unique use-report "poisonous to carnivores" was reported by the Somali and Oromo communities for two Aloe species called Aloe retrospiciens Reynolds & Bally and A. ruspoliana Baker, respectively. The Oromo community used 22 out of the 23 species documented in this study (the exception was A. retrospiciens ). The Somali community used four species ( A. retrospiciens, A . megalacantha subs. megalacantha, A. mcloughlinii , and A. pirottaei ) followed by the Afar community who use two species, A. retrospiciens and A . trichosantha subsp. longiflora and the Harari community who also use two species, A. trichosantha subsp. longiflora and A. macrocarpa. 3.3 Ethno-medicinal and bio-cultural knowledge among gender and age categories The level of knowledge of Aloe species diversity and use-reports are noticeably lower in the younger age categories compared to elderly informants. For example, out of the total citations (1336) for human medicinal uses, 654 were by elderly people (above 60 years, N = 35) followed with 398 use mention by adults (age 41–60, N = 108) and 284 by young informants (age 25–40, N = 67). The highest number of use-report in all major use categories such as medicines (Md), social uses (SU), materials (Mt), environmental uses (EU), vertebrate poisons (VP), and food (Fd) were reported by elderly people ( Figure 5 ). In the medicines category, 21 species were reported by elderly people (above 60 years, N = 35) followed by 16 species by adults (age 41–60, N = 108) and eight species by young informants (age 25–40, N = 67) ( Figure 6 ). Figure 5 Number of use-reports in each major use category for each age group (medicines (Md), social uses (SU), materials (Mt), environmental uses (EU), vertebrate poisons (VP), and food (Fd)). Figure 5 Figure 6 The number of Aloe species reported in each major use category for each age group (Md), social uses (SU), materials (Mt), environmental uses (EU), vertebrate poisons (VP), and food (Fd)). Figure 6 The depth of comprehensive bio-cultural knowledge among different age groups indicated that elderly people (above 60 years) had a much deeper knowledge than the two age group ranging from 25 to 40 and 41–60 (Kruskal-Wallis chi-squared = 12.17, df = 3, p = 0.006∗). There is a significant difference in the depth of ethno-medicinal knowledge between the 25–40 age group and the above 60 age group (p = 0.0004∗∗), and a significant difference between 41-60 year olds and the above 60 age group (p = 0.02∗) (p 0.05). 3.4 Informant consensus factor and value indexes Informant consensus factors (F ic ) of the seven major use categories ranged from 0.86 to 0.22. The highest F ic value 0.86 was for medicines with 22 species and 11 use sub-categories, followed by 0.63 for environmental uses with four species and three sub-categories, and least F ic value was 0.22 for materials with eight species and two sub-categories ( Table 4 ). Table 4 Major use categories with the corresponding sub-categories, number of Aloe species used and informant consensus factor values. Table 4 Major use categories with sub-categories No. of sub-category N. of use reports No. of citation No. of species F ic Medicines : Blood and cardiovascular system, cancer, digestive system, infections and infestations, general ailments with unspecific symptoms, endocrine system, snake and spider bites, pregnancy, birth and puerperial, reproductive system and sex health, sensory system, skin and subcutaneous tissue, & musculo-skeletal system 12 146 (76%) 1607 22 0.86 Social uses : Weaning, illuminant, magic, memorial, tattoo & cosmetics 7 20 (10.2%) 276 13 0.37 Materials: Domestic utensils & soap making 2 10 (5.1%) 107 8 0.22 Environmental uses : Soil improver, boundaries, & barrier 3 9 (4.6%) 112 4 0.63 Vertebrate Poisons : Poison, carnivore prevention, rodent control, & pest control 6 5 (2.6%) 37 3 0.50 Food: Metabolic system & edible 2 3 (1.5%) 19 2 0.50 Aloe species were compared based on the values of importance metrics. Some species like Aloe calidophila, A. megalacantha subs. alticola, A. pirottae, and A. gilbertii subsp. gilbertii showed higher values across the indices due to be mentioned by a higher number of informants, relatively higher number of use-categories, and larger number of use-reports to elucidate species that were of high importance in all indices ( Table 5 ). Table 5 Relative Importance metrics of Aloe species used in the Great Rift Valley floristic regions of Ethiopia. Table 5 Species UC UR UV RFC RI CV Aloe calidophila 4 18 11.72 0.68 0.89 6.20 Aloe megalacantha subsp. alticola 3 23 10.89 0.52 0.88 5.66 Aloe pirottae 3 14 6.92 0.49 0.66 2.06 Aloe gilbertii subsp. gilbertii 3 10 6.00 0.60 0.66 1.55 Aloe pubescens 4 12 5.86 0.49 0.62 1.49 Aloe megalacantha subsp. megalacantha 2 9 5.63 0.63 0.66 1.39 Aloe yavellana 4 8 5.00 0.63 0.63 1.10 Aloe harlana 2 10 4.83 0.48 0.57 1.01 Aloe secundiflora 1 8 4.55 0.57 0.59 0.90 Aloe trichosantha subsp. longiflora 4 20 3.14 0.24 0.61 0.64 Aloe otallensis 2 7 4.92 0.61 0.60 0.91 Aloe macrocarpa 2 9 3.44 0.55 0.60 0.74 Aloe rivae 3 5 3.91 0.58 0.53 0.49 Aloe citrina 1 6 4.12 0.51 0.51 0.55 Aloe rugosifolia 1 5 2.94 0.39 0.39 0.25 Unidentified (AHU51) 2 4 2.83 0.61 0.54 0.30 Unidentified (AHU53) 2 5 2.60 0.52 0.49 0.28 Aloe tewoldei 2 5 2.50 0.33 0.34 0.18 Aloe mcloughlinii 1 6 2.42 0.40 0.43 0.25 Aloe lateritia 2 4 2.25 0.56 0.50 0.22 Aloe welmelensis 1 2 1.50 0.38 0.32 0.12 Aloe retrospiciens 1 3 1.25 0.42 0.37 0.07 Aloe ruspoliana 2 3 1.17 0.28 0.27 0.05 UC = number of major use categories per species; UR = number of use-reports per species; UV = use value of a species; RFC = relative frequency of citation for species; RI = relative importance index; CV = cultural value index of species, which considered the three factors i.e. s (species), i (informants), and u (uses). 4 Discussion 4.1 Ethno-medicinal and bio-cultural values of aloes Most of the Aloe species were reported to have multiple uses, but the majority of uses were medicinal [ 42 , 43 ]. In this study, it has been shown that most common traditional use of Aloe species was in the traditional health care system. This compares closely with the value of 74% of literature-based use records describing the medicinal uses of Aloe [ 28 ] and 73% for human medicines from 11 Aloe species in Tanzania [ 43 ]. There are 12 sub-categories under the medicine (Md) major use category like the most cited applications of Aloe species were for skin and subcutaneous tissue ailments accounted for 29.6%, followed with infections and infestations 26.5%, and digestive system 14.5%. Aloe exudates was the most cited Aloe part used in the medicinal uses deserve further investigation into the phytochemistry and pharmacological activities. The beneficial health-promoting properties present in the exudates of Aloe leaves are attributed to their diverse phytochemical composition [ 44 ]. Phytochemical studies on the genus Aloe indicated that over 200 compounds belonging to different classes such as anthrones, chromones, pyrones, coumarins, alkaloids, glycoproteins, naphthalenes, anthraquinones and flavonoids have so far been reported [ 44 , 45 , 46 , 47 ]. These classes of compounds have been shown to possess antiviral, anti-tumor, and antibacterial activities [ 10 , 13 , 28 , 42 ]. Numerous in vitro and in vivo pharmacological and clinical trials have been revealed the traditional uses of Aloe including wound healing, anti-ulcer, anti-diabetic, hypoglycaemic, anti-cancer, anti-bacterial, anti-viral, and anti-hyperlipidemic activities [ 10 , 42 , 48 ]. A unique use-report: "poisonous to carnivores" was reported for two Aloe species: Aloe retrospiciens and A. ruspoliana . This study is the second to sample A. retrospiciens , 60 years after the first collection of a type specimen for this species. There was no any data on this species except that it was listed as vulnerable (VU) in IUCN category with an unknown population trend [ 22 ]. Meat painted with A. ruspoliana is used as bait to kill hyenas [ 49 ]. Similarly, Aloe buettneri , A. lateritia , A. rabaiensis , A. secundiflora , and A. zebrina have been documented as ingredients in arrow poisons throughout Africa [ 50 ]. The leaf exudates of these two species have an unpleasant smell and found free from the most common Aloes compound called aloin using TLC profile using pure aloin as standard. The higher frequency of citations of some use-reports can be explained by the fact that these Aloe species are best known and have long been used by the majority of informants, representing a source of reliability. For example, the highest FC value for A. otallensis used for weaning breast feeding child could be attributed for the more effectiveness of its exudate, which is much bitter than others. Similarly, the local communities in Kenya claimed that Aloe lateritia is not as bitter as A. secundiflora , and hence not as effective [ 7 ] so the community has preference of specific Aloe species for the purpose. Among the 23 Aloe species reported in this study, five species were reported whose gel is used for cosmetic purposes: Aloe calidophila, A. lateritia, A. macrocarpa , A. rivae , and the unidentified/AHU 51. The cosmetic value reported with higher FC for A. rivae makes this species a good candidate for detailed study of its gel. Taxonomic reports have indicated that A. lateritia and A. macrocarpa are grouped in the Saponaria [ 19 ] or maculate group, a name that came from the Latin "sapo" meaning soap, as the gel makes a soapy lather in water. The gel of AHU51 is used for cosmetic purposes, which has been observed while making soapy lather in water. Though A. calidophila is not in the Saponaria group, its gel is used for soap making and in few sites of the study areas two women associations were observed cultivating this species and used in small-scale soap production for local market, branded as " Yoya " meaning 'peace' in Borena and labelled to treat skin fungus and infections. It has been reported that the gel of Aloe species exhibited faster wound healing effects [ 48 ]. 4.2 Use-reports among the cultural communities People of the five floristic regions use 23 species of Aloe , and the diversity in uses suggests that species are not used interchangeably. The highest number of use-reports, recorded for A. calidophila, A. megalacantha subs. megalacantha, A. trichosantha subsp. longiflora, and A. pirottae could be associated with the wider geographic distribution of species and their use by different ethnic communities, a pattern that has been noted previously in Kenya [ 7 ]. The higher number of Aloe species and use-reports from the Oromo community could be attributed to the diverse ecology and the wider geographic distribution of Aloe species in the floristic regions inhabited by the Oromo community as compared to the other three ethnic communities. It was verified that there were no significant difference (p-value = 0.061) between the uses of Aloe species among the three cultural/ethinic communities (Afar, Somali, and Harari). 4.3 Ethno-medicinal and bio-cultural knowledge among gender and age categories The practice of traditional medicine is restricted to men in Ethiopia [ 51 , 52 ] and, not surprisingly, men were found to hold deeper knowledge of the medicinal uses of Aloe. However, women were found to be more knowledgeable on the cultural uses of aloes such as weaning children from breastfeeding, cosmetics, colouring hand and leg/tattoo, and body fumigation. Most of the young and some adult informants were found to be less knowledgeable about the diversity of Aloe species and their uses as compared to majority of adults and elderly informants. This could be attributed to the current cultural transformation, and increasing accessibility of education and health centers in rural areas that has challenged the acquisition of indigenous knowledge among young people. An alternative interpretation is that knowledge accumulates over time, and that older people have had longer to learn about plant uses. Similar results were reported in some other cultural groups in Ethiopia [ 53 , 54 ] and among users of Aloe species in Tanzania [ 43 ]. Some studies indicated that valuable ethno-medicinal information was shared with researchers mostly from informants over 60 years of age [ 55 ]. In addition, the knowledge of older people might not have been affected by the need to find new subsistence activities, and was thus preserved without external influence. For example, the dried and powdered exudate of Aloe harlana, A. megalacantha subs. alticola, and A. megalacantha subsp. megalacantha locally known as sibrii, which was used as colon cleaner, were sold in the open local market places of nearby towns. The processing, selling, and use of this local product is mainly restricted to elderly people. The knowledge of medicinal values of most Aloe species was particularly evident among elderly informants and also still retained with the majority of adults, but younger participants showed much less knowledge, and if knowledge gain has indeed slowed down, this could have negative consequences for their conservation in future. 4.4 Informant consensus factor and value indexes The most popular indices used to evaluate the relative importance of the different species, which are used in the traditional system are based on "informant consensus," i.e., the degree of agreement among the various interviewees [ 40 , 56 ]. In this respect, there was a strong consensus for the medicinal value of the leaf exudate. Efficacy of traditional medicinal plants is strongly correlated with Fic value, meaning pharmacologically effective remedies are expected to have greater Fic value, and vice versa [ 38 ]. The species with higher values in all indices of UV, RFC, RI, and CV like A. calidophila, A. megalacantha subs. megalacantha, A. gilbertii subsp. gilbertii, and A. pirottae were identified to be the most ethno-medicinal and bio-culturally important species. 4.1 Ethno-medicinal and bio-cultural values of aloes Most of the Aloe species were reported to have multiple uses, but the majority of uses were medicinal [ 42 , 43 ]. In this study, it has been shown that most common traditional use of Aloe species was in the traditional health care system. This compares closely with the value of 74% of literature-based use records describing the medicinal uses of Aloe [ 28 ] and 73% for human medicines from 11 Aloe species in Tanzania [ 43 ]. There are 12 sub-categories under the medicine (Md) major use category like the most cited applications of Aloe species were for skin and subcutaneous tissue ailments accounted for 29.6%, followed with infections and infestations 26.5%, and digestive system 14.5%. Aloe exudates was the most cited Aloe part used in the medicinal uses deserve further investigation into the phytochemistry and pharmacological activities. The beneficial health-promoting properties present in the exudates of Aloe leaves are attributed to their diverse phytochemical composition [ 44 ]. Phytochemical studies on the genus Aloe indicated that over 200 compounds belonging to different classes such as anthrones, chromones, pyrones, coumarins, alkaloids, glycoproteins, naphthalenes, anthraquinones and flavonoids have so far been reported [ 44 , 45 , 46 , 47 ]. These classes of compounds have been shown to possess antiviral, anti-tumor, and antibacterial activities [ 10 , 13 , 28 , 42 ]. Numerous in vitro and in vivo pharmacological and clinical trials have been revealed the traditional uses of Aloe including wound healing, anti-ulcer, anti-diabetic, hypoglycaemic, anti-cancer, anti-bacterial, anti-viral, and anti-hyperlipidemic activities [ 10 , 42 , 48 ]. A unique use-report: "poisonous to carnivores" was reported for two Aloe species: Aloe retrospiciens and A. ruspoliana . This study is the second to sample A. retrospiciens , 60 years after the first collection of a type specimen for this species. There was no any data on this species except that it was listed as vulnerable (VU) in IUCN category with an unknown population trend [ 22 ]. Meat painted with A. ruspoliana is used as bait to kill hyenas [ 49 ]. Similarly, Aloe buettneri , A. lateritia , A. rabaiensis , A. secundiflora , and A. zebrina have been documented as ingredients in arrow poisons throughout Africa [ 50 ]. The leaf exudates of these two species have an unpleasant smell and found free from the most common Aloes compound called aloin using TLC profile using pure aloin as standard. The higher frequency of citations of some use-reports can be explained by the fact that these Aloe species are best known and have long been used by the majority of informants, representing a source of reliability. For example, the highest FC value for A. otallensis used for weaning breast feeding child could be attributed for the more effectiveness of its exudate, which is much bitter than others. Similarly, the local communities in Kenya claimed that Aloe lateritia is not as bitter as A. secundiflora , and hence not as effective [ 7 ] so the community has preference of specific Aloe species for the purpose. Among the 23 Aloe species reported in this study, five species were reported whose gel is used for cosmetic purposes: Aloe calidophila, A. lateritia, A. macrocarpa , A. rivae , and the unidentified/AHU 51. The cosmetic value reported with higher FC for A. rivae makes this species a good candidate for detailed study of its gel. Taxonomic reports have indicated that A. lateritia and A. macrocarpa are grouped in the Saponaria [ 19 ] or maculate group, a name that came from the Latin "sapo" meaning soap, as the gel makes a soapy lather in water. The gel of AHU51 is used for cosmetic purposes, which has been observed while making soapy lather in water. Though A. calidophila is not in the Saponaria group, its gel is used for soap making and in few sites of the study areas two women associations were observed cultivating this species and used in small-scale soap production for local market, branded as " Yoya " meaning 'peace' in Borena and labelled to treat skin fungus and infections. It has been reported that the gel of Aloe species exhibited faster wound healing effects [ 48 ]. 4.2 Use-reports among the cultural communities People of the five floristic regions use 23 species of Aloe , and the diversity in uses suggests that species are not used interchangeably. The highest number of use-reports, recorded for A. calidophila, A. megalacantha subs. megalacantha, A. trichosantha subsp. longiflora, and A. pirottae could be associated with the wider geographic distribution of species and their use by different ethnic communities, a pattern that has been noted previously in Kenya [ 7 ]. The higher number of Aloe species and use-reports from the Oromo community could be attributed to the diverse ecology and the wider geographic distribution of Aloe species in the floristic regions inhabited by the Oromo community as compared to the other three ethnic communities. It was verified that there were no significant difference (p-value = 0.061) between the uses of Aloe species among the three cultural/ethinic communities (Afar, Somali, and Harari). 4.3 Ethno-medicinal and bio-cultural knowledge among gender and age categories The practice of traditional medicine is restricted to men in Ethiopia [ 51 , 52 ] and, not surprisingly, men were found to hold deeper knowledge of the medicinal uses of Aloe. However, women were found to be more knowledgeable on the cultural uses of aloes such as weaning children from breastfeeding, cosmetics, colouring hand and leg/tattoo, and body fumigation. Most of the young and some adult informants were found to be less knowledgeable about the diversity of Aloe species and their uses as compared to majority of adults and elderly informants. This could be attributed to the current cultural transformation, and increasing accessibility of education and health centers in rural areas that has challenged the acquisition of indigenous knowledge among young people. An alternative interpretation is that knowledge accumulates over time, and that older people have had longer to learn about plant uses. Similar results were reported in some other cultural groups in Ethiopia [ 53 , 54 ] and among users of Aloe species in Tanzania [ 43 ]. Some studies indicated that valuable ethno-medicinal information was shared with researchers mostly from informants over 60 years of age [ 55 ]. In addition, the knowledge of older people might not have been affected by the need to find new subsistence activities, and was thus preserved without external influence. For example, the dried and powdered exudate of Aloe harlana, A. megalacantha subs. alticola, and A. megalacantha subsp. megalacantha locally known as sibrii, which was used as colon cleaner, were sold in the open local market places of nearby towns. The processing, selling, and use of this local product is mainly restricted to elderly people. The knowledge of medicinal values of most Aloe species was particularly evident among elderly informants and also still retained with the majority of adults, but younger participants showed much less knowledge, and if knowledge gain has indeed slowed down, this could have negative consequences for their conservation in future. 4.4 Informant consensus factor and value indexes The most popular indices used to evaluate the relative importance of the different species, which are used in the traditional system are based on "informant consensus," i.e., the degree of agreement among the various interviewees [ 40 , 56 ]. In this respect, there was a strong consensus for the medicinal value of the leaf exudate. Efficacy of traditional medicinal plants is strongly correlated with Fic value, meaning pharmacologically effective remedies are expected to have greater Fic value, and vice versa [ 38 ]. The species with higher values in all indices of UV, RFC, RI, and CV like A. calidophila, A. megalacantha subs. megalacantha, A. gilbertii subsp. gilbertii, and A. pirottae were identified to be the most ethno-medicinal and bio-culturally important species. 5 Conclusion Aloe species were reported to have multiple uses, but it has been shown that the most common local use was in the traditional health care system. In priority setting for Aloe -based product development and cultivation, species like, A. calidophila, A. megalacantha subsp. alticola, and A. pirottae were identified to be the top three ethno-medicinal and bio-culturally important endemic Aloe species. In addition, unidentified Aloe samples could instigate taxonomic discussion and investigation. More importantly, the unique use-report: "poisonous to carnivores" from two Aloe species called A. retrospiciens and A. ruspoliana could initiate detailed phytochemical and toxicity studies, which have not been done so far, and are of particular importance for this study. In this study, it has been shown that the leaf exudates were highly valued in medicinal applications. Therefore, research interests on medicinal values of Ethiopian endemic aloes need to focus on exudates for phytochemical and pharmacological analysis. In addition, the result showed the unfortunate decline in bio-cultural and ethnomedicinal knowledge between generations in most of the study areas. The deterioration of indigenous knowledge coupled with declining wild populations of most Aloe species could stimulate an urgent ethnobotanical study for in-depth investigation before it is lost. It should be followed with phytochemical and pharmacological analyses in order to give scientific ground to the ethnomedicinal knowledge as well as to signify conservation attention and future potential utilization. In general, the output of this comprehensive ethno-medicinal and bio-cultural knowledge will encourage the community to conserve, manage, and sustainable use Aloe species in the natural habitat as well as through cultivation. Declarations Author contribution statement A. Belayneh, S. Demissew: Conceived and designed the experiments; Performed the experiments; Analyzed and interpreted the data; Wrote the paper. N.F. Bussa: Conceived and designed the experiments; Analyzed and interpreted the data; Wrote the paper. D. Bisrat: Analyzed and interpreted the data; Contributed reagents, materials, analysis tools or data; Wrote the paper. Funding statement This work was supported by 10.13039/501100004845 Haramaya University through Ministry of Education for postgraduate/PhD studies. Competing interest statement The authors declare no conflict of interest. Additional information No additional information is available for this paper. Author contribution statement A. Belayneh, S. Demissew: Conceived and designed the experiments; Performed the experiments; Analyzed and interpreted the data; Wrote the paper. N.F. Bussa: Conceived and designed the experiments; Analyzed and interpreted the data; Wrote the paper. D. Bisrat: Analyzed and interpreted the data; Contributed reagents, materials, analysis tools or data; Wrote the paper. Funding statement This work was supported by 10.13039/501100004845 Haramaya University through Ministry of Education for postgraduate/PhD studies. Competing interest statement The authors declare no conflict of interest. Additional information No additional information is available for this paper. Appendix A Supplementary data The following is the supplementary data related to this article: Appendix I Appendix I Appendix II Appendix II

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10651316/

CDC Guidelines for the Prevention and Treatment of Anthrax, 2023

Summary This report updates previous CDC guidelines and recommendations on preferred prevention and treatment regimens regarding naturally occurring anthrax. Also provided are a wide range of alternative regimens to first-line antimicrobial drugs for use if patients have contraindications or intolerances or after a wide-area aerosol release of Bacillus anthracis spores if resources become limited or a multidrug-resistant B. anthracis strain is used (Hendricks KA, Wright ME, Shadomy SV, et al.; Workgroup on Anthrax Clinical Guidelines. Centers for Disease Control and Prevention expert panel meetings on prevention and treatment of anthrax in adults. Emerg Infect Dis 2014;20:e130687; Meaney-Delman D, Rasmussen SA, Beigi RH, et al. Prophylaxis and treatment of anthrax in pregnant women. Obstet Gynecol 2013;122:885−900; Bradley JS, Peacock G, Krug SE, et al. Pediatric anthrax clinical management. Pediatrics 2014;133:e1411-36) . Specifically, this report updates antimicrobial drug and antitoxin use for both postexposure prophylaxis (PEP) and treatment from these previous guidelines best practices and is based on systematic reviews of the literature regarding 1) in vitro antimicrobial drug activity against B. anthracis ; 2) in vivo antimicrobial drug efficacy for PEP and treatment; 3) in vivo and human antitoxin efficacy for PEP, treatment, or both; and 4) human survival after antimicrobial drug PEP and treatment of localized anthrax, systemic anthrax, and anthrax meningitis. Changes from previous CDC guidelines and recommendations include an expanded list of alternative antimicrobial drugs to use when first-line antimicrobial drugs are contraindicated or not tolerated or after a bioterrorism event when first-line antimicrobial drugs are depleted or ineffective against a genetically engineered resistant B. anthracis strain. In addition, these updated guidelines include new recommendations regarding special considerations for the diagnosis and treatment of anthrax meningitis, including comorbid, social, and clinical predictors of anthrax meningitis. The previously published CDC guidelines and recommendations described potentially beneficial critical care measures and clinical assessment tools and procedures for persons with anthrax, which have not changed and are not addressed in this update. In addition, no changes were made to the Advisory Committee on Immunization Practices recommendations for use of anthrax vaccine (Bower WA, Schiffer J, Atmar RL, et al. Use of anthrax vaccine in the United States: recommendations of the Advisory Committee on Immunization Practices, 2019. MMWR Recomm Rep 2019;68[No. RR-4]:1–14) . The updated guidelines in this report can be used by health care providers to prevent and treat anthrax and guide emergency preparedness officials and planners as they develop and update plans for a wide-area aerosol release of B. anthracis . This report updates previous CDC guidelines and recommendations on preferred prevention and treatment regimens regarding naturally occurring anthrax. Also provided are a wide range of alternative regimens to first-line antimicrobial drugs for use if patients have contraindications or intolerances or after a wide-area aerosol release of Bacillus anthracis spores if resources become limited or a multidrug-resistant B. anthracis strain is used (Hendricks KA, Wright ME, Shadomy SV, et al.; Workgroup on Anthrax Clinical Guidelines. Centers for Disease Control and Prevention expert panel meetings on prevention and treatment of anthrax in adults. Emerg Infect Dis 2014;20:e130687; Meaney-Delman D, Rasmussen SA, Beigi RH, et al. Prophylaxis and treatment of anthrax in pregnant women. Obstet Gynecol 2013;122:885−900; Bradley JS, Peacock G, Krug SE, et al. Pediatric anthrax clinical management. Pediatrics 2014;133:e1411-36) . Specifically, this report updates antimicrobial drug and antitoxin use for both postexposure prophylaxis (PEP) and treatment from these previous guidelines best practices and is based on systematic reviews of the literature regarding 1) in vitro antimicrobial drug activity against B. anthracis ; 2) in vivo antimicrobial drug efficacy for PEP and treatment; 3) in vivo and human antitoxin efficacy for PEP, treatment, or both; and 4) human survival after antimicrobial drug PEP and treatment of localized anthrax, systemic anthrax, and anthrax meningitis. Changes from previous CDC guidelines and recommendations include an expanded list of alternative antimicrobial drugs to use when first-line antimicrobial drugs are contraindicated or not tolerated or after a bioterrorism event when first-line antimicrobial drugs are depleted or ineffective against a genetically engineered resistant B. anthracis strain. In addition, these updated guidelines include new recommendations regarding special considerations for the diagnosis and treatment of anthrax meningitis, including comorbid, social, and clinical predictors of anthrax meningitis. The previously published CDC guidelines and recommendations described potentially beneficial critical care measures and clinical assessment tools and procedures for persons with anthrax, which have not changed and are not addressed in this update. In addition, no changes were made to the Advisory Committee on Immunization Practices recommendations for use of anthrax vaccine (Bower WA, Schiffer J, Atmar RL, et al. Use of anthrax vaccine in the United States: recommendations of the Advisory Committee on Immunization Practices, 2019. MMWR Recomm Rep 2019;68[No. RR-4]:1–14) . The updated guidelines in this report can be used by health care providers to prevent and treat anthrax and guide emergency preparedness officials and planners as they develop and update plans for a wide-area aerosol release of B. anthracis . Introduction Anthrax is a zoonotic disease caused by infection with Bacillus anthracis and still occurs in agricultural regions of the Americas, sub-Saharan Africa, central and southwestern Asia, and southern and eastern Europe ( 1 ). Sheep, goats, cattle, and other herbivores are primarily affected. Humans are secondarily infected through contact with infected animals, contaminated animal products (e.g., meat or hides), or, rarely, from injection drug use ( 2 ). Anthrax in humans usually is characterized by the route of B. anthracis inoculation ( 3 ). Cutaneous anthrax, which results from direct inoculation of spores through the skin, is the most common form and accounts for >95% of human cases. Ingestion anthrax usually results from consumption of infected meat. Inhalation anthrax results from the inhalation of aerosolized spores. Injection anthrax, which is a relatively new form, results from injection of heroin contaminated with B. anthracis spores. Anthrax meningitis can complicate any form of anthrax or occur alone. In the United States, anthrax has almost been eliminated through livestock vaccination. Wildlife and livestock anthrax still occurs sporadically in an area from southwest Texas through Colorado, North and South Dakota, and Montana ( 4 , 5 ). Since 2006, nine confirmed or probable U.S. cases of anthrax have been reported to CDC: two inhalation ( 6 , 7 ), one ingestion ( 8 ), four cutaneous ( 9 , 10 ), and two with no documented route of infection. In addition, since 1997, seven cases of severe pneumonia have been identified that were caused by Bacillus cereus group species that harbor a plasmid that encodes anthrax toxins similar to those found in B. anthracis ( 11 ). B. anthracis is a Tier 1 select agent and considered one of the most likely bioterrorism agents to be used because it is relatively easy to acquire from the natural environment, mass produce, and disseminate as spores via aerosolization ( 12 ). Although approximately 180 countries have signed on to the 1975 Biological Weapons Convention prohibiting the development, production, acquisition, transfer, stockpiling, and use of bioweapons, a wide-area aerosol release of B. anthracis spores remains a concern. In 2001, B. anthracis spores were sent in letters through the U.S. Postal Service. Exposure to aerosolized spores in these letters resulted in 11 inhalation and 11 cutaneous cases; five inhalation cases were fatal ( 13 , 14 ). A wide-area aerosol release of B. anthracis spores would likely result in a mass-casualty incident ( 15 , 16 ) that could possibly be complicated by use of genetically engineered B. anthracis strains resistant to first-line antimicrobial drugs for postexposure prophylaxis (PEP) and treatment. In 2015, CDC published recommendations for hospital-based acute care that addressed antitoxin and intravenous (IV) antimicrobial drug use and the diagnosis and management of common anthrax-specific complications during a mass-casualty incident that would require a shift from conventional to contingency or crisis standards of care ( 17 ). B. anthracis possesses three primary virulence factors: an extracellular capsule and two bipartite exotoxins (lethal toxin [composed of lethal factor and protective antigen] and edema toxin [composed of edema factor and protective antigen]) ( 18 – 20 ). The capsule prevents phagocytosis of the vegetative form of B. anthracis by macrophages, allowing it to evade the immune system. The two toxins also facilitate immune system evasion by disrupting various immune cell functions (e.g., cellular signaling and cell migration) ( 21 , 22 ). In addition, they impair macrophages, neutrophils, and dendritic cell functions ( 23 , 24 ) and inhibit host B- and T-cell immune responses ( 25 ). Lethal toxin causes apoptosis of endothelial cells within the vascular system and is thought to contribute to hemorrhage ( 26 ). These combined effects allow bacterial proliferation and lead to the high morbidity and mortality associated with anthrax. Historically, antiserum appeared to be an effective treatment. In aggregate data from the preantibiotic era, patients with cutaneous anthrax who were treated with antiserum had a substantially lower mortality rate than those who remained untreated (7.6% for patients who received antiserum during 1903−1941 compared with 23.7% for patients who did not receive antiserum during 1888−1920) ( 27 ). Before the bioterrorism-related inhalation anthrax cases in 2001 ( 13 ), the anthrax mortality rate for the cases before the 1960s approached 90% for inhalation anthrax ( 28 ) and neared 100% for anthrax meningitis ( 29 ). Mortality rates have improved with advancements in critical care; however, even with treatment, mortality ranges from 4/76 >4/76 96 Source: Maxson T, Kongphet-Tran T, Mongkolrattanothai T, et al. Systematic review of in vitro antimicrobial susceptibility testing for Bacillus anthracis , 1947–2019. Clin Infect Dis 2022;75(Suppl 3):S373–8. Abbreviations: MIC 50 = minimum inhibitory concentration required to inhibit the growth of 50% of isolates tested; MIC 90 = minimum inhibitory concentration required to inhibit the growth of 90% of isolates tested. * MIC 50 and MIC 90 values from studies of ciprofloxacin, levofloxacin, moxifloxacin, clarithromycin, penicillin, amoxicillin, amoxicillin/clavulanate, ampicillin/sulbactam, doxycycline, tetracycline, linezolid, clindamycin, and vancomycin align with findings of an unpublished CDC study of 50 wildtype Bacillus anthracis strains by conventional broth microdilution. The CDC MIC 50 and MIC 90 values were within ±1 doubling dilution concentration to those listed. †MIC 50 and MIC 90 values were omitted if 30 breaths per minute. ¶ Administered vasopressors or systolic blood pressure 38.0°C or >100.4°F), tachypnea (>20 breaths per minutes), tachycardia (≥100 beats per minute), systolic and diastolic hypertension (≥130 and ≥90 mm Hg, respectively), and evidence of coagulopathy. Presenting bacteremia was predictive of meningitis in one of the studies and bacteremia at any point throughout hospitalization was predictive in another. TABLE 4 Presenting signs, symptoms, and laboratory test results associated with meningitis for adults with systemic or cutaneous anthrax Symptom/Sign/Laboratory test result Patient population Systemic anthrax* ,†(N = 419) Cutaneous anthrax †,§ (N = 182) Cutaneous anthrax ¶ (N = 126) Time frame 1880–2018 1950–2018 2005–2012 Symptom Fever/chills X X X Nausea/vomiting X X X Headache X X X Abdominal pain ANA X ANA Sign Evidence of coagulopathy (i.e., bleeding) ANA X NE Vital sign Fever (>38.0 ° C) ANA ANA X Tachycardia (>90 beats per minute) NE ANA NE Tachycardia (≥100 beats per minute) ANA NE X Tachypnea (>20 breaths per minute) ANA ANA X Systolic hypertension (≥130 mm Hg) NE NE X Diastolic hypertension (≥90 mm Hg) NE NE X Fulminant X NE NE Dermatologic Lymphadenopathy NE X NE Thoracic edema ANA X ANA Malignant pustule edema ANA X ANA Neurologic Altered mental status X ANA ANA Meningeal signs X NE NE Nonheadache, nonmeningeal neurologic signs X X NE Diagnostic laboratory test result Bacteremia at presentation ANA ANA X Bacteremia at any time throughout hospitalization ANA X NE Abbreviations: ANA = analyzed and not associated; NE = not evaluated; X = predictive of meningitis. * Source: Binney S, Person MK, Traxler RM, Cook R, Bower WA, Hendricks K. Algorithms for the identification of anthrax meningitis during a mass casualty event based on a systematic review of systemic anthrax from 1880 through 2018. Clin Infect Dis 2022;75(Suppl 3):S468–77. †Results from Binney et al. and Thompson et al. shared overlapping data sets. § Source: Thompson JM, Cook R, Person MK, et al. Risk factors for death or meningitis in adults hospitalized for cutaneous anthrax, 1950–2018: a systematic review. Clin Infect Dis 2022;75(Suppl 3):S459–67. ¶ Source: Kutmanova A, Zholdoshev S, Roguski KM, et al. Risk factors for severe cutaneous anthrax in a retrospective case series and use of a clinical algorithm to identify likely meningitis and evaluate treatment outcomes, Kyrgyz Republic, 2005–2012. This patient population included one patient aged 38.5°C for fever, >110 beats per minute for tachycardia, and >14 breaths per minute for tachypnea). Triage Tool for Predicting Meningitis in a Mass Casualty Event In 2016, a study described a 4-item anthrax meningitis assessment (i.e., screening) tool that included severe headache, altered mental status, meningeal signs, and other neurologic signs ( 30 ). Using this tool, patients were deemed likely to have meningitis if two or more of the items were present and unlikely to have meningitis if no items were present. Patients with only one of the items were in a middle group that would need diagnostic testing (e.g., imaging and lumbar puncture) to determine their meningitis status. A study ( 41 ) validated the meningitis screening tool ( 30 ) and another evaluated newly identified risk factors ( 40 ). On re-evaluation, the 2016 screening tool had a sensitivity of 86% (95% CI = 71%–100%) and a specificity of 99% (95% CI = 97%–100%). However, >17% of adults still needed further diagnostic workup to determine presence or absence of meningitis. A new, more complex screening tool was developed to minimize the number of patients who would require further testing ( Figure ). This screening tool had a sensitivity of 86% (95% CI = 71%–100%) and a specificity of 92% (95% CI = 85%–99%) but left only 2.5% of adults in need of further testing ( 41 ). FIGURE Screening tool to identify potential anthrax meningitis cases by presenting signs and symptoms after a mass casualty event when diagnostic capability is limited — CDC, 2023 Abbreviations: CSF = cerebrospinal fluid; CT = computed tomography. Figure illustrates the steps in a screening tool to identify potential anthrax meningitis cases by presenting signs and symptoms after a mass casualty event when diagnostic capability is limited. Source: Adapted from Figure 2 in Binney S, Person MK, Traxler RM, Cook R, Bower WA, Hendricks K. Algorithms for the identification of anthrax meningitis during a mass casualty event based on a systematic review of systemic anthrax from 1880 through 2018. Clin Infect Dis 2022;75(Suppl 3):S468–77. Antimicrobial Drug Selection In the Kyrgyz Republic case series, 37 patients were categorized by an algorithm as likely having meningitis and 89 were considered unlikely to have meningitis ( 52 ). Twenty-three (62%) of the 37 received monotherapy, and all 23 survived. The remainder (38%) received 2-drug combination therapy and survived. Only five of 14 survivors who received combination therapy received a bactericidal and a PSI antimicrobial drug, and none received three concurrent antimicrobial drugs. In the systematic review of human treatment data, survival was not improved by adding a PSI to a bactericidal antimicrobial drug in adults with anthrax meningitis ( 39 ). However, median length of stay for the three patients who received both a PSI and a bactericidal antimicrobial drug was 10 days (IQR = 8–21 days) compared with 30 days (IQR = 15–50 days) for the four who received bactericidal antimicrobial drugs alone, although this difference was not statistically significant (p = 0.26) ( 39 ). Population PK Modeling to Predict Drug Exposures in Cerebrospinal Fluid for Naturally Occurring B. anthracis Strains Antimicrobial drugs with published human cerebrospinal fluid (CSF) penetration data were categorized according to their ability to achieve robust drug exposures in both plasma and CSF based on population PK modeling followed by Monte Carlo simulations. Because PK/PD target values for B. anthracis are scarce, established targets from other pathogens were used to evaluate drug exposures. One study suggested the PK/PD targets for the carbapenem class of antibiotics (e.g., faropenem) were comparable between B. anthracis and gram-negative pathogens. The blood-brain barrier can be impaired considerably during anthrax meningitis; therefore, both the plasma and the CSF exposures were considered. Because of the need for between-pathogen extrapolations of PK/PD targets, absence of CSF-specific targets, and relative sparsity of the data, these Monte Carlo simulation results should be interpreted conservatively. The Monte Carlo simulations indicated the probability that patients with suspected or documented anthrax meningitis would achieve microbiologic cure of CSF when treated with the recommended dosage was high (≥95%) for ampicillin, ampicillin/sulbactam, ciprofloxacin, doxycycline, imipenem/cilastatin, levofloxacin 500 mg every 12 hours, meropenem, minocycline, and penicillin G; acceptable (≥90% to MIC = duration of the unbound (i.e. free) concentration above the minimum inhibitory concentration (expressed as a fraction of the dosing interval); IV = intravenous; MIC = minimum inhibitory concentration; MIC 90 = minimum inhibitory concentration required to inhibit the growth of 90% of organisms; PD = pharmacodynamic; PK = pharmacokinetic; S = bacteriostatic. * Highest MIC ( µ g/mL) with a >95% probability of successful CSF treatment. †PK/PD targets used were fAUC/MIC ≥18 for doxycycline and minocycline, fAUC/MIC ≥75 for quinolones, fT>MIC ≥40% for carbapenems, fT>MIC ≥50% for penicillins, fAUC/MIC ≥90 for amikacin, fAUC/MIC ≥27 for clindamycin, AUC/MIC ≥110 for linezolid, and AUC/MIC ≥400 for vancomycin. § High probability is defined as a ≥95% probability of treatment success. ¶ PK/PD target achievement and the highest treatable MICs were similar (i.e., within twofold) in CSF and plasma for ciprofloxacin, levofloxacin, doxycycline, and minocycline. ** For the studied ß-lactams (i.e., meropenem, imipenem, penicillin G, ampicillin, and piperacillin) and for amikacin, the highest treatable MICs based on the plasma targets were twofold to thirty-two-fold higher compared with those in CSF. ††Previous susceptibility testing to infecting strain is recommended. Bacillus anthracis naturally carries penicillin-resistance genes for a ß-lactamase that is expressed in up to 10% of naturally occurring strains, creating resistance to penicillin G, ampicillin, and piperacillin. §§ Acceptable probability is defined as a ≥90% to 4/76 >4/76 96 Source: Maxson T, Kongphet-Tran T, Mongkolrattanothai T, et al. Systematic review of in vitro antimicrobial susceptibility testing for Bacillus anthracis , 1947–2019. Clin Infect Dis 2022;75(Suppl 3):S373–8. Abbreviations: MIC 50 = minimum inhibitory concentration required to inhibit the growth of 50% of isolates tested; MIC 90 = minimum inhibitory concentration required to inhibit the growth of 90% of isolates tested. * MIC 50 and MIC 90 values from studies of ciprofloxacin, levofloxacin, moxifloxacin, clarithromycin, penicillin, amoxicillin, amoxicillin/clavulanate, ampicillin/sulbactam, doxycycline, tetracycline, linezolid, clindamycin, and vancomycin align with findings of an unpublished CDC study of 50 wildtype Bacillus anthracis strains by conventional broth microdilution. The CDC MIC 50 and MIC 90 values were within ±1 doubling dilution concentration to those listed. †MIC 50 and MIC 90 values were omitted if 30 breaths per minute. ¶ Administered vasopressors or systolic blood pressure 30 breaths per minute. ¶ Administered vasopressors or systolic blood pressure 38.0°C or >100.4°F), tachypnea (>20 breaths per minutes), tachycardia (≥100 beats per minute), systolic and diastolic hypertension (≥130 and ≥90 mm Hg, respectively), and evidence of coagulopathy. Presenting bacteremia was predictive of meningitis in one of the studies and bacteremia at any point throughout hospitalization was predictive in another. TABLE 4 Presenting signs, symptoms, and laboratory test results associated with meningitis for adults with systemic or cutaneous anthrax Symptom/Sign/Laboratory test result Patient population Systemic anthrax* ,†(N = 419) Cutaneous anthrax †,§ (N = 182) Cutaneous anthrax ¶ (N = 126) Time frame 1880–2018 1950–2018 2005–2012 Symptom Fever/chills X X X Nausea/vomiting X X X Headache X X X Abdominal pain ANA X ANA Sign Evidence of coagulopathy (i.e., bleeding) ANA X NE Vital sign Fever (>38.0 ° C) ANA ANA X Tachycardia (>90 beats per minute) NE ANA NE Tachycardia (≥100 beats per minute) ANA NE X Tachypnea (>20 breaths per minute) ANA ANA X Systolic hypertension (≥130 mm Hg) NE NE X Diastolic hypertension (≥90 mm Hg) NE NE X Fulminant X NE NE Dermatologic Lymphadenopathy NE X NE Thoracic edema ANA X ANA Malignant pustule edema ANA X ANA Neurologic Altered mental status X ANA ANA Meningeal signs X NE NE Nonheadache, nonmeningeal neurologic signs X X NE Diagnostic laboratory test result Bacteremia at presentation ANA ANA X Bacteremia at any time throughout hospitalization ANA X NE Abbreviations: ANA = analyzed and not associated; NE = not evaluated; X = predictive of meningitis. * Source: Binney S, Person MK, Traxler RM, Cook R, Bower WA, Hendricks K. Algorithms for the identification of anthrax meningitis during a mass casualty event based on a systematic review of systemic anthrax from 1880 through 2018. Clin Infect Dis 2022;75(Suppl 3):S468–77. †Results from Binney et al. and Thompson et al. shared overlapping data sets. § Source: Thompson JM, Cook R, Person MK, et al. Risk factors for death or meningitis in adults hospitalized for cutaneous anthrax, 1950–2018: a systematic review. Clin Infect Dis 2022;75(Suppl 3):S459–67. ¶ Source: Kutmanova A, Zholdoshev S, Roguski KM, et al. Risk factors for severe cutaneous anthrax in a retrospective case series and use of a clinical algorithm to identify likely meningitis and evaluate treatment outcomes, Kyrgyz Republic, 2005–2012. This patient population included one patient aged 38.5°C for fever, >110 beats per minute for tachycardia, and >14 breaths per minute for tachypnea). Triage Tool for Predicting Meningitis in a Mass Casualty Event In 2016, a study described a 4-item anthrax meningitis assessment (i.e., screening) tool that included severe headache, altered mental status, meningeal signs, and other neurologic signs ( 30 ). Using this tool, patients were deemed likely to have meningitis if two or more of the items were present and unlikely to have meningitis if no items were present. Patients with only one of the items were in a middle group that would need diagnostic testing (e.g., imaging and lumbar puncture) to determine their meningitis status. A study ( 41 ) validated the meningitis screening tool ( 30 ) and another evaluated newly identified risk factors ( 40 ). On re-evaluation, the 2016 screening tool had a sensitivity of 86% (95% CI = 71%–100%) and a specificity of 99% (95% CI = 97%–100%). However, >17% of adults still needed further diagnostic workup to determine presence or absence of meningitis. A new, more complex screening tool was developed to minimize the number of patients who would require further testing ( Figure ). This screening tool had a sensitivity of 86% (95% CI = 71%–100%) and a specificity of 92% (95% CI = 85%–99%) but left only 2.5% of adults in need of further testing ( 41 ). FIGURE Screening tool to identify potential anthrax meningitis cases by presenting signs and symptoms after a mass casualty event when diagnostic capability is limited — CDC, 2023 Abbreviations: CSF = cerebrospinal fluid; CT = computed tomography. Figure illustrates the steps in a screening tool to identify potential anthrax meningitis cases by presenting signs and symptoms after a mass casualty event when diagnostic capability is limited. Source: Adapted from Figure 2 in Binney S, Person MK, Traxler RM, Cook R, Bower WA, Hendricks K. Algorithms for the identification of anthrax meningitis during a mass casualty event based on a systematic review of systemic anthrax from 1880 through 2018. Clin Infect Dis 2022;75(Suppl 3):S468–77. Antimicrobial Drug Selection In the Kyrgyz Republic case series, 37 patients were categorized by an algorithm as likely having meningitis and 89 were considered unlikely to have meningitis ( 52 ). Twenty-three (62%) of the 37 received monotherapy, and all 23 survived. The remainder (38%) received 2-drug combination therapy and survived. Only five of 14 survivors who received combination therapy received a bactericidal and a PSI antimicrobial drug, and none received three concurrent antimicrobial drugs. In the systematic review of human treatment data, survival was not improved by adding a PSI to a bactericidal antimicrobial drug in adults with anthrax meningitis ( 39 ). However, median length of stay for the three patients who received both a PSI and a bactericidal antimicrobial drug was 10 days (IQR = 8–21 days) compared with 30 days (IQR = 15–50 days) for the four who received bactericidal antimicrobial drugs alone, although this difference was not statistically significant (p = 0.26) ( 39 ). Population PK Modeling to Predict Drug Exposures in Cerebrospinal Fluid for Naturally Occurring B. anthracis Strains Antimicrobial drugs with published human cerebrospinal fluid (CSF) penetration data were categorized according to their ability to achieve robust drug exposures in both plasma and CSF based on population PK modeling followed by Monte Carlo simulations. Because PK/PD target values for B. anthracis are scarce, established targets from other pathogens were used to evaluate drug exposures. One study suggested the PK/PD targets for the carbapenem class of antibiotics (e.g., faropenem) were comparable between B. anthracis and gram-negative pathogens. The blood-brain barrier can be impaired considerably during anthrax meningitis; therefore, both the plasma and the CSF exposures were considered. Because of the need for between-pathogen extrapolations of PK/PD targets, absence of CSF-specific targets, and relative sparsity of the data, these Monte Carlo simulation results should be interpreted conservatively. The Monte Carlo simulations indicated the probability that patients with suspected or documented anthrax meningitis would achieve microbiologic cure of CSF when treated with the recommended dosage was high (≥95%) for ampicillin, ampicillin/sulbactam, ciprofloxacin, doxycycline, imipenem/cilastatin, levofloxacin 500 mg every 12 hours, meropenem, minocycline, and penicillin G; acceptable (≥90% to MIC = duration of the unbound (i.e. free) concentration above the minimum inhibitory concentration (expressed as a fraction of the dosing interval); IV = intravenous; MIC = minimum inhibitory concentration; MIC 90 = minimum inhibitory concentration required to inhibit the growth of 90% of organisms; PD = pharmacodynamic; PK = pharmacokinetic; S = bacteriostatic. * Highest MIC ( µ g/mL) with a >95% probability of successful CSF treatment. †PK/PD targets used were fAUC/MIC ≥18 for doxycycline and minocycline, fAUC/MIC ≥75 for quinolones, fT>MIC ≥40% for carbapenems, fT>MIC ≥50% for penicillins, fAUC/MIC ≥90 for amikacin, fAUC/MIC ≥27 for clindamycin, AUC/MIC ≥110 for linezolid, and AUC/MIC ≥400 for vancomycin. § High probability is defined as a ≥95% probability of treatment success. ¶ PK/PD target achievement and the highest treatable MICs were similar (i.e., within twofold) in CSF and plasma for ciprofloxacin, levofloxacin, doxycycline, and minocycline. ** For the studied ß-lactams (i.e., meropenem, imipenem, penicillin G, ampicillin, and piperacillin) and for amikacin, the highest treatable MICs based on the plasma targets were twofold to thirty-two-fold higher compared with those in CSF. ††Previous susceptibility testing to infecting strain is recommended. Bacillus anthracis naturally carries penicillin-resistance genes for a ß-lactamase that is expressed in up to 10% of naturally occurring strains, creating resistance to penicillin G, ampicillin, and piperacillin. §§ Acceptable probability is defined as a ≥90% to 38.0°C or >100.4°F), tachypnea (>20 breaths per minutes), tachycardia (≥100 beats per minute), systolic and diastolic hypertension (≥130 and ≥90 mm Hg, respectively), and evidence of coagulopathy. Presenting bacteremia was predictive of meningitis in one of the studies and bacteremia at any point throughout hospitalization was predictive in another. TABLE 4 Presenting signs, symptoms, and laboratory test results associated with meningitis for adults with systemic or cutaneous anthrax Symptom/Sign/Laboratory test result Patient population Systemic anthrax* ,†(N = 419) Cutaneous anthrax †,§ (N = 182) Cutaneous anthrax ¶ (N = 126) Time frame 1880–2018 1950–2018 2005–2012 Symptom Fever/chills X X X Nausea/vomiting X X X Headache X X X Abdominal pain ANA X ANA Sign Evidence of coagulopathy (i.e., bleeding) ANA X NE Vital sign Fever (>38.0 ° C) ANA ANA X Tachycardia (>90 beats per minute) NE ANA NE Tachycardia (≥100 beats per minute) ANA NE X Tachypnea (>20 breaths per minute) ANA ANA X Systolic hypertension (≥130 mm Hg) NE NE X Diastolic hypertension (≥90 mm Hg) NE NE X Fulminant X NE NE Dermatologic Lymphadenopathy NE X NE Thoracic edema ANA X ANA Malignant pustule edema ANA X ANA Neurologic Altered mental status X ANA ANA Meningeal signs X NE NE Nonheadache, nonmeningeal neurologic signs X X NE Diagnostic laboratory test result Bacteremia at presentation ANA ANA X Bacteremia at any time throughout hospitalization ANA X NE Abbreviations: ANA = analyzed and not associated; NE = not evaluated; X = predictive of meningitis. * Source: Binney S, Person MK, Traxler RM, Cook R, Bower WA, Hendricks K. Algorithms for the identification of anthrax meningitis during a mass casualty event based on a systematic review of systemic anthrax from 1880 through 2018. Clin Infect Dis 2022;75(Suppl 3):S468–77. †Results from Binney et al. and Thompson et al. shared overlapping data sets. § Source: Thompson JM, Cook R, Person MK, et al. Risk factors for death or meningitis in adults hospitalized for cutaneous anthrax, 1950–2018: a systematic review. Clin Infect Dis 2022;75(Suppl 3):S459–67. ¶ Source: Kutmanova A, Zholdoshev S, Roguski KM, et al. Risk factors for severe cutaneous anthrax in a retrospective case series and use of a clinical algorithm to identify likely meningitis and evaluate treatment outcomes, Kyrgyz Republic, 2005–2012. This patient population included one patient aged 38.5°C for fever, >110 beats per minute for tachycardia, and >14 breaths per minute for tachypnea). Triage Tool for Predicting Meningitis in a Mass Casualty Event In 2016, a study described a 4-item anthrax meningitis assessment (i.e., screening) tool that included severe headache, altered mental status, meningeal signs, and other neurologic signs ( 30 ). Using this tool, patients were deemed likely to have meningitis if two or more of the items were present and unlikely to have meningitis if no items were present. Patients with only one of the items were in a middle group that would need diagnostic testing (e.g., imaging and lumbar puncture) to determine their meningitis status. A study ( 41 ) validated the meningitis screening tool ( 30 ) and another evaluated newly identified risk factors ( 40 ). On re-evaluation, the 2016 screening tool had a sensitivity of 86% (95% CI = 71%–100%) and a specificity of 99% (95% CI = 97%–100%). However, >17% of adults still needed further diagnostic workup to determine presence or absence of meningitis. A new, more complex screening tool was developed to minimize the number of patients who would require further testing ( Figure ). This screening tool had a sensitivity of 86% (95% CI = 71%–100%) and a specificity of 92% (95% CI = 85%–99%) but left only 2.5% of adults in need of further testing ( 41 ). FIGURE Screening tool to identify potential anthrax meningitis cases by presenting signs and symptoms after a mass casualty event when diagnostic capability is limited — CDC, 2023 Abbreviations: CSF = cerebrospinal fluid; CT = computed tomography. Figure illustrates the steps in a screening tool to identify potential anthrax meningitis cases by presenting signs and symptoms after a mass casualty event when diagnostic capability is limited. Source: Adapted from Figure 2 in Binney S, Person MK, Traxler RM, Cook R, Bower WA, Hendricks K. Algorithms for the identification of anthrax meningitis during a mass casualty event based on a systematic review of systemic anthrax from 1880 through 2018. Clin Infect Dis 2022;75(Suppl 3):S468–77. Antimicrobial Drug Selection In the Kyrgyz Republic case series, 37 patients were categorized by an algorithm as likely having meningitis and 89 were considered unlikely to have meningitis ( 52 ). Twenty-three (62%) of the 37 received monotherapy, and all 23 survived. The remainder (38%) received 2-drug combination therapy and survived. Only five of 14 survivors who received combination therapy received a bactericidal and a PSI antimicrobial drug, and none received three concurrent antimicrobial drugs. In the systematic review of human treatment data, survival was not improved by adding a PSI to a bactericidal antimicrobial drug in adults with anthrax meningitis ( 39 ). However, median length of stay for the three patients who received both a PSI and a bactericidal antimicrobial drug was 10 days (IQR = 8–21 days) compared with 30 days (IQR = 15–50 days) for the four who received bactericidal antimicrobial drugs alone, although this difference was not statistically significant (p = 0.26) ( 39 ). Population PK Modeling to Predict Drug Exposures in Cerebrospinal Fluid for Naturally Occurring B. anthracis Strains Antimicrobial drugs with published human cerebrospinal fluid (CSF) penetration data were categorized according to their ability to achieve robust drug exposures in both plasma and CSF based on population PK modeling followed by Monte Carlo simulations. Because PK/PD target values for B. anthracis are scarce, established targets from other pathogens were used to evaluate drug exposures. One study suggested the PK/PD targets for the carbapenem class of antibiotics (e.g., faropenem) were comparable between B. anthracis and gram-negative pathogens. The blood-brain barrier can be impaired considerably during anthrax meningitis; therefore, both the plasma and the CSF exposures were considered. Because of the need for between-pathogen extrapolations of PK/PD targets, absence of CSF-specific targets, and relative sparsity of the data, these Monte Carlo simulation results should be interpreted conservatively. The Monte Carlo simulations indicated the probability that patients with suspected or documented anthrax meningitis would achieve microbiologic cure of CSF when treated with the recommended dosage was high (≥95%) for ampicillin, ampicillin/sulbactam, ciprofloxacin, doxycycline, imipenem/cilastatin, levofloxacin 500 mg every 12 hours, meropenem, minocycline, and penicillin G; acceptable (≥90% to MIC = duration of the unbound (i.e. free) concentration above the minimum inhibitory concentration (expressed as a fraction of the dosing interval); IV = intravenous; MIC = minimum inhibitory concentration; MIC 90 = minimum inhibitory concentration required to inhibit the growth of 90% of organisms; PD = pharmacodynamic; PK = pharmacokinetic; S = bacteriostatic. * Highest MIC ( µ g/mL) with a >95% probability of successful CSF treatment. †PK/PD targets used were fAUC/MIC ≥18 for doxycycline and minocycline, fAUC/MIC ≥75 for quinolones, fT>MIC ≥40% for carbapenems, fT>MIC ≥50% for penicillins, fAUC/MIC ≥90 for amikacin, fAUC/MIC ≥27 for clindamycin, AUC/MIC ≥110 for linezolid, and AUC/MIC ≥400 for vancomycin. § High probability is defined as a ≥95% probability of treatment success. ¶ PK/PD target achievement and the highest treatable MICs were similar (i.e., within twofold) in CSF and plasma for ciprofloxacin, levofloxacin, doxycycline, and minocycline. ** For the studied ß-lactams (i.e., meropenem, imipenem, penicillin G, ampicillin, and piperacillin) and for amikacin, the highest treatable MICs based on the plasma targets were twofold to thirty-two-fold higher compared with those in CSF. ††Previous susceptibility testing to infecting strain is recommended. Bacillus anthracis naturally carries penicillin-resistance genes for a ß-lactamase that is expressed in up to 10% of naturally occurring strains, creating resistance to penicillin G, ampicillin, and piperacillin. §§ Acceptable probability is defined as a ≥90% to 90% for anthrax meningitis ( 29 , 52 ). On the basis of efficacy described from in vivo data and human clinical experience and known potential risks, the benefits of antimicrobial drugs for PEP or treatment of anthrax outweigh the known risks. These guidelines address anthrax PEP and treatment for both natural and intentional exposures (e.g., a wide-area aerosol release of B. anthracis spores). The evidence base linking recommendations to data is available (Supplementary Material, https://stacks.cdc.gov/view/cdc/132182 ). Previously, all B. anthracis strains from a naturally occurring source or an intentional release were thought to be susceptible to the recommended first-line antimicrobial drugs (except for penicillins). However, over the past few decades, studies have demonstrated that antimicrobial-resistant B. anthracis strains can be created with relative ease through serial passaging on selective media ( 121 , 122 ). Consequently, bioterrorists could mass produce a multidrug-resistant B. anthracis strain capable of evading previously recommended first-line antimicrobial medical countermeasures. These updated CDC guidelines provide PEP and treatment recommendations that include numerous antimicrobial drugs from multiple classes. The antimicrobial drugs recommended as first-line agents are expected to address most scenarios. The alternative antimicrobial drugs provide contingencies for contraindications, intolerances, unavailability, and natural or genetically engineered resistance. The recommended medical countermeasures are preferentially ordered based on 1) in vitro effectiveness against B. anthracis ( 34 ) ( Table 1 ); 2) in vivo efficacy against B. anthracis exposures as demonstrated by ORs and CIs for survival compared with no therapy or therapy with a positive control ( 35 , 37 ) ( Table 2 ); 3) the animal model used to generate efficacy data (nonhuman primate or rabbit models were preferred over mouse, guinea pig, or hamster models); 4) treatment outcomes for published human cases ( 39 ) ( Table 3 ); 5) the percentage of patients expected to achieve microbiologic CSF cure at recommended antimicrobial drug dosing based on Monte Carlo simulations ( Table 5 ); 6) the safety profiles of the antimicrobial drugs ( 42 ); 7) logistical considerations (e.g., available formulations [including availability and palatability of liquid formulations], dosing intervals, cost, and supply and availability patterns); and 8) expert opinion. In addition, certain antimicrobial drugs are included for PEPAbx or treatment on the basis of class efficacy (e.g., levofloxacin, moxifloxacin, and ofloxacin) or for treatment on the basis of demonstrated PEPAbx efficacy (e.g., minocycline). Early diagnosis of anthrax and initiation of appropriate treatment are critical to improving survival. Although empiric treatment of anthrax or prophylaxis after exposure is needed to save lives, antimicrobial drug susceptibility testing is vital; antimicrobial drug choices might need to be modified based on the results. Data indicate penicillin-class antimicrobial drugs are as effective as other bactericidal agents for PEPAbx and treatment and might be preferred in certain populations. However, although 65 years exposed to anthrax. In July 2023, a second-generation anthrax vaccine, anthrax vaccine adsorbed, adjuvanted, was FDA approved for PEPVx against inhalation anthrax. Anthrax vaccine adsorbed, adjuvanted is administered by the IM route as a 2-dose series 2 weeks apart, in conjunction with PEPAbx for adults aged 18–65 years. In persons aged >65 years, anthrax vaccine adsorbed, adjuvanted elicited a higher immune response compared with AVA ( 131 ). Anthrax vaccine use in older adults (aged >65 years), pregnant or lactating persons, and children (aged 18 years with cutaneous anthrax without signs and symptoms of meningitis, by descending order of preference — CDC recommendations, 2023 Treatment (listed drugs joined by "or" are considered equivalent) Dosage First-line antimicrobial drug Doxycycline ¶, ** 100 mg every 12 hours orally or Ciprofloxacin ¶ 500 mg every 12 hours orally or Levofloxacin ¶ 750 mg every 24 hours orally PCN-S only: Amoxicillin ††,§§ 1 g every 8 hours orally or Penicillin VK ††500 mg every 6 hours orally Alternative antimicrobial drug ¶¶ Amoxicillin/clavulanate ††16:1 formulation (1 g/62.5 mg) in 2 tablets every 12 hours orally or Amoxicillin/clavulanate ††7:1 formulation (875/125 mg) every 12 hours orally Moxifloxacin** ,††400 mg every 24 hours orally Ofloxacin ††400 mg every 12 hours orally Clindamycin ††600 mg every 8 hours orally Omadacycline ††450 mg every 24 hours orally x 2 days, then 300 mg every 24 hours orally Linezolid ††600 mg every 12 hours orally Clarithromycin ††, *** 500 mg every 12 hours orally (only initiate after at least 3 days of treatment with any of the other antimicrobial drugs listed) Dalbavancin ††1 g x 1 dose IV, then 500 mg weekly IV Imipenem/cilastatin ††2 g every 8 hours IV or Meropenem ††2 g every 8 hours IV Vancomycin ††15 mg/kg every 12 hours IV over a period of 1–2 hours (target AUC 24 of 400–600 µ g x h/mL [preferred]; if AUC 24 is not available, maintain serum trough concentrations of 15–20 µ g/mL) Antitoxin (only to be used if antimicrobial drugs are not available or not appropriate; listed antitoxins joined by "or" are considered equivalent) Raxibacumab †††40 mg/kg in a single dose IV or Obiltoxaximab †††16 mg/kg in a single dose IV AIGIV §§§ 420 units IV Abbreviations: AIGIV = anthrax immunoglobulin intravenous; AUC 24 = area under the concentration-time curve from 0 to 24 hours; FDA = Food and Drug Administration; IV = intravenous; PCN-S = penicillin-susceptible strains; PEPAbx = antimicrobial postexposure prophylaxis for anthrax. * Definitive therapy should be directed by antibiotic susceptibility test results, when available. †For pregnant adolescents, refer to pediatric guidelines for weight-based dosing (see Table 13 ). § Dosing recommended for pregnant persons regardless of trimester. ¶ Approved by FDA for anthrax PEPAbx, treatment, or both, but specific uses (e.g., doses, dosing schedules, and patient populations) recommended in this report might differ from the FDA-approved labeling. ** If liquid formulations are not available for adults who cannot swallow pills, instructions are available for preparing oral suspensions of moxifloxacin ( Source: Hutchinson DJ, Johnson CE, Klein KC. Stability of extemporaneously prepared moxifloxacin oral suspensions. Am J Health Syst Pharm 2009;66:665–7.121) and doxycycline ( Source: CDC. In an anthrax emergency: how to prepare doxycycline hyclate for children and adults who cannot swallow pills. Atlanta, GA: US Department of Health and Human Services, CDC; 2020. https://www.cdc.gov/anthrax/public-health/doxy-crushing-instruction-pamphlet.html ). ††Not approved by FDA for anthrax PEPAbx or treatment. §§ Ampicillin 500 mg every 6 hours orally can be used as an alternative to amoxicillin, if available. ¶¶ Alternative selections are for patients who have contraindications to or cannot tolerate first-line antimicrobial drugs or if first-line antimicrobial drugs are not available. *** Clarithromycin is unlikely to be effective if the patient has bacteremia, thus a different antimicrobial drug must be used initially to clear bacteremia. †††Premedicate with IV or oral diphenhydramine within 1 hour before administration. Hypersensitivity and anaphylaxis have been reported after raxibacumab and obiltoxaximab administration. §§§ An 840-unit dose of AIGIV can be considered for severe cases. TABLE 11 Empiric* treatment regimens for pregnant or lactating persons aged ≥18 years †,§ with systemic ¶ anthrax with or without meningitis, by descending order of preference — CDC recommendations, 2023 Regimen Example Regimen 1. Two bactericidal drugs from different antimicrobial drug classes plus a PSI or an RNAI Ciprofloxacin plus meropenem plus doxycycline or omadacycline** Regimen 2. One bactericidal drug plus a PSI Meropenem plus linezolid or doxycycline Regimen 3. One bactericidal drug plus a second bactericidal drug from a different antimicrobial drug class Meropenem plus ciprofloxacin Regimen 4. One bactericidal drug plus an RNAI (rifampin should not be used as monotherapy) Meropenem plus rifampin Regimen 5. A PSI plus an RNAI (rifampin should not be used as monotherapy) Linezolid or doxycycline or clindamycin plus rifampin Regimen 6. Two PSIs from different antimicrobial drug classes Linezolid plus doxycycline Regimen 7. A single bactericidal drug Meropenem Regimen 8. A single PSI Linezolid or doxycycline or clindamycin First-line antimicrobial drug ††Bactericidal drug PSI Treatment (listed drugs joined by "or" are considered equivalent) Dosage Treatment Dosage Meropenem §§ 2 g every 8 hours IV Doxycycline ¶¶ 200 mg x 1 dose IV, then 100 mg every 12 hours IV or Ciprofloxacin ¶¶ 400 mg every 8 hours IV or Levofloxacin ¶¶ 500 mg every 12 hours IV PCN-S only: Penicillin G ¶¶ 4 million units every 4 hours IV or Ampicillin §§ 2 g every 4 hours IV Imipenem/cilastatin §§ 1 g every 6 hours IV or Ampicillin/sulbactam §§ 3 g every 6 hours IV Alternative antimicrobial drug*** Bactericidal drug PSI/RNAI Treatment Dosage Treatment Dosage Piperacillin/tazobactam §§ 3.375 g every 4 hours IV Omadacycline §§,†††200 mg x 1 dose IV on day 1, then 100 mg every 24 hours IV Moxifloxacin §§ 400 mg every 24 hours IV Eravacycline §§,†††1 mg/kg every 12 hours IV Vancomycin §§,†††15 mg/kg every 12 hours IV over a period of 1–2 hours (target AUC 24 of 400–600 µ g x h/mL [preferred]; if AUC 24 is not available, maintain serum trough concentrations of 15–20 µ g/mL). Consider a loading dose of 20-35 mg/kg for critically ill patients. Clindamycin §§ 900 mg every 8 hours IV Linezolid §§ 600 mg every 12 hours IV Rifampin §§,§§§ 600 mg every 12 hours IV plus Antitoxin (single dose as an adjunct to antimicrobial drug; listed antitoxins joined by "or" are considered equivalent) Treatment Dosage Raxibacumab ¶¶¶ 40 mg/kg IV or Obiltoxaximab ¶¶¶ 16 mg/kg IV AIGIV 420 units IV Abbreviations: AIGIV = anthrax immunoglobulin intravenous; AUC 24 = area under the concentration-time curve from 0 to 24 hours; FDA = Food and Drug Administration; IV = intravenous; PCN-S = penicillin-susceptible strains; PEPAbx = antimicrobial postexposure prophylaxis for anthrax; PSI = protein synthesis inhibitor; RNAI = RNA synthesis inhibitor. * Definitive therapy should be directed by antibiotic susceptibility test results, when available. †Pregnant adolescents: refer to pediatric guidelines for weight-based dosing (see Table 14 ). § Dosing recommended for pregnant persons regardless of trimester. If meningitis is not suspected and susceptibilities are known, start at regimen 2. ¶ "Systemic" was defined as including evidence of organ damage or any of the following: hyperthermia or hypothermia, tachycardia, tachypnea, hypotension, or leukocytosis or leukopenia ( Source: Katharios-Lanwermeyer S, Holty JE, Person M, et al. Identifying meningitis during an anthrax mass casualty incident: systematic review of systemic anthrax since 1880. Clin Infect Dis 2016;62:1537–45). ** For anthrax meningitis, consider using antimicrobial drugs that have demonstrated potential neuroprotective benefits in vivo (e.g., minocycline, doxycycline, clindamycin, and ß-lactamase inhibitors). ††For highly bioavailable antimicrobial drugs (e.g., ciprofloxacin, doxycycline, and linezolid), if the IV formulation is not available, oral formulations can be considered for patients with an intact gastrointestinal tract where absorption is expected to be complete after oral administration. §§ Not approved by FDA for anthrax PEPAbx or treatment. ¶¶ Approved by FDA for anthrax PEPAbx, treatment, or both, but specific uses (e.g., doses, dosing schedules, and patient populations) recommended in this report might differ from the FDA-approved labeling. *** Alternative selections are for patients who have contraindications to or cannot tolerate first-line antimicrobial drugs or if first-line antimicrobial drugs are not available. †††This antimicrobial does not cross an intact blood-brain barrier but can cross with meningitis because of breakdown of the barrier. §§§ Rifampin is an RNAI and also bactericidal; however, it should not be used as monotherapy. ¶¶¶ Premedicate with IV or oral diphenhydramine within 1 hour before administration. Hypersensitivity and anaphylaxis have been reported after raxibacumab and obiltoxaximab administration. **** An 840-unit dose of AIGIV can be considered for severe cases. PEP and Treatment Regimens for Cutaneous Anthrax Without Signs and Symptoms of Meningitis For pregnant and lactating persons aged ≥18 years, empiric PEP ( Table 9 ) and empiric cutaneous anthrax treatment ( Table 10 ) regimens include either a single antimicrobial drug or a single antitoxin. These regimens are summarized as follows: Antimicrobial drug: Choose a single antimicrobial drug. º Antimicrobial drugs are listed in descending order of preference in the table. Listed drugs joined by "or" are considered equivalent. ° Continue or switch antimicrobial drug based on susceptibility testing once available. ° Only choose a "PCN-S only" antimicrobial drug after the strain has been determined to be penicillin susceptible. Antitoxin: Choose a single antitoxin if no antimicrobial drugs are available. For all pregnant and lactating persons who have an aerosol exposure (e.g., a bioterrorism-related incident or animal skin drum–related event), PEPAbx should continue for 60 days from the exposure whether or not vaccine is given ( 33 ). For nonaerosol (i.e., cutaneous or ingestion) exposures, PEPAbx should continue for 7 days and vaccine is not needed. For pregnant and lactating persons with cutaneous anthrax without signs and symptoms of meningitis, the treatment regimen should continue for 7–10 days, or until clinical criteria for stability are met. If an aerosol exposure might have occurred, patients should transition from a treatment to a PEP regimen ( Table 9 ); the combined regimen should total 60 days from exposure. If no aerosolizing event occurred, patients with cutaneous anthrax do not need to continue PEPAbx. Treatment Regimens for Systemic Anthrax With or Without Meningitis For pregnant or lactating persons aged ≥18 years, empiric treatment regimens for those with systemic anthrax with or without meningitis ( Table 11 ) summarized as follows: Antimicrobial drugs: Choose two bactericidal drugs from different antimicrobial drug classes plus a PSI or an RNAI. º Antimicrobial drugs are listed in descending order of preference in the table. Listed drugs joined by "or" are considered equivalent. º Continue or switch antimicrobial drugs based on susceptibility testing once available. º Only choose a "PCN-S only" antimicrobial drug after the strain has been determined to be penicillin susceptible. Antitoxin: Choose a single antitoxin as adjunctive therapy. If an appropriate combination of bactericidal drugs plus a PSI or an RNAI is contraindicated, not well tolerated, or not available or if meningitis is considered unlikely, consider the following regimens in descending order of preference: One bactericidal drug plus a PSI (start with this regimen if meningitis is not suspected and susceptibilities are known) One bactericidal drug plus a second bactericidal drug from a different antimicrobial drug class One bactericidal drug plus an RNAI A PSI plus an RNAI Two PSIs from different antimicrobial drug classes A single bactericidal drug A single PSI From a PK/PD perspective, minocycline and doxycycline are the preferred PSIs because they provide more robust drug exposures in plasma and CSF compared with macrolides or clindamycin. A single RNAI (i.e., rifampin) should not be used as monotherapy because of the potential for rapid development of resistance ( 135 ). In addition, when meningitis is not suspected, certain oral formulations are included as alternatives in case IV formulations are not available. Duration of antimicrobial drug treatment should be for 2 weeks or longer; however, duration can be shortened and IV administration transitioned to oral medication based on patient improvement and clinical judgment. Patients with naturally acquired noninhalation cases do not need continuation of antimicrobial drugs for PEP. If an aerosol exposure might have occurred (i.e., a bioterrorism-related incident or animal skin drum–related event), healthy patients treated for systemic disease need no further antimicrobial drugs for PEP because they will have developed natural immunity. However, patients with compromised immune systems should transition to an oral PEP regimen ( Table 9 ). The total duration of antimicrobial drugs (i.e., treatment plus PEP) should be 60 days from onset of illness. Children Aged ≥1 Month to 12 to 12 to 8 years: 450 mg every 24 hours orally x 2 days, then 300 mg every 24 hours orally Clarithromycin** , **** 75 mg/kg every 12 hours orally (maximum 500 mg/dose; only initiate after at least 3 days of treatment with any of the other antimicrobial drugs listed) Dalbavancin** Aged ≥3 months to 10 kg to 40 kg: 60 mg/kg as a single dose IV >40 kg: 40 mg/kg as a single dose IV or Obiltoxaximab ††††≤15 kg: 32 mg/kg as a single dose IV >15 kg to 40 kg: 24 mg/kg as a single dose IV >40 kg: 16 mg/kg as a single dose IV Abbreviations: FDA = Food and Drug Administration; PCN-S = penicillin-susceptible strains; IV = intravenous; PEP = postexposure prophylaxis; PEPAbx = antimicrobial postexposure prophylaxis for anthrax. * Definitive therapy should be directed by antibiotic susceptibility test results, when available. †Approved by FDA for anthrax PEPAbx, treatment, or both, but specific uses (e.g., doses, dosing schedules, and patient populations) recommended in this report might differ from the FDA-approved labeling. § If liquid formulations are not available for children who cannot swallow pills, instructions are available for preparing oral suspensions of moxifloxacin ( Source: Hutchinson DJ, Johnson CE, Klein KC. Stability of extemporaneously prepared moxifloxacin oral suspensions. Am J Health Syst Pharm 2009;66:665–7.121) and doxycycline ( Source: CDC. In an anthrax emergency: how to prepare doxycycline hyclate for children and adults who cannot swallow pills. Atlanta, GA: US Department of Health and Human Services, CDC; 2020. https://www.cdc.gov/anthrax/public-health/doxy-crushing-instruction-pamphlet.html ). ¶ Up to 10% of naturally occurring strains of anthrax are resistant to penicillin and amoxicillin; bioterror strains can be engineered to generate resistance to multiple antibiotics. Susceptibility results reported from CDC within 48–72 hours of initial isolation of anthrax. ** Not approved by FDA for anthrax PEPAbx or treatment. ††Ampicillin 25 mg/kg every 6 hours orally (maximum 500 mg/dose) can be used as an alternative to amoxicillin, if available. §§ To minimize potential side effects from clavulanate, the amoxicillin/clavulanate 875/125 mg every 12 hours orally and amoxicillin/clavulanate extended-release 2,000/125 mg every 12 hours orally treatments are preferred over 500/125 mg every 8 hours orally. ¶¶ Alternative selections are for patients who have contraindications to or cannot tolerate first-line antimicrobial drugs or if first-line antimicrobial drugs are not available. *** Not approved by FDA for any pediatric indications. Not approved by FDA for anthrax PEPAbx or treatment. †††For children aged 12–17 years who weigh ≥45 kg with risk factors for cardiac events, consider 200 mg twice daily. §§§ The use of tetracycline-class antibiotics during tooth development (i.e., last half of pregnancy, infancy, and childhood up to age 8 years) might cause permanent discoloration of the teeth (yellow, gray, or brown) and enamel hypoplasia. This adverse effect appears to occur less often with doxycycline but might occur more with longer durations of therapy. ¶¶¶ Linezolid can be considered for PEP in scenarios when patients can receive regular laboratory testing to monitor for myelosuppression or neurotoxicity, which might occur within 14–28 days of use. If possible, switch to a different drug when available. **** Clarithromycin is unlikely to be effective if the patient has bacteremia, thus a different antimicrobial drug must be used initially to clear bacteremia. ††††Premedicate with IV or oral diphenhydramine within 1 hour before administration. Hypersensitivity and anaphylaxis have been reported after raxibacumab and obiltoxaximab administration. TABLE 13 Empiric* treatment regimens for children aged ≥1 month to 12 to 12 to 8 years: 450 mg every 24 hours orally x 2 days, then 300 mg every 24 hours orally Clarithromycin** ,§§§ 7.5 mg/kg every 12 hours orally (maximum 500 mg/dose; only initiate after at least 3 days of treatment with any of the other antimicrobial drugs listed) Dalbavancin** Aged ≥3 months to 10–40 kg: 60 mg/kg as a single dose IV >40 kg: 40 mg/kg as a single dose IV or Obiltoxaximab ¶¶¶ ≤15 kg: 32 mg/kg as a single dose IV >15–40 kg: 24 mg/kg as a single dose IV >40 kg: 16 mg/kg as a single dose IV AIGIV**** >10 kg: 1 vial (approximately 60 units) IV 10 to 5 kg. TABLE 14 Empiric* treatment regimens for children aged ≥1 month to 8 years: 1 mg/kg every 12 hours IV or Linezolid ††Aged 10 to 40 kg: 60 mg/kg as a single dose IV >40 kg: 40 mg/kg as a single dose IV or Obiltoxaximab §§§§ ≤15 kg: 32 mg/kg as a single dose IV >15 to 40 kg: 24 mg/kg as a single dose IV >40 kg: 16 mg/kg as a single dose IV AIGIV ¶¶¶¶ >10 kg: 1 vial (approximately 60 units) IV 10 to 5 kg. PEP and Treatment Regimens for Cutaneous Anthrax Without Signs and Symptoms of Meningitis For children aged ≥1 month to 1 month to 65 years exposed to anthrax. In July 2023, a second-generation anthrax vaccine, anthrax vaccine adsorbed, adjuvanted, was FDA approved for PEPVx against inhalation anthrax. Anthrax vaccine adsorbed, adjuvanted is administered by the IM route as a 2-dose series 2 weeks apart, in conjunction with PEPAbx for adults aged 18–65 years. In persons aged >65 years, anthrax vaccine adsorbed, adjuvanted elicited a higher immune response compared with AVA ( 131 ). Anthrax vaccine use in older adults (aged >65 years), pregnant or lactating persons, and children (aged 65 years exposed to anthrax. In July 2023, a second-generation anthrax vaccine, anthrax vaccine adsorbed, adjuvanted, was FDA approved for PEPVx against inhalation anthrax. Anthrax vaccine adsorbed, adjuvanted is administered by the IM route as a 2-dose series 2 weeks apart, in conjunction with PEPAbx for adults aged 18–65 years. In persons aged >65 years, anthrax vaccine adsorbed, adjuvanted elicited a higher immune response compared with AVA ( 131 ). Anthrax vaccine use in older adults (aged >65 years), pregnant or lactating persons, and children (aged 18 years with cutaneous anthrax without signs and symptoms of meningitis, by descending order of preference — CDC recommendations, 2023 Treatment (listed drugs joined by "or" are considered equivalent) Dosage First-line antimicrobial drug Doxycycline ¶, ** 100 mg every 12 hours orally or Ciprofloxacin ¶ 500 mg every 12 hours orally or Levofloxacin ¶ 750 mg every 24 hours orally PCN-S only: Amoxicillin ††,§§ 1 g every 8 hours orally or Penicillin VK ††500 mg every 6 hours orally Alternative antimicrobial drug ¶¶ Amoxicillin/clavulanate ††16:1 formulation (1 g/62.5 mg) in 2 tablets every 12 hours orally or Amoxicillin/clavulanate ††7:1 formulation (875/125 mg) every 12 hours orally Moxifloxacin** ,††400 mg every 24 hours orally Ofloxacin ††400 mg every 12 hours orally Clindamycin ††600 mg every 8 hours orally Omadacycline ††450 mg every 24 hours orally x 2 days, then 300 mg every 24 hours orally Linezolid ††600 mg every 12 hours orally Clarithromycin ††, *** 500 mg every 12 hours orally (only initiate after at least 3 days of treatment with any of the other antimicrobial drugs listed) Dalbavancin ††1 g x 1 dose IV, then 500 mg weekly IV Imipenem/cilastatin ††2 g every 8 hours IV or Meropenem ††2 g every 8 hours IV Vancomycin ††15 mg/kg every 12 hours IV over a period of 1–2 hours (target AUC 24 of 400–600 µ g x h/mL [preferred]; if AUC 24 is not available, maintain serum trough concentrations of 15–20 µ g/mL) Antitoxin (only to be used if antimicrobial drugs are not available or not appropriate; listed antitoxins joined by "or" are considered equivalent) Raxibacumab †††40 mg/kg in a single dose IV or Obiltoxaximab †††16 mg/kg in a single dose IV AIGIV §§§ 420 units IV Abbreviations: AIGIV = anthrax immunoglobulin intravenous; AUC 24 = area under the concentration-time curve from 0 to 24 hours; FDA = Food and Drug Administration; IV = intravenous; PCN-S = penicillin-susceptible strains; PEPAbx = antimicrobial postexposure prophylaxis for anthrax. * Definitive therapy should be directed by antibiotic susceptibility test results, when available. †For pregnant adolescents, refer to pediatric guidelines for weight-based dosing (see Table 13 ). § Dosing recommended for pregnant persons regardless of trimester. ¶ Approved by FDA for anthrax PEPAbx, treatment, or both, but specific uses (e.g., doses, dosing schedules, and patient populations) recommended in this report might differ from the FDA-approved labeling. ** If liquid formulations are not available for adults who cannot swallow pills, instructions are available for preparing oral suspensions of moxifloxacin ( Source: Hutchinson DJ, Johnson CE, Klein KC. Stability of extemporaneously prepared moxifloxacin oral suspensions. Am J Health Syst Pharm 2009;66:665–7.121) and doxycycline ( Source: CDC. In an anthrax emergency: how to prepare doxycycline hyclate for children and adults who cannot swallow pills. Atlanta, GA: US Department of Health and Human Services, CDC; 2020. https://www.cdc.gov/anthrax/public-health/doxy-crushing-instruction-pamphlet.html ). ††Not approved by FDA for anthrax PEPAbx or treatment. §§ Ampicillin 500 mg every 6 hours orally can be used as an alternative to amoxicillin, if available. ¶¶ Alternative selections are for patients who have contraindications to or cannot tolerate first-line antimicrobial drugs or if first-line antimicrobial drugs are not available. *** Clarithromycin is unlikely to be effective if the patient has bacteremia, thus a different antimicrobial drug must be used initially to clear bacteremia. †††Premedicate with IV or oral diphenhydramine within 1 hour before administration. Hypersensitivity and anaphylaxis have been reported after raxibacumab and obiltoxaximab administration. §§§ An 840-unit dose of AIGIV can be considered for severe cases. TABLE 11 Empiric* treatment regimens for pregnant or lactating persons aged ≥18 years †,§ with systemic ¶ anthrax with or without meningitis, by descending order of preference — CDC recommendations, 2023 Regimen Example Regimen 1. Two bactericidal drugs from different antimicrobial drug classes plus a PSI or an RNAI Ciprofloxacin plus meropenem plus doxycycline or omadacycline** Regimen 2. One bactericidal drug plus a PSI Meropenem plus linezolid or doxycycline Regimen 3. One bactericidal drug plus a second bactericidal drug from a different antimicrobial drug class Meropenem plus ciprofloxacin Regimen 4. One bactericidal drug plus an RNAI (rifampin should not be used as monotherapy) Meropenem plus rifampin Regimen 5. A PSI plus an RNAI (rifampin should not be used as monotherapy) Linezolid or doxycycline or clindamycin plus rifampin Regimen 6. Two PSIs from different antimicrobial drug classes Linezolid plus doxycycline Regimen 7. A single bactericidal drug Meropenem Regimen 8. A single PSI Linezolid or doxycycline or clindamycin First-line antimicrobial drug ††Bactericidal drug PSI Treatment (listed drugs joined by "or" are considered equivalent) Dosage Treatment Dosage Meropenem §§ 2 g every 8 hours IV Doxycycline ¶¶ 200 mg x 1 dose IV, then 100 mg every 12 hours IV or Ciprofloxacin ¶¶ 400 mg every 8 hours IV or Levofloxacin ¶¶ 500 mg every 12 hours IV PCN-S only: Penicillin G ¶¶ 4 million units every 4 hours IV or Ampicillin §§ 2 g every 4 hours IV Imipenem/cilastatin §§ 1 g every 6 hours IV or Ampicillin/sulbactam §§ 3 g every 6 hours IV Alternative antimicrobial drug*** Bactericidal drug PSI/RNAI Treatment Dosage Treatment Dosage Piperacillin/tazobactam §§ 3.375 g every 4 hours IV Omadacycline §§,†††200 mg x 1 dose IV on day 1, then 100 mg every 24 hours IV Moxifloxacin §§ 400 mg every 24 hours IV Eravacycline §§,†††1 mg/kg every 12 hours IV Vancomycin §§,†††15 mg/kg every 12 hours IV over a period of 1–2 hours (target AUC 24 of 400–600 µ g x h/mL [preferred]; if AUC 24 is not available, maintain serum trough concentrations of 15–20 µ g/mL). Consider a loading dose of 20-35 mg/kg for critically ill patients. Clindamycin §§ 900 mg every 8 hours IV Linezolid §§ 600 mg every 12 hours IV Rifampin §§,§§§ 600 mg every 12 hours IV plus Antitoxin (single dose as an adjunct to antimicrobial drug; listed antitoxins joined by "or" are considered equivalent) Treatment Dosage Raxibacumab ¶¶¶ 40 mg/kg IV or Obiltoxaximab ¶¶¶ 16 mg/kg IV AIGIV 420 units IV Abbreviations: AIGIV = anthrax immunoglobulin intravenous; AUC 24 = area under the concentration-time curve from 0 to 24 hours; FDA = Food and Drug Administration; IV = intravenous; PCN-S = penicillin-susceptible strains; PEPAbx = antimicrobial postexposure prophylaxis for anthrax; PSI = protein synthesis inhibitor; RNAI = RNA synthesis inhibitor. * Definitive therapy should be directed by antibiotic susceptibility test results, when available. †Pregnant adolescents: refer to pediatric guidelines for weight-based dosing (see Table 14 ). § Dosing recommended for pregnant persons regardless of trimester. If meningitis is not suspected and susceptibilities are known, start at regimen 2. ¶ "Systemic" was defined as including evidence of organ damage or any of the following: hyperthermia or hypothermia, tachycardia, tachypnea, hypotension, or leukocytosis or leukopenia ( Source: Katharios-Lanwermeyer S, Holty JE, Person M, et al. Identifying meningitis during an anthrax mass casualty incident: systematic review of systemic anthrax since 1880. Clin Infect Dis 2016;62:1537–45). ** For anthrax meningitis, consider using antimicrobial drugs that have demonstrated potential neuroprotective benefits in vivo (e.g., minocycline, doxycycline, clindamycin, and ß-lactamase inhibitors). ††For highly bioavailable antimicrobial drugs (e.g., ciprofloxacin, doxycycline, and linezolid), if the IV formulation is not available, oral formulations can be considered for patients with an intact gastrointestinal tract where absorption is expected to be complete after oral administration. §§ Not approved by FDA for anthrax PEPAbx or treatment. ¶¶ Approved by FDA for anthrax PEPAbx, treatment, or both, but specific uses (e.g., doses, dosing schedules, and patient populations) recommended in this report might differ from the FDA-approved labeling. *** Alternative selections are for patients who have contraindications to or cannot tolerate first-line antimicrobial drugs or if first-line antimicrobial drugs are not available. †††This antimicrobial does not cross an intact blood-brain barrier but can cross with meningitis because of breakdown of the barrier. §§§ Rifampin is an RNAI and also bactericidal; however, it should not be used as monotherapy. ¶¶¶ Premedicate with IV or oral diphenhydramine within 1 hour before administration. Hypersensitivity and anaphylaxis have been reported after raxibacumab and obiltoxaximab administration. **** An 840-unit dose of AIGIV can be considered for severe cases. PEP and Treatment Regimens for Cutaneous Anthrax Without Signs and Symptoms of Meningitis For pregnant and lactating persons aged ≥18 years, empiric PEP ( Table 9 ) and empiric cutaneous anthrax treatment ( Table 10 ) regimens include either a single antimicrobial drug or a single antitoxin. These regimens are summarized as follows: Antimicrobial drug: Choose a single antimicrobial drug. º Antimicrobial drugs are listed in descending order of preference in the table. Listed drugs joined by "or" are considered equivalent. ° Continue or switch antimicrobial drug based on susceptibility testing once available. ° Only choose a "PCN-S only" antimicrobial drug after the strain has been determined to be penicillin susceptible. Antitoxin: Choose a single antitoxin if no antimicrobial drugs are available. For all pregnant and lactating persons who have an aerosol exposure (e.g., a bioterrorism-related incident or animal skin drum–related event), PEPAbx should continue for 60 days from the exposure whether or not vaccine is given ( 33 ). For nonaerosol (i.e., cutaneous or ingestion) exposures, PEPAbx should continue for 7 days and vaccine is not needed. For pregnant and lactating persons with cutaneous anthrax without signs and symptoms of meningitis, the treatment regimen should continue for 7–10 days, or until clinical criteria for stability are met. If an aerosol exposure might have occurred, patients should transition from a treatment to a PEP regimen ( Table 9 ); the combined regimen should total 60 days from exposure. If no aerosolizing event occurred, patients with cutaneous anthrax do not need to continue PEPAbx. Treatment Regimens for Systemic Anthrax With or Without Meningitis For pregnant or lactating persons aged ≥18 years, empiric treatment regimens for those with systemic anthrax with or without meningitis ( Table 11 ) summarized as follows: Antimicrobial drugs: Choose two bactericidal drugs from different antimicrobial drug classes plus a PSI or an RNAI. º Antimicrobial drugs are listed in descending order of preference in the table. Listed drugs joined by "or" are considered equivalent. º Continue or switch antimicrobial drugs based on susceptibility testing once available. º Only choose a "PCN-S only" antimicrobial drug after the strain has been determined to be penicillin susceptible. Antitoxin: Choose a single antitoxin as adjunctive therapy. If an appropriate combination of bactericidal drugs plus a PSI or an RNAI is contraindicated, not well tolerated, or not available or if meningitis is considered unlikely, consider the following regimens in descending order of preference: One bactericidal drug plus a PSI (start with this regimen if meningitis is not suspected and susceptibilities are known) One bactericidal drug plus a second bactericidal drug from a different antimicrobial drug class One bactericidal drug plus an RNAI A PSI plus an RNAI Two PSIs from different antimicrobial drug classes A single bactericidal drug A single PSI From a PK/PD perspective, minocycline and doxycycline are the preferred PSIs because they provide more robust drug exposures in plasma and CSF compared with macrolides or clindamycin. A single RNAI (i.e., rifampin) should not be used as monotherapy because of the potential for rapid development of resistance ( 135 ). In addition, when meningitis is not suspected, certain oral formulations are included as alternatives in case IV formulations are not available. Duration of antimicrobial drug treatment should be for 2 weeks or longer; however, duration can be shortened and IV administration transitioned to oral medication based on patient improvement and clinical judgment. Patients with naturally acquired noninhalation cases do not need continuation of antimicrobial drugs for PEP. If an aerosol exposure might have occurred (i.e., a bioterrorism-related incident or animal skin drum–related event), healthy patients treated for systemic disease need no further antimicrobial drugs for PEP because they will have developed natural immunity. However, patients with compromised immune systems should transition to an oral PEP regimen ( Table 9 ). The total duration of antimicrobial drugs (i.e., treatment plus PEP) should be 60 days from onset of illness. Children Aged ≥1 Month to 12 to 12 to 8 years: 450 mg every 24 hours orally x 2 days, then 300 mg every 24 hours orally Clarithromycin** , **** 75 mg/kg every 12 hours orally (maximum 500 mg/dose; only initiate after at least 3 days of treatment with any of the other antimicrobial drugs listed) Dalbavancin** Aged ≥3 months to 10 kg to 40 kg: 60 mg/kg as a single dose IV >40 kg: 40 mg/kg as a single dose IV or Obiltoxaximab ††††≤15 kg: 32 mg/kg as a single dose IV >15 kg to 40 kg: 24 mg/kg as a single dose IV >40 kg: 16 mg/kg as a single dose IV Abbreviations: FDA = Food and Drug Administration; PCN-S = penicillin-susceptible strains; IV = intravenous; PEP = postexposure prophylaxis; PEPAbx = antimicrobial postexposure prophylaxis for anthrax. * Definitive therapy should be directed by antibiotic susceptibility test results, when available. †Approved by FDA for anthrax PEPAbx, treatment, or both, but specific uses (e.g., doses, dosing schedules, and patient populations) recommended in this report might differ from the FDA-approved labeling. § If liquid formulations are not available for children who cannot swallow pills, instructions are available for preparing oral suspensions of moxifloxacin ( Source: Hutchinson DJ, Johnson CE, Klein KC. Stability of extemporaneously prepared moxifloxacin oral suspensions. Am J Health Syst Pharm 2009;66:665–7.121) and doxycycline ( Source: CDC. In an anthrax emergency: how to prepare doxycycline hyclate for children and adults who cannot swallow pills. Atlanta, GA: US Department of Health and Human Services, CDC; 2020. https://www.cdc.gov/anthrax/public-health/doxy-crushing-instruction-pamphlet.html ). ¶ Up to 10% of naturally occurring strains of anthrax are resistant to penicillin and amoxicillin; bioterror strains can be engineered to generate resistance to multiple antibiotics. Susceptibility results reported from CDC within 48–72 hours of initial isolation of anthrax. ** Not approved by FDA for anthrax PEPAbx or treatment. ††Ampicillin 25 mg/kg every 6 hours orally (maximum 500 mg/dose) can be used as an alternative to amoxicillin, if available. §§ To minimize potential side effects from clavulanate, the amoxicillin/clavulanate 875/125 mg every 12 hours orally and amoxicillin/clavulanate extended-release 2,000/125 mg every 12 hours orally treatments are preferred over 500/125 mg every 8 hours orally. ¶¶ Alternative selections are for patients who have contraindications to or cannot tolerate first-line antimicrobial drugs or if first-line antimicrobial drugs are not available. *** Not approved by FDA for any pediatric indications. Not approved by FDA for anthrax PEPAbx or treatment. †††For children aged 12–17 years who weigh ≥45 kg with risk factors for cardiac events, consider 200 mg twice daily. §§§ The use of tetracycline-class antibiotics during tooth development (i.e., last half of pregnancy, infancy, and childhood up to age 8 years) might cause permanent discoloration of the teeth (yellow, gray, or brown) and enamel hypoplasia. This adverse effect appears to occur less often with doxycycline but might occur more with longer durations of therapy. ¶¶¶ Linezolid can be considered for PEP in scenarios when patients can receive regular laboratory testing to monitor for myelosuppression or neurotoxicity, which might occur within 14–28 days of use. If possible, switch to a different drug when available. **** Clarithromycin is unlikely to be effective if the patient has bacteremia, thus a different antimicrobial drug must be used initially to clear bacteremia. ††††Premedicate with IV or oral diphenhydramine within 1 hour before administration. Hypersensitivity and anaphylaxis have been reported after raxibacumab and obiltoxaximab administration. TABLE 13 Empiric* treatment regimens for children aged ≥1 month to 12 to 12 to 8 years: 450 mg every 24 hours orally x 2 days, then 300 mg every 24 hours orally Clarithromycin** ,§§§ 7.5 mg/kg every 12 hours orally (maximum 500 mg/dose; only initiate after at least 3 days of treatment with any of the other antimicrobial drugs listed) Dalbavancin** Aged ≥3 months to 10–40 kg: 60 mg/kg as a single dose IV >40 kg: 40 mg/kg as a single dose IV or Obiltoxaximab ¶¶¶ ≤15 kg: 32 mg/kg as a single dose IV >15–40 kg: 24 mg/kg as a single dose IV >40 kg: 16 mg/kg as a single dose IV AIGIV**** >10 kg: 1 vial (approximately 60 units) IV 10 to 5 kg. TABLE 14 Empiric* treatment regimens for children aged ≥1 month to 8 years: 1 mg/kg every 12 hours IV or Linezolid ††Aged 10 to 40 kg: 60 mg/kg as a single dose IV >40 kg: 40 mg/kg as a single dose IV or Obiltoxaximab §§§§ ≤15 kg: 32 mg/kg as a single dose IV >15 to 40 kg: 24 mg/kg as a single dose IV >40 kg: 16 mg/kg as a single dose IV AIGIV ¶¶¶¶ >10 kg: 1 vial (approximately 60 units) IV 10 to 5 kg. PEP and Treatment Regimens for Cutaneous Anthrax Without Signs and Symptoms of Meningitis For children aged ≥1 month to 1 month to 18 years with cutaneous anthrax without signs and symptoms of meningitis, by descending order of preference — CDC recommendations, 2023 Treatment (listed drugs joined by "or" are considered equivalent) Dosage First-line antimicrobial drug Doxycycline ¶, ** 100 mg every 12 hours orally or Ciprofloxacin ¶ 500 mg every 12 hours orally or Levofloxacin ¶ 750 mg every 24 hours orally PCN-S only: Amoxicillin ††,§§ 1 g every 8 hours orally or Penicillin VK ††500 mg every 6 hours orally Alternative antimicrobial drug ¶¶ Amoxicillin/clavulanate ††16:1 formulation (1 g/62.5 mg) in 2 tablets every 12 hours orally or Amoxicillin/clavulanate ††7:1 formulation (875/125 mg) every 12 hours orally Moxifloxacin** ,††400 mg every 24 hours orally Ofloxacin ††400 mg every 12 hours orally Clindamycin ††600 mg every 8 hours orally Omadacycline ††450 mg every 24 hours orally x 2 days, then 300 mg every 24 hours orally Linezolid ††600 mg every 12 hours orally Clarithromycin ††, *** 500 mg every 12 hours orally (only initiate after at least 3 days of treatment with any of the other antimicrobial drugs listed) Dalbavancin ††1 g x 1 dose IV, then 500 mg weekly IV Imipenem/cilastatin ††2 g every 8 hours IV or Meropenem ††2 g every 8 hours IV Vancomycin ††15 mg/kg every 12 hours IV over a period of 1–2 hours (target AUC 24 of 400–600 µ g x h/mL [preferred]; if AUC 24 is not available, maintain serum trough concentrations of 15–20 µ g/mL) Antitoxin (only to be used if antimicrobial drugs are not available or not appropriate; listed antitoxins joined by "or" are considered equivalent) Raxibacumab †††40 mg/kg in a single dose IV or Obiltoxaximab †††16 mg/kg in a single dose IV AIGIV §§§ 420 units IV Abbreviations: AIGIV = anthrax immunoglobulin intravenous; AUC 24 = area under the concentration-time curve from 0 to 24 hours; FDA = Food and Drug Administration; IV = intravenous; PCN-S = penicillin-susceptible strains; PEPAbx = antimicrobial postexposure prophylaxis for anthrax. * Definitive therapy should be directed by antibiotic susceptibility test results, when available. †For pregnant adolescents, refer to pediatric guidelines for weight-based dosing (see Table 13 ). § Dosing recommended for pregnant persons regardless of trimester. ¶ Approved by FDA for anthrax PEPAbx, treatment, or both, but specific uses (e.g., doses, dosing schedules, and patient populations) recommended in this report might differ from the FDA-approved labeling. ** If liquid formulations are not available for adults who cannot swallow pills, instructions are available for preparing oral suspensions of moxifloxacin ( Source: Hutchinson DJ, Johnson CE, Klein KC. Stability of extemporaneously prepared moxifloxacin oral suspensions. Am J Health Syst Pharm 2009;66:665–7.121) and doxycycline ( Source: CDC. In an anthrax emergency: how to prepare doxycycline hyclate for children and adults who cannot swallow pills. Atlanta, GA: US Department of Health and Human Services, CDC; 2020. https://www.cdc.gov/anthrax/public-health/doxy-crushing-instruction-pamphlet.html ). ††Not approved by FDA for anthrax PEPAbx or treatment. §§ Ampicillin 500 mg every 6 hours orally can be used as an alternative to amoxicillin, if available. ¶¶ Alternative selections are for patients who have contraindications to or cannot tolerate first-line antimicrobial drugs or if first-line antimicrobial drugs are not available. *** Clarithromycin is unlikely to be effective if the patient has bacteremia, thus a different antimicrobial drug must be used initially to clear bacteremia. †††Premedicate with IV or oral diphenhydramine within 1 hour before administration. Hypersensitivity and anaphylaxis have been reported after raxibacumab and obiltoxaximab administration. §§§ An 840-unit dose of AIGIV can be considered for severe cases. TABLE 11 Empiric* treatment regimens for pregnant or lactating persons aged ≥18 years †,§ with systemic ¶ anthrax with or without meningitis, by descending order of preference — CDC recommendations, 2023 Regimen Example Regimen 1. Two bactericidal drugs from different antimicrobial drug classes plus a PSI or an RNAI Ciprofloxacin plus meropenem plus doxycycline or omadacycline** Regimen 2. One bactericidal drug plus a PSI Meropenem plus linezolid or doxycycline Regimen 3. One bactericidal drug plus a second bactericidal drug from a different antimicrobial drug class Meropenem plus ciprofloxacin Regimen 4. One bactericidal drug plus an RNAI (rifampin should not be used as monotherapy) Meropenem plus rifampin Regimen 5. A PSI plus an RNAI (rifampin should not be used as monotherapy) Linezolid or doxycycline or clindamycin plus rifampin Regimen 6. Two PSIs from different antimicrobial drug classes Linezolid plus doxycycline Regimen 7. A single bactericidal drug Meropenem Regimen 8. A single PSI Linezolid or doxycycline or clindamycin First-line antimicrobial drug ††Bactericidal drug PSI Treatment (listed drugs joined by "or" are considered equivalent) Dosage Treatment Dosage Meropenem §§ 2 g every 8 hours IV Doxycycline ¶¶ 200 mg x 1 dose IV, then 100 mg every 12 hours IV or Ciprofloxacin ¶¶ 400 mg every 8 hours IV or Levofloxacin ¶¶ 500 mg every 12 hours IV PCN-S only: Penicillin G ¶¶ 4 million units every 4 hours IV or Ampicillin §§ 2 g every 4 hours IV Imipenem/cilastatin §§ 1 g every 6 hours IV or Ampicillin/sulbactam §§ 3 g every 6 hours IV Alternative antimicrobial drug*** Bactericidal drug PSI/RNAI Treatment Dosage Treatment Dosage Piperacillin/tazobactam §§ 3.375 g every 4 hours IV Omadacycline §§,†††200 mg x 1 dose IV on day 1, then 100 mg every 24 hours IV Moxifloxacin §§ 400 mg every 24 hours IV Eravacycline §§,†††1 mg/kg every 12 hours IV Vancomycin §§,†††15 mg/kg every 12 hours IV over a period of 1–2 hours (target AUC 24 of 400–600 µ g x h/mL [preferred]; if AUC 24 is not available, maintain serum trough concentrations of 15–20 µ g/mL). Consider a loading dose of 20-35 mg/kg for critically ill patients. Clindamycin §§ 900 mg every 8 hours IV Linezolid §§ 600 mg every 12 hours IV Rifampin §§,§§§ 600 mg every 12 hours IV plus Antitoxin (single dose as an adjunct to antimicrobial drug; listed antitoxins joined by "or" are considered equivalent) Treatment Dosage Raxibacumab ¶¶¶ 40 mg/kg IV or Obiltoxaximab ¶¶¶ 16 mg/kg IV AIGIV 420 units IV Abbreviations: AIGIV = anthrax immunoglobulin intravenous; AUC 24 = area under the concentration-time curve from 0 to 24 hours; FDA = Food and Drug Administration; IV = intravenous; PCN-S = penicillin-susceptible strains; PEPAbx = antimicrobial postexposure prophylaxis for anthrax; PSI = protein synthesis inhibitor; RNAI = RNA synthesis inhibitor. * Definitive therapy should be directed by antibiotic susceptibility test results, when available. †Pregnant adolescents: refer to pediatric guidelines for weight-based dosing (see Table 14 ). § Dosing recommended for pregnant persons regardless of trimester. If meningitis is not suspected and susceptibilities are known, start at regimen 2. ¶ "Systemic" was defined as including evidence of organ damage or any of the following: hyperthermia or hypothermia, tachycardia, tachypnea, hypotension, or leukocytosis or leukopenia ( Source: Katharios-Lanwermeyer S, Holty JE, Person M, et al. Identifying meningitis during an anthrax mass casualty incident: systematic review of systemic anthrax since 1880. Clin Infect Dis 2016;62:1537–45). ** For anthrax meningitis, consider using antimicrobial drugs that have demonstrated potential neuroprotective benefits in vivo (e.g., minocycline, doxycycline, clindamycin, and ß-lactamase inhibitors). ††For highly bioavailable antimicrobial drugs (e.g., ciprofloxacin, doxycycline, and linezolid), if the IV formulation is not available, oral formulations can be considered for patients with an intact gastrointestinal tract where absorption is expected to be complete after oral administration. §§ Not approved by FDA for anthrax PEPAbx or treatment. ¶¶ Approved by FDA for anthrax PEPAbx, treatment, or both, but specific uses (e.g., doses, dosing schedules, and patient populations) recommended in this report might differ from the FDA-approved labeling. *** Alternative selections are for patients who have contraindications to or cannot tolerate first-line antimicrobial drugs or if first-line antimicrobial drugs are not available. †††This antimicrobial does not cross an intact blood-brain barrier but can cross with meningitis because of breakdown of the barrier. §§§ Rifampin is an RNAI and also bactericidal; however, it should not be used as monotherapy. ¶¶¶ Premedicate with IV or oral diphenhydramine within 1 hour before administration. Hypersensitivity and anaphylaxis have been reported after raxibacumab and obiltoxaximab administration. **** An 840-unit dose of AIGIV can be considered for severe cases. PEP and Treatment Regimens for Cutaneous Anthrax Without Signs and Symptoms of Meningitis For pregnant and lactating persons aged ≥18 years, empiric PEP ( Table 9 ) and empiric cutaneous anthrax treatment ( Table 10 ) regimens include either a single antimicrobial drug or a single antitoxin. These regimens are summarized as follows: Antimicrobial drug: Choose a single antimicrobial drug. º Antimicrobial drugs are listed in descending order of preference in the table. Listed drugs joined by "or" are considered equivalent. ° Continue or switch antimicrobial drug based on susceptibility testing once available. ° Only choose a "PCN-S only" antimicrobial drug after the strain has been determined to be penicillin susceptible. Antitoxin: Choose a single antitoxin if no antimicrobial drugs are available. For all pregnant and lactating persons who have an aerosol exposure (e.g., a bioterrorism-related incident or animal skin drum–related event), PEPAbx should continue for 60 days from the exposure whether or not vaccine is given ( 33 ). For nonaerosol (i.e., cutaneous or ingestion) exposures, PEPAbx should continue for 7 days and vaccine is not needed. For pregnant and lactating persons with cutaneous anthrax without signs and symptoms of meningitis, the treatment regimen should continue for 7–10 days, or until clinical criteria for stability are met. If an aerosol exposure might have occurred, patients should transition from a treatment to a PEP regimen ( Table 9 ); the combined regimen should total 60 days from exposure. If no aerosolizing event occurred, patients with cutaneous anthrax do not need to continue PEPAbx. Treatment Regimens for Systemic Anthrax With or Without Meningitis For pregnant or lactating persons aged ≥18 years, empiric treatment regimens for those with systemic anthrax with or without meningitis ( Table 11 ) summarized as follows: Antimicrobial drugs: Choose two bactericidal drugs from different antimicrobial drug classes plus a PSI or an RNAI. º Antimicrobial drugs are listed in descending order of preference in the table. Listed drugs joined by "or" are considered equivalent. º Continue or switch antimicrobial drugs based on susceptibility testing once available. º Only choose a "PCN-S only" antimicrobial drug after the strain has been determined to be penicillin susceptible. Antitoxin: Choose a single antitoxin as adjunctive therapy. If an appropriate combination of bactericidal drugs plus a PSI or an RNAI is contraindicated, not well tolerated, or not available or if meningitis is considered unlikely, consider the following regimens in descending order of preference: One bactericidal drug plus a PSI (start with this regimen if meningitis is not suspected and susceptibilities are known) One bactericidal drug plus a second bactericidal drug from a different antimicrobial drug class One bactericidal drug plus an RNAI A PSI plus an RNAI Two PSIs from different antimicrobial drug classes A single bactericidal drug A single PSI From a PK/PD perspective, minocycline and doxycycline are the preferred PSIs because they provide more robust drug exposures in plasma and CSF compared with macrolides or clindamycin. A single RNAI (i.e., rifampin) should not be used as monotherapy because of the potential for rapid development of resistance ( 135 ). In addition, when meningitis is not suspected, certain oral formulations are included as alternatives in case IV formulations are not available. Duration of antimicrobial drug treatment should be for 2 weeks or longer; however, duration can be shortened and IV administration transitioned to oral medication based on patient improvement and clinical judgment. Patients with naturally acquired noninhalation cases do not need continuation of antimicrobial drugs for PEP. If an aerosol exposure might have occurred (i.e., a bioterrorism-related incident or animal skin drum–related event), healthy patients treated for systemic disease need no further antimicrobial drugs for PEP because they will have developed natural immunity. However, patients with compromised immune systems should transition to an oral PEP regimen ( Table 9 ). The total duration of antimicrobial drugs (i.e., treatment plus PEP) should be 60 days from onset of illness. PEP and Treatment Regimens for Cutaneous Anthrax Without Signs and Symptoms of Meningitis For pregnant and lactating persons aged ≥18 years, empiric PEP ( Table 9 ) and empiric cutaneous anthrax treatment ( Table 10 ) regimens include either a single antimicrobial drug or a single antitoxin. These regimens are summarized as follows: Antimicrobial drug: Choose a single antimicrobial drug. º Antimicrobial drugs are listed in descending order of preference in the table. Listed drugs joined by "or" are considered equivalent. ° Continue or switch antimicrobial drug based on susceptibility testing once available. ° Only choose a "PCN-S only" antimicrobial drug after the strain has been determined to be penicillin susceptible. Antitoxin: Choose a single antitoxin if no antimicrobial drugs are available. For all pregnant and lactating persons who have an aerosol exposure (e.g., a bioterrorism-related incident or animal skin drum–related event), PEPAbx should continue for 60 days from the exposure whether or not vaccine is given ( 33 ). For nonaerosol (i.e., cutaneous or ingestion) exposures, PEPAbx should continue for 7 days and vaccine is not needed. For pregnant and lactating persons with cutaneous anthrax without signs and symptoms of meningitis, the treatment regimen should continue for 7–10 days, or until clinical criteria for stability are met. If an aerosol exposure might have occurred, patients should transition from a treatment to a PEP regimen ( Table 9 ); the combined regimen should total 60 days from exposure. If no aerosolizing event occurred, patients with cutaneous anthrax do not need to continue PEPAbx. Treatment Regimens for Systemic Anthrax With or Without Meningitis For pregnant or lactating persons aged ≥18 years, empiric treatment regimens for those with systemic anthrax with or without meningitis ( Table 11 ) summarized as follows: Antimicrobial drugs: Choose two bactericidal drugs from different antimicrobial drug classes plus a PSI or an RNAI. º Antimicrobial drugs are listed in descending order of preference in the table. Listed drugs joined by "or" are considered equivalent. º Continue or switch antimicrobial drugs based on susceptibility testing once available. º Only choose a "PCN-S only" antimicrobial drug after the strain has been determined to be penicillin susceptible. Antitoxin: Choose a single antitoxin as adjunctive therapy. If an appropriate combination of bactericidal drugs plus a PSI or an RNAI is contraindicated, not well tolerated, or not available or if meningitis is considered unlikely, consider the following regimens in descending order of preference: One bactericidal drug plus a PSI (start with this regimen if meningitis is not suspected and susceptibilities are known) One bactericidal drug plus a second bactericidal drug from a different antimicrobial drug class One bactericidal drug plus an RNAI A PSI plus an RNAI Two PSIs from different antimicrobial drug classes A single bactericidal drug A single PSI From a PK/PD perspective, minocycline and doxycycline are the preferred PSIs because they provide more robust drug exposures in plasma and CSF compared with macrolides or clindamycin. A single RNAI (i.e., rifampin) should not be used as monotherapy because of the potential for rapid development of resistance ( 135 ). In addition, when meningitis is not suspected, certain oral formulations are included as alternatives in case IV formulations are not available. Duration of antimicrobial drug treatment should be for 2 weeks or longer; however, duration can be shortened and IV administration transitioned to oral medication based on patient improvement and clinical judgment. Patients with naturally acquired noninhalation cases do not need continuation of antimicrobial drugs for PEP. If an aerosol exposure might have occurred (i.e., a bioterrorism-related incident or animal skin drum–related event), healthy patients treated for systemic disease need no further antimicrobial drugs for PEP because they will have developed natural immunity. However, patients with compromised immune systems should transition to an oral PEP regimen ( Table 9 ). The total duration of antimicrobial drugs (i.e., treatment plus PEP) should be 60 days from onset of illness. Children Aged ≥1 Month to 12 to 12 to 8 years: 450 mg every 24 hours orally x 2 days, then 300 mg every 24 hours orally Clarithromycin** , **** 75 mg/kg every 12 hours orally (maximum 500 mg/dose; only initiate after at least 3 days of treatment with any of the other antimicrobial drugs listed) Dalbavancin** Aged ≥3 months to 10 kg to 40 kg: 60 mg/kg as a single dose IV >40 kg: 40 mg/kg as a single dose IV or Obiltoxaximab ††††≤15 kg: 32 mg/kg as a single dose IV >15 kg to 40 kg: 24 mg/kg as a single dose IV >40 kg: 16 mg/kg as a single dose IV Abbreviations: FDA = Food and Drug Administration; PCN-S = penicillin-susceptible strains; IV = intravenous; PEP = postexposure prophylaxis; PEPAbx = antimicrobial postexposure prophylaxis for anthrax. * Definitive therapy should be directed by antibiotic susceptibility test results, when available. †Approved by FDA for anthrax PEPAbx, treatment, or both, but specific uses (e.g., doses, dosing schedules, and patient populations) recommended in this report might differ from the FDA-approved labeling. § If liquid formulations are not available for children who cannot swallow pills, instructions are available for preparing oral suspensions of moxifloxacin ( Source: Hutchinson DJ, Johnson CE, Klein KC. Stability of extemporaneously prepared moxifloxacin oral suspensions. Am J Health Syst Pharm 2009;66:665–7.121) and doxycycline ( Source: CDC. In an anthrax emergency: how to prepare doxycycline hyclate for children and adults who cannot swallow pills. Atlanta, GA: US Department of Health and Human Services, CDC; 2020. https://www.cdc.gov/anthrax/public-health/doxy-crushing-instruction-pamphlet.html ). ¶ Up to 10% of naturally occurring strains of anthrax are resistant to penicillin and amoxicillin; bioterror strains can be engineered to generate resistance to multiple antibiotics. Susceptibility results reported from CDC within 48–72 hours of initial isolation of anthrax. ** Not approved by FDA for anthrax PEPAbx or treatment. ††Ampicillin 25 mg/kg every 6 hours orally (maximum 500 mg/dose) can be used as an alternative to amoxicillin, if available. §§ To minimize potential side effects from clavulanate, the amoxicillin/clavulanate 875/125 mg every 12 hours orally and amoxicillin/clavulanate extended-release 2,000/125 mg every 12 hours orally treatments are preferred over 500/125 mg every 8 hours orally. ¶¶ Alternative selections are for patients who have contraindications to or cannot tolerate first-line antimicrobial drugs or if first-line antimicrobial drugs are not available. *** Not approved by FDA for any pediatric indications. Not approved by FDA for anthrax PEPAbx or treatment. †††For children aged 12–17 years who weigh ≥45 kg with risk factors for cardiac events, consider 200 mg twice daily. §§§ The use of tetracycline-class antibiotics during tooth development (i.e., last half of pregnancy, infancy, and childhood up to age 8 years) might cause permanent discoloration of the teeth (yellow, gray, or brown) and enamel hypoplasia. This adverse effect appears to occur less often with doxycycline but might occur more with longer durations of therapy. ¶¶¶ Linezolid can be considered for PEP in scenarios when patients can receive regular laboratory testing to monitor for myelosuppression or neurotoxicity, which might occur within 14–28 days of use. If possible, switch to a different drug when available. **** Clarithromycin is unlikely to be effective if the patient has bacteremia, thus a different antimicrobial drug must be used initially to clear bacteremia. ††††Premedicate with IV or oral diphenhydramine within 1 hour before administration. Hypersensitivity and anaphylaxis have been reported after raxibacumab and obiltoxaximab administration. TABLE 13 Empiric* treatment regimens for children aged ≥1 month to 12 to 12 to 8 years: 450 mg every 24 hours orally x 2 days, then 300 mg every 24 hours orally Clarithromycin** ,§§§ 7.5 mg/kg every 12 hours orally (maximum 500 mg/dose; only initiate after at least 3 days of treatment with any of the other antimicrobial drugs listed) Dalbavancin** Aged ≥3 months to 10–40 kg: 60 mg/kg as a single dose IV >40 kg: 40 mg/kg as a single dose IV or Obiltoxaximab ¶¶¶ ≤15 kg: 32 mg/kg as a single dose IV >15–40 kg: 24 mg/kg as a single dose IV >40 kg: 16 mg/kg as a single dose IV AIGIV**** >10 kg: 1 vial (approximately 60 units) IV 10 to 5 kg. TABLE 14 Empiric* treatment regimens for children aged ≥1 month to 8 years: 1 mg/kg every 12 hours IV or Linezolid ††Aged 10 to 40 kg: 60 mg/kg as a single dose IV >40 kg: 40 mg/kg as a single dose IV or Obiltoxaximab §§§§ ≤15 kg: 32 mg/kg as a single dose IV >15 to 40 kg: 24 mg/kg as a single dose IV >40 kg: 16 mg/kg as a single dose IV AIGIV ¶¶¶¶ >10 kg: 1 vial (approximately 60 units) IV 10 to 5 kg. PEP and Treatment Regimens for Cutaneous Anthrax Without Signs and Symptoms of Meningitis For children aged ≥1 month to 1 month to 1 month to <18 years with systemic anthrax with or without meningitis ( Table 14 ) are summarized as follows: Antimicrobial drugs: Choose two bactericidal drugs from different antimicrobial drug classes plus a PSI or an RNAI. º Antimicrobial drugs are listed in descending order of preference in the table. Listed drugs joined by "or" are considered equivalent. º Continue or switch antimicrobial drugs based on susceptibility testing once available. º Only choose a "PCN-S only" antimicrobial drug after the strain has been determined to be penicillin susceptible. Antitoxin: Choose a single antitoxin as adjunctive therapy. If an appropriate combination of bactericidal drug plus a PSI or an RNAI is contraindicated, not well tolerated, or not available for treatment of noncutaneous systemic anthrax, consider the following regimens in descending order of preference: One bactericidal drug plus a PSI (start with this regimen if meningitis is not suspected) One bactericidal drug plus a second bactericidal agent from a different antimicrobial drug class One bactericidal drug plus an RNAI A PSI plus an RNAI Two PSIs from different antimicrobial drug classes A single bactericidal drug A single PSI From a PK/PD perspective, minocycline and doxycycline are the preferred PSIs because they provide more robust drug exposures in plasma and CSF compared with macrolides or clindamycin. A single RNAI (i.e., rifampin) should not be used as monotherapy because of the potential for rapid development of resistance ( 135 ). When meningitis is not suspected, certain oral formulations are included as alternatives in case IV formulations are not available. Duration of antimicrobial drug treatment should be for 2 weeks or longer; however, duration can be shortened and IV administration transitioned to oral medication based on patient improvement and clinical judgment. Patients with naturally acquired noninhalation anthrax do not need continuation of antimicrobial drug therapy for PEP. If an aerosol exposure might have occurred (i.e., a bioterrorism-related incident or animal skin drum–related event), patients who are immunocompetent do not need further antimicrobial drug therapy because they will have developed natural immunity. Patients who are immunocompromised should transition to an oral PEP regimen ( Table 12 ). The total duration of antimicrobial drug therapy (i.e., treatment plus PEP) should be 60 days from onset of illness. Preterm and Full-Term Newborns Virtually no data are available on antimicrobial drug dosing in neonates and premature infants. Dosing guidance for anthrax in newborn infants is based on extrapolation of data from older populations by using pharmacologic data modeling that incorporates antimicrobial drug kinetics, safety, and efficacy in newborns and how the broad range of developmental changes in this immature population affects therapy ( 32 , 141 – 171 ). Recommendations for both preterm and full-term newborns 32–44 weeks' postmenstrual age (i.e., gestational age plus chronologic age) are available for empiric PEP ( Table 15 ), empiric treatment of cutaneous anthrax without signs and symptoms of meningitis ( Table 16 ), and empiric treatment of systemic anthrax with or without meningitis ( Table 17 ). For neonates of earlier gestational age or without developmentally appropriate renal and hepatic function, providers should consult with a neonatologist, pharmacologist, or infectious diseases physician for appropriate dosing. TABLE 15 Empiric* postexposure prophylaxis for preterm and full-term neonates 32–44 weeks' postmenstrual age (gestational age plus chronologic age) after exposure to Bacillus anthracis , by descending order preference — CDC recommendations, 2023 Treatment 32 to <34 weeks' gestational age 34 to <37 weeks' gestational age Full-term infant 0 to <1 week 1–4 weeks 0 to <1 week 1–4 weeks 0 to <1 week 1–4 weeks Dosage Dosage Dosage Dosage Dosage Dosage First-line antimicrobial drug PCN-S only †: Amoxicillin §,¶ 15 mg/kg every 12 hours orally 15 mg/kg every 8 hours orally 15 mg/kg every 8 hours orally 15 mg/kg every 8 hours orally 15 mg/kg every 8 hours orally 15 mg/kg every 8 hours orally Penicillin VK § 25 mg/kg every 12 hours orally 25 mg/kg every 8 hours orally 25 mg/kg every 12 hours orally 25 mg/kg every 8 hours orally 25 mg/kg every 8 hours orally 25 mg/kg every 6 hours orally Penicillin G** aqueous 25,000 units/kg every 12 hours IM 25,000 units/kg every 12 hours IM 25,000 units/kg every 12 hours IM 25,000 units/kg every 12 hours IM 25,000 units/kg every 12 hours IM 25,000 units/kg every 12 hours IM Amoxicillin/clavulanate § 25 mg amoxicillin/kg every 12 hours orally 25 mg amoxicillin/kg every 8 hours orally 25 mg amoxicillin/kg every 8 hours orally 25 mg amoxicillin/kg every 8 hours orally 25 mg amoxicillin/kg every 8 hours orally 25 mg amoxicillin/kg every 8 hours orally Ciprofloxacin** 7.5 mg/kg every 12 hours orally 12.5 mg/kg every 12 hours orally 12.5 mg/kg every 12 hours orally 12.5 mg/kg every 12 hours orally 12.5 mg/kg every 12 hours orally 12.5 mg/kg every 12 hours orally Clindamycin § 7 mg/kg every 8 hours orally 7 mg/kg every 8 hours orally 7 mg/kg every 8 hours orally 7 mg/kg every 8 hours orally 9 mg/kg every 8 hours orally 9 mg/kg every 8 hours orally Doxycycline** ,††, §§ 5 mg/kg every 12 hours orally 5 mg/kg every 12 hours orally 5 mg/kg every 12 hours orally 5 mg/kg every 12 hours orally 5 mg/kg x 1dose orally, then 2.5 mg/kg every 12 hours orally 5 mg/kg x 1 dose orally, then 2.5 mg/kg every 12 hours orally Levofloxacin** 10 mg/kg every 12 hours orally 10 mg/kg every 8 hours orally 10 mg/kg every 12 hours orally 10 mg/kg every 8 hours orally 10 mg/kg every 12 hours orally 10 mg/kg every 8 hours orally Alternative antimicrobial drug ¶¶ Moxifloxacin ††, *** 25 mg/kg every 24 hours orally 25 mg/kg every 24 hours orally 25 mg/kg every 24 hours orally 25 mg/kg every 24 hours orally 25 mg/kg every 24 hours orally 25 mg/kg every 24 hours orally Linezolid §,†††10 mg/kg every 8 hours orally 10 mg/kg every 8 hours orally 10 mg/kg every 8 hours orally 10 mg/kg every 8 hours orally 10 mg/kg every 8 hours orally 10 mg/kg every 8 hours orally Antitoxin (only to be used if antimicrobial drugs are not available or not appropriate; listed antitoxins joined by "or" are considered equivalent) Raxibacumab §§§ 55 mg/kg as a single dose IV 55 mg/kg as a single dose IV 55 mg/kg as a single dose IV 55 mg/kg as a single dose IV 55 mg/kg as a single dose IV 55 mg/kg as a single dose IV or Obiltoxaximab §§§ 16 mg/kg as a single dose IV 16 mg/kg as a single dose IV 16 mg/kg as a single dose IV 16 mg/kg as a single dose IV 16 mg/kg as a single dose IV 16 mg/kg as a single dose IV Abbreviations: FDA = Food and Drug Administration; IM = intramuscular; IV = intravenous; PCN-S = penicillin-susceptible strains; PEP = postexposure prophylaxis; PEPAbx = antimicrobial postexposure prophylaxis for anthrax. * Definitive therapy should be directed by antibiotic susceptibility test results, when available. †Up to 10% of naturally occurring strains of anthrax are penicillin- and amoxicillin-resistant; bioterror strains might be engineered to generate resistance to multiple antibiotics. Susceptibility results reported from CDC within 48–72 hours of initial isolation of anthrax. § Not approved by FDA for anthrax PEPAbx or treatment. ¶ Ampicillin can be used as an alternative to amoxicillin, if available. ** Approved by FDA for anthrax PEPAbx, treatment, or both, but specific uses (e.g., doses, dosing schedules, and patient populations) recommended in this report might differ from the FDA-approved labeling. ††Instructions are available for preparing oral suspensions of moxifloxacin ( Source: Hutchinson DJ, Johnson CE, Klein KC. Stability of extemporaneously prepared moxifloxacin oral suspensions. Am J Health Syst Pharm 2009;66:665–7.121) and doxycycline ( Source: CDC. In an anthrax emergency: how to prepare doxycycline hyclate for children and adults who cannot swallow pills. Atlanta, GA: US Department of Health and Human Services, CDC; 2020. https://www.cdc.gov/anthrax/public-health/doxy-crushing-instruction-pamphlet.html ). §§ The use of tetracycline-class antibiotics during tooth development (i.e., last half of pregnancy, infancy, and childhood up to age 8 years) might cause permanent discoloration of the teeth (yellow, gray, or brown) and enamel hypoplasia. This adverse effect appears to occur less often with doxycycline but might occur more with longer durations of therapy. ¶¶ Alternative selections are for patients who have contraindications to or cannot tolerate first-line antimicrobial drugs or if first-line antimicrobial drugs are not available. *** Not approved by FDA for any pediatric indications. Not approved by FDA for anthrax PEPAbx or treatment. †††Linezolid can be considered for PEP in scenarios when patients can receive regular monitoring for myelosuppression or neurotoxicity, which might occur within 14–28 days of use. If possible, switch to a different drug when available. §§§ Premedicate with IV or oral diphenhydramine within 1 hour before administration. Hypersensitivity and anaphylaxis have been reported after raxibacumab and obiltoxaximab administration. TABLE 16 Empiric* treatment regimens for preterm and full-term neonates 32–44 weeks' postmenstrual age (gestational age plus chronologic age) with cutaneous anthrax without signs and symptoms of meningitis, by descending order of preference — CDC recommendations, 2023 Treatment 32 to <34 weeks' gestational age 34 to <37 weeks' gestational age Full-term infant 0 to <1 week 1–4 weeks 0 to <1 week 1–4 weeks 0 to <1 week 1–4 weeks Dosage Dosage Dosage Dosage Dosage Dosage First-line antimicrobial drug Ciprofloxacin †10 mg/kg every 12 hours orally 10 mg/kg every 12 hours orally 10 mg/kg every 12 hours orally 10 mg/kg every 12 hours orally 15 mg/kg every 12 hours orally 15 mg/kg every 12 hours orally PCN-S only § : Amoxicillin ¶, ** 25 mg/kg every 12 hours orally 25 mg/kg every 8 hours orally 25 mg/kg every 8 hours orally 25 mg/kg every 8 hours orally 25 mg/kg every 8 hours orally 25 mg/kg every 8 hours orally Penicillin VK ¶ 25 mg/kg every 12 hours orally 25 mg/kg every 8 hours orally 25 mg/kg every 12 hours orally 25 mg/kg every 8 hours orally 25 mg/kg every 8 hours orally 25 mg/kg every 6 hours orally Amoxicillin/clavulanate ¶ 25 mg amoxicillin/kg every 12 hours orally 25 mg amoxicillin/kg every 8 hours orally 25 mg amoxicillin/kg every 12 hours orally 25 mg amoxicillin/kg every 8 hours orally 25 mg amoxicillin/kg every 8 hours orally 25 mg amoxicillin/kg every 8 hours orally Doxycycline †,††5 mg/kg every 12 hours orally 5 mg/kg every 12 hours orally 5 mg/kg every 12 hours orally 5 mg/kg every 12 hours orally 5 mg/kg x 1 dose orally, then 2.5 mg/kg every 12 hours orally 5 mg/kg x 1 dose orally, then 2.5 mg/kg every 12 hours orally Clindamycin ¶ 7 mg/kg every 8 hours orally 7 mg/kg every 8 hours orally 7 mg/kg every 8 hours orally 7 mg/kg every 8 hours orally 9 mg/kg every 8 hours orally 9 mg/kg every 8 hours orally Levofloxacin †10 mg/kg every 12 hours orally 10 mg/kg every 8 hours orally 10 mg/kg every 12 hours orally 10 mg/kg every 8 hours orally 10 mg/kg every 12 hours orally 10 mg/kg every 8 hours orally Alternative antimicrobial drug §§ Moxifloxacin ††,¶¶ 25 mg/kg every 24 hours orally 25 mg/kg every 24 hours orally 25 mg/kg every 24 hours orally 25 mg/kg every 24 hours orally 25 mg/kg every 24 hours orally 25 mg/kg every 24 hours orally Linezolid ¶ 10 mg/kg every 12 hours IV 10 mg/kg every 8 hours IV 10 mg/kg every 12 hours IV 10 mg/kg every 8 hours IV 10 mg/kg every 12 hours IV 10 mg/kg every 8 hours IV Meropenem ¶ 13.3 mg/kg every 8 hours IV 20 mg/kg every 8 hours IV 20 mg/kg every 8 hours IV 20 mg/kg every 8 hours IV 20 mg/kg every 8 hours IV 20 mg/kg every 8 hours IV Vancomycin ¶, *** 20 mg/kg loading dose IV, then 15 mg/kg every 12 hours IV 20 mg/kg loading dose IV, then 15 mg/kg every 12 hours IV 20 mg/kg loading dose IV, then 15 mg/kg every 8 hours IV 20 mg/kg loading dose IV, then 15 mg/kg every 8 hours IV 20 mg/kg loading dose IV, then 15 mg/kg every 8 hours IV 20 mg/kg loading dose IV, then 15 mg/kg every 8 hours IV Administer over a period of 1–2 hours. After dose 3 of vancomycin, adjust dosages to target AUC 24 of 400 µ g x h/mL [preferred]; if AUC 24 is not available, maintain trough concentrations of 10–15 µ g/mL. Check concentrations earlier if renal function is impaired. During the first 7–10 days, serum creatinine represents maternal concentration. Omadacycline ¶¶,†††NA NA NA 5.5 mg/kg loading dose IV x 1, then 3.85 mg/kg every 24 hours IV 5.5 mg/kg loading dose IV x 1, then 3.85 mg/kg every 24 hours IV 5.5 mg/kg loading dose IV x 1, then 3.85 mg/kg every 24 hours IV Dalbavancin ¶ NA NA NA NA NA 22.5 mg/kg x 1 dose IV Antitoxin (only to be used if antimicrobial drugs are not available or not appropriate; listed antitoxins joined by "or" are considered equivalent) Raxibacumab §§§ 55 mg/kg as a single dose IV 55 mg/kg as a single dose IV 55 mg/kg as a single dose IV 55 mg/kg as a single dose IV 55 mg/kg as a single dose IV 55 mg/kg as a single dose IV or Obiltoxaximab §§§ 16 mg/kg as a single dose IV 16 mg/kg as a single dose IV 16 mg/kg as a single dose IV 16 mg/kg as a single dose IV 16 mg/kg as a single dose IV 16 mg/kg as a single dose IV AIGIV 1 vial (approximately 60 units) as a single dose IV 1 vial (approximately 60 units) as a single dose IV 1 vial (approximately 60 units) as a single dose IV 1 vial (approximately 60 units) as a single dose IV 1 vial (approximately 60 units) as a single dose IV 1 vial (approximately 60 units) as a single dose IV Source: Bradley JS, Nelson JD, eds. Nelson's pediatric antimicrobial therapy. 29th ed. Itasca, IL: American Academy of Pediatrics; 2023. Abbreviations: FDA = Food and Drug Administration; IV = intravenous; NA = not applicable; PCN-S = penicillin-susceptible strains; PEPAbx = antimicrobial postexposure prophylaxis for anthrax. * Definitive therapy should be directed by antibiotic susceptibility test results, when available. †Approved by FDA for anthrax PEPAbx, treatment, or both, but specific uses (e.g., doses, dosing schedules, and patient populations) recommended in this report might differ from the FDA-approved labeling. § Up to 10% of naturally occurring strains of anthrax are resistant to penicillin and amoxicillin; bioterror strains might be engineered to generate resistance to multiple antibiotics. Susceptibility results reported from CDC within 48–72 hours of initial isolation of anthrax. ¶ Not approved by FDA for anthrax PEPAbx or treatment. ** Ampicillin can be used as an alternative to amoxicillin, if available. ††Instructions are available for preparing oral suspensions of moxifloxacin ( Source: Hutchinson DJ, Johnson CE, Klein KC. Stability of extemporaneously prepared moxifloxacin oral suspensions. Am J Health Syst Pharm 2009;66:665–7.121) and doxycycline ( Source: CDC. In an anthrax emergency: how to prepare doxycycline hyclate for children and adults who cannot swallow pills. Atlanta, GA: US Department of Health and Human Services, CDC; 2020. https://www.cdc.gov/anthrax/public-health/doxy-crushing-instruction-pamphlet.html ). §§ Alternative selections are for patients who have contraindications to or cannot tolerate first-line antimicrobial drugs or if first-line antimicrobial drugs are not available. ¶¶ Not approved by FDA for any pediatric indications. Not approved by FDA for anthrax PEPAbx or treatment. *** Allergic reactions are rare in neonates, but neonates can release histamine that causes hypotension after rapid infusions of vancomycin; thus, it can be safest to pretreat with an antihistamine. †††The use of tetracycline-class antibiotics during tooth development (i.e., last half of pregnancy, infancy, and childhood up to age 8 years) might cause permanent discoloration of the teeth (yellow, gray, or brown) and enamel hypoplasia. This adverse effect appears to occur less often with doxycycline but might occur more with longer durations of therapy. §§§ Premedicate with IV or oral diphenhydramine within 1 hour before administration. Hypersensitivity and anaphylaxis have been reported after raxibacumab and obiltoxaximab administration. TABLE 17 Empiric* treatment regimens for preterm and full-term neonates 32–44 weeks' postmenstrual age (gestational age plus chronologic age) with systemic †anthrax with or without meningitis, by descending order of preference — CDC recommendations, 2023 Regimen Example Regimen 1. Two bactericidal drugs from different antimicrobial drug classes plus a PSI or an RNAI Ciprofloxacin plus meropenem plus clindamycin § Regimen 2. One bactericidal drug plus a PSI Meropenem plus doxycycline Regimen 3. One bactericidal drug plus a second bactericidal drug from a different antimicrobial drug class Meropenem plus ciprofloxacin Regimen 4. One bactericidal drug plus an RNAI (rifampin should not be used as monotherapy) Meropenem plus rifampin Regimen 5. A PSI plus an RNAI (rifampin should not be used as monotherapy) Doxycycline plus rifampin Regimen 6. Two PSIs from different antimicrobial drug classes Linezolid plus doxycycline Regimen 7. A single bactericidal drug Meropenem Regimen 8. A single PSI Doxycycline or clindamycin Treatment Mechanism of action 32 to <34 weeks' gestational age 34 to <37 weeks' gestational age Full-term infant 0 to <1 week 1–4 weeks 0 to <1 week 1–4 weeks 0 to <1 week 1–4 weeks Dosage Dosage Dosage Dosage Dosage Dosage First-line antimicrobial drug ¶ Ciprofloxacin** C 7.5 mg/kg every 12 hours IV 12.5 mg/kg every 12 hours IV 12.5 mg/kg every 12 hours IV 12.5 mg/kg every 12 hours IV 12.5 mg/kg every 12 hours IV 12.5 mg/kg every 12 hours IV Levofloxacin** C 10 mg/kg every 12 hours IV 10 mg/kg every 8 hours IV 10 mg/kg every 12 hours IV 10 mg/kg every 8 hours IV 10 mg/kg every 12 hours IV 10 mg/kg every 8 hours IV Meropenem ††C 13.3 mg/kg every 8 hours IV 20 mg/kg every 8 hours IV 20 mg/kg every 8 hours IV 20 mg/kg every 8 hours IV 20 mg/kg every 8 hours IV 20 mg/kg every 8 hours IV PNC-S only §§ : Penicillin G** ,¶¶ aqueous C 100,000 units/kg every 12 hours IV/IM 100,000 units/kg every 8 hours IV/IM 100,000 units/kg every 8 hours IV/IM 100,000 units/kg every 6 hours IV/IM 100,000 units/kg every 8 hours IV/IM 100,000 units/kg every 6 hours IV/IM Ampicillin ††C 50 mg/kg every 12 hours IV/IM 75 mg/kg every 12 hours IV/IM 50 mg/kg every 8 hours IV/IM 50 mg/kg every 8 hours IV/IM 50 mg/kg every 8 hours IV/IM 50 mg/kg every 8 hours IV/IM Vancomycin ††, *** C 20 mg/kg loading dose IV, then 15 mg/kg every 12 hours IV 20 mg/kg loading dose IV, then 15 mg/kg every 12 hours IV 20 mg/kg loading dose IV, then 15 mg/kg every 8 hours IV 20 mg/kg loading dose IV, then 15 mg/kg every 8 hours IV 20 mg/kg loading dose IV, then 15 mg/kg every 8 hours IV 20 mg/kg loading dose IV, then 15 mg/kg every 8 hours IV Administer over a period of 1–2 hours. After dose 3 of vancomycin, adjust dosages to target AUC 24 of 400 µ g x h/mL [preferred]; if AUC 24 is not available, maintain trough concentrations of 10–15 µ g/mL. Check concentrations earlier if renal function is impaired. During the first 7–10 days, serum creatinine represents maternal concentration. Doxycycline** PSI 5 mg/kg every 12 hours IV 5 mg/kg every 12 hours IV 5 mg/kg every 12 hours IV 5 mg/kg every 12 hours IV 5 mg/kg x 1 dose IV, then 2.5 mg/kg every 12 hours 5 mg/kg x 1 dose IV, then 2.5 mg/kg every 12 hours Alternative antimicrobial drug †††Moxifloxacin §§§ C 25 mg/kg every 24 hours IV 25 mg/kg every 24 hours IV 25 mg/kg every 24 hours IV 25 mg/kg every 24 hours IV 25 mg/kg every 24 hours IV 25 mg/kg every 24 hours IV Imipenem ††,¶¶¶ C 25 mg/kg every 8 hours IV infused over 1.5 hours 25 mg/kg every 8 hours IV infused over 1.5 hours 25 mg/kg every 8 hours IV infused over 1.5 hours 25 mg/kg every 8 hours IV infused over 1.5 hours 25 mg/kg every 8 hours IV infused over 1.5 hours 25 mg/kg every 8 hours IV infused over 1.5 hours Dalbavancin ††C NA NA NA NA NA 22.5 mg/kg x 1 dose IV PCN-S only ††: Ampicillin/sulbactam ††C 50 mg ampicillin/kg every 12 hours IV 50 mg ampicillin/kg every 12 hours IV 50 mg ampicillin/kg every 12 hours IV 50 mg ampicillin/kg every 12 hours IV 50 mg ampicillin/kg every 12 hours IV 50 mg ampicillin/kg every 12 hours IV Clindamycin ††PSI 7 mg/kg every 8 hours IV/IM 7 mg/kg every 8 hours IV/IM 7 mg/kg every 8 hours IV/IM 7 mg/kg every 8 hours IV/IM 9 mg/kg every 8 hours IV/IM 9 mg/kg every 8 hours IV/IM Linezolid ††PSI 10 mg/kg every 12 hours IV 10 mg/kg every 8 hours IV 10 mg/kg every 12 hours IV 10 mg/kg every 8 hours IV 10 mg/kg every 12 hours IV 10 mg/kg every 8 hours IV Rifampin ††, **** RNAI 10 mg/kg every 24 hours IV 10 mg/kg every 24 hours IV 10 mg/kg every 24 hours IV 10 mg/kg every 24 hours IV 10 mg/kg every 24 hours IV 10 mg/kg every 24 hours IV Omadacycline §§§,††††PSI NA NA NA 5.5 mg/kg loading dose orally x 1, then 3.85 mg/kg dose daily orally 5.5 mg/kg loading dose orally x 1, then 3.85 mg/kg dose daily orally 5.5 mg/kg loading dose orally x 1, then 3.85 mg/kg dose daily orally Antitoxin (only to be used if antimicrobial drugs are not available or not appropriate; listed antitoxins joined by "or" are considered equivalent) Raxibacumab §§§§ 55 mg/kg as a single dose IV 55 mg/kg as a single dose IV 55 mg/kg as a single dose IV 55 mg/kg as a single dose IV 55 mg/kg as a single dose IV 55 mg/kg as a single dose IV or Obiltoxaximab §§§§ 32 mg/kg as a single dose IV 32 mg/kg as a single dose IV 32 mg/kg as a single dose IV 32 mg/kg as a single dose IV 32 mg/kg as a single dose IV 32 mg/kg as a single dose IV AIGIV 1 vial (approximately 60 units) as a single dose IV 1 vial (approximately 60 units) as a single dose IV 1 vial (approximately 60 units) as a single dose IV 1 vial (approximately 60 units) as a single dose IV 1 vial (approximately 60 units) as a single dose IV 1 vial (approximately 60 units) as a single dose IV Abbreviations: AIGIV = anthrax immunoglobulin intravenous; C = bactericidal; FDA = Food and Drug Administration; IM = intramuscular; IV = intravenous; NA = not applicable; PCN-S = penicillin-susceptible strains; PEPAbx = antimicrobial postexposure prophylaxis for anthrax; PSI = protein synthesis inhibitor; RNAI = RNA synthesis inhibitor. * Definitive therapy should be directed by antibiotic susceptibility test results, when available. †"Systemic" was defined as one or more of the following using cutoffs for adults aged ≥18 years: hyperthermia or hypothermia, tachycardia, tachypnea, hypotension, or neutrophilia or neutropenia. ( Source: Katharios-Lanwermeyer S, Holty JE, Person M, et al. Identifying meningitis during an anthrax mass casualty incident: systematic review of systemic anthrax since 1880. Clin Infect Dis 2016;62:1537–45). § For anthrax meningitis, consider using antimicrobial drugs that have demonstrated potential neuroprotective benefits in vivo (e.g., doxycycline, clindamycin, and ß-lactam antimicrobial drugs). ¶ For highly bioavailable antimicrobial drugs (e.g., ciprofloxacin, doxycycline, and linezolid), if the IV formulation is not available, oral formulations can be considered for patients with an intact gastrointestinal tract where absorption is expected to be complete after oral administration. ** Approved by FDA for anthrax PEPAbx, treatment, or both, but specific uses (e.g., doses, dosing schedules, and patient populations) recommended in this report might differ from the FDA-approved labeling. ††Not approved by FDA for anthrax PEPAbx or treatment. §§ Up to 10% of naturally occurring strains of anthrax are resistant to penicillin and amoxicillin; bioterror strains might be engineered to generate resistance to multiple antibiotics. Susceptibility results reported from CDC within 48–72 hours of initial isolation of anthrax. ¶¶ Penicillin G benzathine or procaine should never be administered intravenously. *** Vancomycin does not cross an intact blood-brain barrier but might cross with meningitis because of breakdown of the barrier. Allergic reactions are rare in neonates; however, neonates can release histamine that causes hypotension after rapid infusions of vancomycin, thus it might be safest to pretreat with an antihistamine. †††Alternative selections are for patients who have contraindications to or cannot tolerate first-line antimicrobial drugs or if first-line antimicrobial drugs are not available. §§§ Not approved by FDA for any pediatric indications. Not approved by FDA for anthrax PEPAbx or treatment. ¶¶¶ If meningitis is confirmed, imipenem moves to a first-line antimicrobial drug. **** Rifampin is an RNAI and also bactericidal; however, it should not be used as monotherapy. ††††The use of tetracycline-class antibiotics during tooth development (i.e., last half of pregnancy, infancy, and childhood up to age 8 years) might cause permanent discoloration of the teeth (yellow, gray, or brown) and enamel hypoplasia. This adverse effect appears to occur less often with doxycycline but might occur more with longer durations of therapy. §§§§ Premedicate with IV or oral diphenhydramine within 1 hour before administration. Hypersensitivity and anaphylaxis have been reported after raxibacumab and obiltoxaximab administration. PEP and Treatment Regimens for Cutaneous Anthrax Without Signs and Symptoms of Meningitis For preterm and full-term newborns 32–44 weeks' postmenstrual age (i.e., gestational age plus chronologic age), empiric PEP ( Table 15 ), and empiric cutaneous anthrax treatment ( Table 16 ) regimens include either a single antimicrobial drug or a single antitoxin. These regimens are summarized as follows: Antimicrobial drug: Choose a single antimicrobial drug. ° Antimicrobial drugs are listed in descending order of preference in the table. Listed drugs joined by "or" are considered equivalent. ° Continue or switch antimicrobial drug based on susceptibility testing once available. ° Only choose a "PCN-S only" antimicrobial drug after the strain has been determined to be penicillin susceptible. Antitoxin: Choose a single antitoxin if no antimicrobial drugs are available. For preterm and full-term newborns 32–44 weeks' postmenstrual age (i.e., gestational age plus chronologic age), PEPAbx after aerosol exposure should continue for 60 days ( 33 ). Vaccine is not currently indicated for this age group. PEPAbx after nonaerosol exposure should continue for 7 days. For preterm and full-term newborns with cutaneous anthrax without signs and symptoms of meningitis, the treatment regimen should continue for 7–10 days, or until clinical criteria for stability are met. If an aerosol exposure might have occurred, patients should transition from a treatment to a PEP regimen ( Table 15 ); the combined regimen should total 60 days from exposure. If no aerosolizing event occurred, patients with cutaneous anthrax do not need to continue PEPAbx. Treatment Regimens for Systemic Anthrax With or Without Meningitis For preterm and full-term newborns 32–44 weeks' postmenstrual age (i.e., gestational age plus chronologic age), empiric treatment regimens for those with systemic anthrax with or without meningitis ( Table 17 ) are summarized as follows: Antimicrobial drugs: Choose two bactericidal drugs from different antimicrobial drug classes plus a PSI or an RNAI. º Antimicrobial drugs are listed in descending order of preference in the table. Listed drugs joined by "or" are considered equivalent. º Continue or switch antimicrobial drugs based on susceptibility testing once available. º Only choose a "PCN-S only" antimicrobial drug after the strain has been determined to be penicillin susceptible. Antitoxin: Choose a single antitoxin as adjunctive therapy. If an appropriate combination of bactericidal drugs plus a PSI or an RNAI is contraindicated, not well tolerated, or not available or if meningitis is considered unlikely, consider the following regimens in descending order of preference: One bactericidal drug plus a PSI (start with this regimen if meningitis is not suspected) One bactericidal drug plus a second bactericidal agent from a different antimicrobial drug class One bactericidal drug plus an RNAI A PSI plus an RNAI Two PSIs from different antimicrobial drug classes A single bactericidal drug A single PSI From a PK/PD perspective, minocycline and doxycycline are the preferred PSIs because they provide more robust drug exposures in plasma and CSF compared with macrolides or clindamycin. A single RNAI (i.e., rifampin) should not be used as monotherapy because of the potential for rapid development of resistance ( 135 ). In addition, when meningitis is not suspected, certain oral formulations are included as alternatives in case IV formulations are not available. Duration of antimicrobial drug treatment should be for 2 weeks or longer, although as immune-compromised hosts, neonates might require a longer duration of therapy to achieve cure. Transition from IV administration to oral medication for neonates tolerating regular feeding should be based on patient improvement and clinical judgment. Patients with naturally acquired noninhalation cases do not need continuation of antimicrobial drug therapy for PEP. If an aerosol exposure might have occurred (i.e., a bioterrorism-related incident or animal skin drum–related event), preterm and full-term newborns (who are not considered fully immunocompetent) should transition to an oral PEP regimen ( Table 15 ). The total duration of antimicrobial drug therapy (i.e., treatment plus PEP) should be 60 days from onset of illness. PEP and Treatment Regimens for Cutaneous Anthrax Without Signs and Symptoms of Meningitis For preterm and full-term newborns 32–44 weeks' postmenstrual age (i.e., gestational age plus chronologic age), empiric PEP ( Table 15 ), and empiric cutaneous anthrax treatment ( Table 16 ) regimens include either a single antimicrobial drug or a single antitoxin. These regimens are summarized as follows: Antimicrobial drug: Choose a single antimicrobial drug. ° Antimicrobial drugs are listed in descending order of preference in the table. Listed drugs joined by "or" are considered equivalent. ° Continue or switch antimicrobial drug based on susceptibility testing once available. ° Only choose a "PCN-S only" antimicrobial drug after the strain has been determined to be penicillin susceptible. Antitoxin: Choose a single antitoxin if no antimicrobial drugs are available. For preterm and full-term newborns 32–44 weeks' postmenstrual age (i.e., gestational age plus chronologic age), PEPAbx after aerosol exposure should continue for 60 days ( 33 ). Vaccine is not currently indicated for this age group. PEPAbx after nonaerosol exposure should continue for 7 days. For preterm and full-term newborns with cutaneous anthrax without signs and symptoms of meningitis, the treatment regimen should continue for 7–10 days, or until clinical criteria for stability are met. If an aerosol exposure might have occurred, patients should transition from a treatment to a PEP regimen ( Table 15 ); the combined regimen should total 60 days from exposure. If no aerosolizing event occurred, patients with cutaneous anthrax do not need to continue PEPAbx. Treatment Regimens for Systemic Anthrax With or Without Meningitis For preterm and full-term newborns 32–44 weeks' postmenstrual age (i.e., gestational age plus chronologic age), empiric treatment regimens for those with systemic anthrax with or without meningitis ( Table 17 ) are summarized as follows: Antimicrobial drugs: Choose two bactericidal drugs from different antimicrobial drug classes plus a PSI or an RNAI. º Antimicrobial drugs are listed in descending order of preference in the table. Listed drugs joined by "or" are considered equivalent. º Continue or switch antimicrobial drugs based on susceptibility testing once available. º Only choose a "PCN-S only" antimicrobial drug after the strain has been determined to be penicillin susceptible. Antitoxin: Choose a single antitoxin as adjunctive therapy. If an appropriate combination of bactericidal drugs plus a PSI or an RNAI is contraindicated, not well tolerated, or not available or if meningitis is considered unlikely, consider the following regimens in descending order of preference: One bactericidal drug plus a PSI (start with this regimen if meningitis is not suspected) One bactericidal drug plus a second bactericidal agent from a different antimicrobial drug class One bactericidal drug plus an RNAI A PSI plus an RNAI Two PSIs from different antimicrobial drug classes A single bactericidal drug A single PSI From a PK/PD perspective, minocycline and doxycycline are the preferred PSIs because they provide more robust drug exposures in plasma and CSF compared with macrolides or clindamycin. A single RNAI (i.e., rifampin) should not be used as monotherapy because of the potential for rapid development of resistance ( 135 ). In addition, when meningitis is not suspected, certain oral formulations are included as alternatives in case IV formulations are not available. Duration of antimicrobial drug treatment should be for 2 weeks or longer, although as immune-compromised hosts, neonates might require a longer duration of therapy to achieve cure. Transition from IV administration to oral medication for neonates tolerating regular feeding should be based on patient improvement and clinical judgment. Patients with naturally acquired noninhalation cases do not need continuation of antimicrobial drug therapy for PEP. If an aerosol exposure might have occurred (i.e., a bioterrorism-related incident or animal skin drum–related event), preterm and full-term newborns (who are not considered fully immunocompetent) should transition to an oral PEP regimen ( Table 15 ). The total duration of antimicrobial drug therapy (i.e., treatment plus PEP) should be 60 days from onset of illness. Special Considerations for Inhalation and Ingestion Anthrax Pleural effusion and other fluid collections are common complications of anthrax ( 28 , 36 ). Hypothetically, draining pleural fluid or ascites might reduce the amount of lethal factor, thereby reducing illness severity and decreasing mortality. In addition, drainage of pleural fluid is believed to improve survival by decreasing mechanical lung compression. Early and aggressive drainage of any clinically or radiographically apparent pleural effusion is recommended; chest tube drainage is preferred over thoracentesis because many effusions will require prolonged drainage. Thoracotomy or video-assisted thoracic surgery might be required to remove gelatinous or loculated collections. Ascites should also be drained, if feasible, and monitored for reaccumulation; continuous drainage might be required. Standard precautions are sufficient when caring for anthrax patients. The exception is when the patient is potentially contaminated with B. anthracis spores. In such cases, the patient should be isolated in an airborne infection isolation room until decontamination is completed ( 172 ). Special Considerations for Anthrax Meningitis Diagnosis Anthrax meningitis has a mortality rate that approaches 100% ( 29 , 38 ) and is a common complication of anthrax. Meningitis can either be primary (i.e., have no obvious route of transmission) or secondary (i.e., develop as a complication of any other form of anthrax). Depending on the route of transmission, 14%–37% of patients with injection, ingestion, systemic cutaneous, or inhalation anthrax develop meningitis ( 38 ). Thus, all patients with symptoms or signs of systemic disease should be evaluated for meningitis. In a wide-area aerosol release of B. anthracis spores mass casualty event, conventional standards of care for diagnosing meningitis (i.e., imaging and lumbar puncture followed by analysis of CSF) might be limited or not available. For such situations, a screening tool has been developed to identify patients likely to have anthrax meningitis. On the basis of this screening tool ( Figure ), patients are likely to have meningitis if they have either ≥2 of the following signs or symptoms: severe headache, altered mental status, meningeal signs, or other neurologic deficits, or ≥1 of the following signs or symptoms: severe headache, altered mental status, meningeal signs, or other neurologic deficits and ≥1 of the following signs or symptoms: nausea/vomiting, abdominal pain, or fever (either subjective or measured) or chills. Patients are unlikely to have meningitis if they do not have severe headache, altered mental status, meningeal signs, and other neurologic deficits. Patients who have bacteremia; those with obesity, diabetes, hypertension, and chronic obstructive pulmonary disease; and current and former smokers appear to be at increased risk for meningitis. Adjunctive Therapy The combined effects of infection and intercranial bleeding predispose patients to malignant, rapidly fatal brain swelling and elevated intracranial pressure. Mannitol or hypertonic saline should be considered for patients with anthrax meningitis and evidence of cerebral edema ( 98 ). The data did not demonstrate a survival benefit in those who received steroids compared with those who did not. However, steroids did not appear to cause harm and should be used if clinically indicated. In addition, therapies that target intracranial bleeding and swelling (e.g., nimodipine) have been reported to improve outcomes in aneurysmal subarachnoid hemorrhage and intracerebral hemorrhage and might be applicable to the treatment of hemorrhagic anthrax meningitis. However, at present no data from animal studies or human patients with anthrax are available to support the theoretical benefit of these treatments ( 51 ). Diagnosis Anthrax meningitis has a mortality rate that approaches 100% ( 29 , 38 ) and is a common complication of anthrax. Meningitis can either be primary (i.e., have no obvious route of transmission) or secondary (i.e., develop as a complication of any other form of anthrax). Depending on the route of transmission, 14%–37% of patients with injection, ingestion, systemic cutaneous, or inhalation anthrax develop meningitis ( 38 ). Thus, all patients with symptoms or signs of systemic disease should be evaluated for meningitis. In a wide-area aerosol release of B. anthracis spores mass casualty event, conventional standards of care for diagnosing meningitis (i.e., imaging and lumbar puncture followed by analysis of CSF) might be limited or not available. For such situations, a screening tool has been developed to identify patients likely to have anthrax meningitis. On the basis of this screening tool ( Figure ), patients are likely to have meningitis if they have either ≥2 of the following signs or symptoms: severe headache, altered mental status, meningeal signs, or other neurologic deficits, or ≥1 of the following signs or symptoms: severe headache, altered mental status, meningeal signs, or other neurologic deficits and ≥1 of the following signs or symptoms: nausea/vomiting, abdominal pain, or fever (either subjective or measured) or chills. Patients are unlikely to have meningitis if they do not have severe headache, altered mental status, meningeal signs, and other neurologic deficits. Patients who have bacteremia; those with obesity, diabetes, hypertension, and chronic obstructive pulmonary disease; and current and former smokers appear to be at increased risk for meningitis. Adjunctive Therapy The combined effects of infection and intercranial bleeding predispose patients to malignant, rapidly fatal brain swelling and elevated intracranial pressure. Mannitol or hypertonic saline should be considered for patients with anthrax meningitis and evidence of cerebral edema ( 98 ). The data did not demonstrate a survival benefit in those who received steroids compared with those who did not. However, steroids did not appear to cause harm and should be used if clinically indicated. In addition, therapies that target intracranial bleeding and swelling (e.g., nimodipine) have been reported to improve outcomes in aneurysmal subarachnoid hemorrhage and intracerebral hemorrhage and might be applicable to the treatment of hemorrhagic anthrax meningitis. However, at present no data from animal studies or human patients with anthrax are available to support the theoretical benefit of these treatments ( 51 ). Research Needs Development of a wide array of medical countermeasures that can lessen the morbidity and mortality of anthrax, especially from bioengineered B. anthracis strains , will make anthrax a less desirable bioweapon. To attain preparedness goals, additional diverse countermeasures are needed to address multidrug-resistant B. anthracis . Newer antimicrobial drugs with novel mechanisms might largely mitigate certain challenges presented by multidrug-resistant strains. However, PK/PD data, on which rational dosing would be based, are limited for these newer antimicrobial drugs. More research is needed on CSF penetration and PK/PD topics (including PK/PD targets for B. anthracis ) to enhance current and newly developed medical countermeasures. These research gaps are especially profound for children and pregnant and lactating persons. In addition, animal and in vitro studies are needed to assess potential synergism (or antagonism) of antimicrobial drug combinations and antitoxin dosing and the benefit of inhibiting toxin production by a PSI. Considerable gaps also exist in knowledge of how ß-lactam antimicrobial drugs bind to multiple target receptors in B. anthracis . Acquisition of receptor binding data would support optimizing ß-lactam combination therapies with and without addition of established or newer ß-lactamase inhibitors. The utility of current antitoxin countermeasures is considerably compromised when anthrax toxins become intracellular. To overcome this gap, studies are needed to develop small-molecule inhibitors. Such agents might be of use in patients during the fulminant phase of disease. Studies are needed on the immunogenicity and safety of currently approved anthrax vaccines in special populations, including pregnant and lactating persons, children, and older adults. Vaccines that enhance immunogenicity sooner also would be desirable. Most deaths after a wide-area aerosol release of B. anthracis spores would likely be from complications of anthrax meningitis (e.g., intracranial bleeding). Unlike other bacterial meningitides, anthrax meningitis is characterized by substantial bleeding and cerebral edema. Previously, neither clinical guidelines nor basic research have focused on the bleeding or cerebral edema that accompanies meningitis or on drugs or molecules that might be neuroprotective for this complication. The best available PK/PD recommendations are provided by predicting antimicrobial drug exposures in plasma and CSF via Monte Carlo simulations. Whereas these analyses were helpful to rank antimicrobial drugs, the inherent limitations of borrowing PK/PD exposure targets from pathogens other than B. anthracis require the PK/PD results to be interpreted conservatively and considered in conjunction with animal efficacy and clinical data. Finally, little is known about host factors that might predispose a person to develop anthrax meningitis. Conclusion Anthrax continues to occur in certain places around the world, with an estimated 20,000–100,000 cases occurring annually ( 173 ). B. anthracis also continues to be considered the most likely bioweapon to be used during a bioterrorist event because of its availability, ease of dissemination, and high mortality rate associated with systemic anthrax. Biopreparedness efforts are made more challenging by the ease with which B. anthracis can be made resistant to first-line antimicrobial drugs for PEP and treatment. This report describes updated CDC guidelines and recommendations for the preferred prevention and treatment regimens for naturally occurring anthrax. Also provided are a wide range of alternative regimens to first-line antimicrobial drugs for use if patients have contraindications or intolerances or after a wide-area aerosol release of B. anthracis spores if resources become limited or a multidrug-resistant B. anthracis strain is used. Future revisions to these guidelines will be supported by new research and technological advancements for prevention and clinical management of anthrax. CDC Anthrax Work Group Steering Committee Randy Elder, PhD; Romeo Galang, MD; Kate Glynn, DVM; Peggy Honein, PhD; Jenny McQuiston, DVM; Dana Meaney-Delman, MD; Christina Nelson, MD; Georgina Peacock, MD; Nicki Pesik, MD; Brett Petersen, MD; Jeremy Sobel, MD; Henry Walke, MD . CDC Anthrax Clinical Guidelines Team Co-Leads : Katherine Hendricks, MD; Yon Yu, PharmD. Members: William A. Bower, MD; Kevin Chatham Stephens, MD; Rachel Cook, MS; Shana Godfred Cato, DO; Elisabeth Hesse, MD; Jordan L. Kennedy, MSPH; Corinne M. Parker, PharmD; Marissa K. Person, MSPH; David Sue, PhD; Patricia Yu, MPH. Anthrax Work Group CDC Co-Leads: William A. Bower, MD; Katherine Hendricks, MD. CDC Members: Kevin Chatham Stephens, MD; Rachel Cook, MS; Shana Godfred Cato, DO; Elisabeth Hesse, MD; Jordan L. Kennedy, MSPH; Corinne M. Parker, PharmD; Marissa K. Person, MSPH; Joanna Prasher, PhD; David Sue, PhD; Yon Yu, PharmD. Non-CDC Members : John Bradley, MD, University of California San Diego School of Medicine; Arthur Friedlander, MD, United States Army Medical Research Institute of Infectious Diseases; Brenna Hughes, MD, Duke University; Adolf W. Karchmer, MD, Beth Israel Deaconess Medical Center; Andrew Pavia, MD, Robert M. Ward, MD, University of Utah School of Medicine; Ray Slay, PhD, National Institutes of Health; Yuliya Yasinskaya, MD, Food and Drug Administration. Pediatric Anthrax Work Group CDC Lead : Kevin Chatham Stephens, MD. Members : John Bradley, MD, University of California San Diego School of Medicine; Margaret (Meg) Fisher, MD, Monmouth Medical Center; Steven Krug, MD, Todd Semla, PharmD, Northwestern University Feinberg School of Medicine; Shannon Manzi, PharmD, Boston Children's Hospital; Scott Needle, MD, Elica Health Centers; Christina Nelson, MD; Rosemary Roberts, MD, Food and Drug Administration; Robert M. Ward, MD, University of Utah School of Medicine. Obstetrics and Gynecology Anthrax Work Group CDC Lead : Shana Godfred Cato, DO. Members : Martina L. Badell, MD, Emory Midtown Perinatal Center; Kristin C. Klein, PharmD, University of Michigan; Laura E. Riley, MD, Weill Cornell Medicine; Jeanne Sheffield, MD, Johns Hopkins Medicine; Robert M. Ward, MD, University of Utah School of Medicine. 2021 Anthrax Expert Forum CDC Co-Leads : Katherine Hendricks, MD; Yon Yu, PharmD. CDC Participants : William A. Bower, MD; Kevin Chatham Stephens, MD; Rachel Cook, MS; Romeo Galang, MPH; Shana Godfred Cato, DO; Elisabeth Hesse, MD; Alex Hoffmaster, PhD; Jordan L. Kennedy, MSPH; Christina Nelson, MD; Corinne M. Parker, PharmD; Marissa K. Person, MSPH; Joanna Prasher, PhD; Jeremy Sobel, MD; David Sue, PhD; Cassidy Whitson, MPH; Patricia Yu, MPH. Non-CDC Participants : Mark Albrecht, PhD, David Boucher, PhD, Michael Merchlinsky, PhD, Biomedical Advanced Research and Development Authority; Martina Badell, MD, Emory University; Richard Beigi, MD, University of Pittsburgh; John Bradley, MD, University of California San Diego School of Medicine; Juergen Bulitta, PhD, University of Florida; Steve Cantrill, MD, Denver Health Medical Center; Herbert (Dele) Davies, MD, University of Nebraska Medical Center; Mehmet Doganay, MD, Erciyes University, Turkey; Ed Eitzen, MD, Martin, Blanck & Associates; Margaret (Meg) Fisher, MD, Monmouth Medical Center; Arthur Friedlander, MD, United States Army Medical Research Institute of Infectious Diseases; Susan Gorman, PharmD, David Christian (Chris) Hassell, PhD, Assistant Secretary for Preparedness and Response; Adolf W. Karchmer, MD, Beth Israel Deaconess Medical Center; Kristin C Klein, PharmD, University of Michigan; Susan Kline, MD, University of Minnesota Medical Center; Steven Krug, MD, Todd Semla, PharmD, Northwestern University Feinberg School of Medicine; Ainura Kutmanova, MD, International School of Medicine, Kyrgyz Republic; Ruth Lynfield, MD, Council of State and Territorial Epidemiologists; Shannon Manzi, PharmD, Boston Children's Hospital; Ryan Maves, MD, Naval Medical Center San Diego; Scott Needle, MD, Elica Health Centers; Ann Christine (Chris) Nyquist, MD, Children's Hospital Colorado; Andrew Pavia, MD, Robert M. Ward, MD, University of Utah School of Medicine; Laura Riley, MD, Weill Cornell Medicine; Rosemary Roberts, MD, Yuliya Yasinskaya, MD, Food and Drug Administration; Jeanne Sheffield, MD, Johns Hopkins Medicine; Kevin Sheth, MD, Yale School of Medicine; J. Marc Simard, MD, PhD, University of Maryland School of Medicine; Ray Slay, PhD, Mary Wright, MD, National Institutes of Health; Julie Thompson, DVM, Tulane University School of Medicine; Nicholas Vietri, MD, Boise Veterans Administration Medical Center.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2706995/

Functions of Phenylalanine Residues within the β-Barrel Stem of the Anthrax Toxin Pore

Background A key step of anthrax toxin action involves the formation of a protein-translocating pore within the endosomal membrane by the Protective Antigen (PA) moiety. Formation of this transmembrane pore by PA involves interaction of the seven 2β2–2β3 loops of the heptameric precursor to generate a 14-strand transmembrane β barrel. Methodology/Principal Findings We examined the effects on pore formation, protein translocation, and cytotoxicity, of mutating two phenylalanines, F313 and F314, that lie at the tip the β barrel, and a third one, F324, that lies part way up the barrel. Conclusions/Significance Our results show that the function of these phenylalanine residues is to mediate membrane insertion and formation of stable transmembrane channels. Unlike F427, a key luminal residue in the cap of the pore, F313, F314, and F324 do not directly affect protein translocation through the pore. Our findings add to our knowledge of structure-function relationships of a key virulence factor of the anthrax bacillus. Background A key step of anthrax toxin action involves the formation of a protein-translocating pore within the endosomal membrane by the Protective Antigen (PA) moiety. Formation of this transmembrane pore by PA involves interaction of the seven 2β2–2β3 loops of the heptameric precursor to generate a 14-strand transmembrane β barrel. Methodology/Principal Findings We examined the effects on pore formation, protein translocation, and cytotoxicity, of mutating two phenylalanines, F313 and F314, that lie at the tip the β barrel, and a third one, F324, that lies part way up the barrel. Conclusions/Significance Our results show that the function of these phenylalanine residues is to mediate membrane insertion and formation of stable transmembrane channels. Unlike F427, a key luminal residue in the cap of the pore, F313, F314, and F324 do not directly affect protein translocation through the pore. Our findings add to our knowledge of structure-function relationships of a key virulence factor of the anthrax bacillus. Introduction Bacillus anthracis , the causative agent of anthrax, produces two major virulence factors: a poly-D-glutamic acid capsule and anthrax toxin. The latter comprises three large monomeric proteins. The Lethal Factor (LF) and the Edema Factor (EF), are enzymes that act on substrates within the cytosol of mammalian cell. The third, Protective Antigen (PA; 83 kDa), is a receptor-binding and pore-forming protein that binds and transports LF and EF from the extracellular milieu of cells to the cytosol [1] . PA binds to a receptor at the cell surface and is proteolytically activated by a furin-family protease, yielding a 63 kDa form (PA 63 ), which self-associates to form a ring-shaped heptamer, termed the prepore. The prepore binds 1–3 copies of EF and/or LF and undergoes receptor mediated endocytosis. Residence within the acidic environment of the endosome induces a conformational change in the PA moiety from the soluble prepore to a membrane-inserted, protein-conducting channel. The PA pore functions as a translocase, mediating unfolding and translocation of bound EF and LF across the membrane [2] . The PA pore is a mushroom-shaped structure, with a globular cap and a 100à -long, 14-stranded, β-barrel stem formed from the seven 2β2–2β3 loops of the prepore ( Fig. 1 ) [3] , [4] . Within the cap there exists a solvent-exposed Phe residue, Phe427, that has been shown to play a crucial role in protein translocation through the pore [5] . The Phe427 side chains of the seven subunits form what has been called the Phe clamp and have been proposed to interact directly with the unfolded translocating polypeptide to form a seal against the passage of ions. This seal may function by preserving the transmembrane pH gradient within the pore, previously demonstrated to be the primary driving force for protein translocation [2] . 10.1371/journal.pone.0006280.g001 Figure 1 Locations of F313, F314, F324 and β strands 1–4 in the β barrel and the pore. The approximate boundaries of the membrane-spanning region are indicated by the dashed lines. The illustration of the pore structure (left), reconstructed from single-pore images obtained by electron microscopy, is modified from Figure 4 of reference [4] . The amino acid sequence of the ® hairpin is as follows: E(302)VHGNAEVHASFFDIGGSVSAGFS(325). At the turn region of the 2β2–2β3 β hairpins that form the stem of the pore there are two Phe residues, F313 and F314 ( Fig. 1 ). Their locations suggest they form a hydrophobic tip of the 14-stranded β barrel that may aid in its insertion into the membrane and/or anchor the tip in the cap-distal leaflet of the bilayer. It is also conceivable, however, that one or both of these residues might face the lumen of the pore and play an active role in translocation, perhaps resembling that of Phe427. We therefore examined the functional consequences of mutating F313 and F314 to various other residues. We also examined the effect of mutating to Ala a third Phe residue in the β barrel proposed to lie on the endosomal leaflet of the membrane interface, F324. Materials and Methods Proteins – Mutations were generated using site-directed mutagenesis (Stratagene) and confirmed by DNA sequencing. Recombinant PA, von Willebrand factor A (VWA) domain, residues 1–263 of LF (LF N ), and LF N fused to the catalytic domain of diphtheria toxin (LF N -DTA) were expressed and purified, and the heptameric PA prepore was prepared from the full length PA, as described [6] . Protein concentrations were estimated from A 280 measurements with the following calculated molar extinction coefficients: PA 83 , 75,670 M −1 cm −1 ; PA 63 , 49,640 M −1 cm −1 ; VWA, 12,485 M −1 cm −1 ; LF N 17,920 M −1 cm −1 ; LF N –DTA, 43,600 M −1 cm −1 . Liposome preparation – Phospholipid (1,2-dioleoyl-sn-glycero-3-phosphocholine) was dried under a nitrogen gas stream, followed by desiccation overnight. The lipid film was hydrated with 1 mL 10 mM HEPES, 100 mM KCl, pH 7.5 to a final concentration of 25 mg/ml, followed by three freeze-thaw cycles and extrusion 11 times through a 200 µm pore size polycarbonate filter (Whatman). The resulting liposomes were stored at 4°C. Immediately before the experiment, the liposomes were exchanged into 10 mM Tris, 100 mM NaCl, pH 8.5, using a G50 desalting column (GE Healthcare) and adjusted to a final concentration of 5 mg/ml. K + release assay – PA prepore (3 nM ) was incubated with 40 nM VWA domain (molar ratio of VWA domain to PA 63 = 2) at room temperature for 15 min, and 20 µl of the sample was mixed with 200 µl liposomes. The mixture was then incubated 5 min and added to 5 ml working solution (50 mM sodium acetate, 100 mM NaCl, pH 5.0). K + release was monitored by means of a K + -selective electrode [6] . The initial rate of K + release (k) was calculated by linearly fitting the K + release curve. LF N translocation across planar lipid bilayers – Planar lipid bilayers were generated as described [7] . The cis and trans compartments were bathed in 1 ml Universal Bilayer Buffer (UBB: 100 mM KCl, 1 mM EDTA, 10 mM each of potassium oxalate, potassium phosphate, 2-(N-morpholino)ethanesulfonic acid, and pH 5.5). Δψ (Δψ = ψ cis −ψ trans ), the membrane potential, was set to +20 mV. PA 63 heptamer (15 ng) was added to the cis compartment. Mutants that produced no channels after addition of1.5 µg PA 63 heptamer were deemed to be devoid of channel-forming activity. For the mutants capable of forming channels, 0.1 nmol LF N was added to the cis compartment after (PA 63 ) 7 channel formation stabilized. Unbound LF N was removed by perfusion with 10 ml UBB at 2 ml/min. To initiate translocation, 7 µl 2 M KOH was added to the trans chamber to raise the pH to 7.2, and the change in current was monitored. Both compartments were stirred continuously throughout the experiments. The half time of translocation ( t 1/2 ) was calculated from sigmoidal fits of averaged normalized data. Single-channel bilayer measurements - Single-channel measurements were obtained under conditions similar to macroscopic current recording with a 100 µm aperture membrane. The cis and trans compartments contained solutions of 1 M KCl, 10 mM 2-(N-morpholino)ethanesulfonic acid at pH 5.5, and current was meassured over a range of positive potentials ( cis -positive). Increments of ∼0.1 ng of PA 63 heptamer were added until a single channel was observed. Data were analyzed using Clampfit, version 10.0, software (Axon Instruments, Sunnyvale, CA), and Microsoft Excel. Analysis was performed on 10-s records. Single-channel conductance was calculated from Gaussian fits to current amplitude histograms. Experiments were repeated 3 times each. Cellular LF N -DTA intoxication assay – The LF N -DTA intoxication assay was performed as described [8] . WT or mutant PA (0.1 µM–10 −9 µM) was incubated with CHO-K1 cells in the presence of 0.1 µM LF N -DTA for 4 h at 37°C. The medium was then replaced with leucine-free F-12 medium supplemented with 1 µCi/ml 3 H-leucine, and the cells were incubated at 37°C for 1 h. Translocation of the diphtheria toxin catalytic domain into the cells was measured by 3 H-leucine incorporation into total cellular protein. Results We used directed mutagenesis to replace F313 and F314 with various other amino acid residues and F324 with Ala. The mutant proteins were expressed in Escherichia coli , purified, and compared with wild-type PA in various assays. For cell-culture toxicity assays we used the purified monomeric proteins, and for assays in model membranes we used the heptameric prepore obtained by activating the monomers with trypsin and isolating the PA prepore by ion-exchange chromatography. To test the effects of mutations on pore formation in a model membrane, we assayed for K + release from KCl-charged liposomes at pH 5.5. The prepore was complexed with the PA-binding VWA domain from anthrax toxin receptor ANTXR2. Binding of the VWA domain, besides approximating the in vivo state, improved the quality of data on the kinetics of K + release by stabilizing the prepore and slowing its conversion to the pore conformation. As shown in Fig. 2 and Table 1 , mutating both F313 and F314 to either Trp (WW) or Tyr (YY) had little effect on the kinetics of K + release, whereas replacing them with Leu caused a two-fold inhibition of initial rate of release. Mutating both of these Phe residues to His (HH), Asp (DD) or Arg (RR), or deleting them (mutant S1) virtually ablated permeablization activity. Deletion of the entire ® strand segment proposed to insert into the membrane (residues 302–325) resulted in a mutant, the "loopless" mutant, that was incapable of permeablizing the membrane. Individual mutations of F313 or F314 to Ala caused 25–50% reduction in the initial rate of permeabilization, and the double Ala mutant reduced the initial rate ∼3-fold. Thus, efficient channel formation depended upon having hydrophobic residues at these positions, aromatic residues being the most active. 10.1371/journal.pone.0006280.g002 Figure 2 Effects of F313 and F314 mutations on PA permeabilization of membranes to K + . For clarity, only selected mutants are shown. The double Trp (WW) and double Tyr (YY) replacements had approximately the same kinetics of K + release as the WT (FF). The kinetics of release by the double Leu (LL), double Ala (AA), and double Arg (RR), as well as the F313A and F314A mutants are also shown. The double His (HH), double Asp (DD), and the buffer control results were indistinguishable from those of RR. 10.1371/journal.pone.0006280.t001 Table 1 Pore formation, LF N translocation and LF N -DTA-dependent cytotoxicity mediated by F313X/F314X mutants. Mutation K + release, initial rate (s −1 ) LF N translocation, t 1/2 (s) Cytotoxicity, EC 50 (pM) FF (wild-type) 24.4±4.0 12±.1 1.6±0.03 WW 22.1±1.9 12±.1 6.5±2.2 YY 23.5±4.7 13±.1 12±3.7 LL 11.2±1.4 12±.1 0.9±0.4 AA 7.3±3.3 * 10±2.2 F313A 17.5±1.0 12±.1 3.7±1.7 F314A 10.9±1.4 12±.1 1.7±0.3 HH − ** 270±72 DD − ** 590±230 RR − ** − S1 − ** 2300±880 "Loopless" − ** − − no activity relative to control. * low pore forming activity; translocation not tested. ** no pores formed in planar bilayers. Activity of these mutants in forming channels in planar phospholipid bilayers correlated well with activity observed in the K + release assay ( Table 1 ). Stable pores were found only with the double Trp, Tyr, and Leu mutants and the single F313A and F314A mutants. Few pores were seen with the double Ala mutant. For a subset of the mutants we measured single-channel currents in planar bilayers. Wild-type PA elicited discrete channel openings with a single-channel conductance of 153±2 pS (in symmetric 1 M KCl). Single-channel conductance values for the double Leu (153±2 pS), double Ala (155±2 pS), and the single F313A (154±2 pS) mutants were indistinguishable from the wild-type (data not shown). Also, the probability of residing in the open state did not vary from the wild-type pore. For each mutant that formed stable pores in planar bilayers, we examined its ability to translocate LF N , the N-terminal domain of LF, across the bilayer. Channels were formed in the membrane upon addition of prepore under an applied potential of +20 mV and symmetric pH 5.5. Subsequent addition of LF N caused current blockage as this protein bound to the channel. After perfusion to remove excess LF N , the pH of the trans compartment was raised to pH 7.2, and translocation was monitored by the alleviation of channel block. The t 1/2 of the translocation reaction deviated less than 10% from that of the wild type (∼12 s) ( Table 1 ). These results imply that mutations at positions 313 and 314 that do not inhibit formation of stable channels are fully competent for protein translocation. To characterize these mutations in a cellular assay we measured the ability of the mutated proteins to transport a model intracellular effector protein, LF N -DTA, to the cytosolic compartment of CHO-K1. LF N -DTA is a fusion protein containing the N-terminal, prepore-binding domain of LF fused to DTA, the catalytic domain of diphtheria toxin. Delivery of LF N -DTA to the cytosol causes inhibition of protein synthesis, resulting in cell death at ∼1 pM PA and 0.1 ▒M LF N -DTA ( Fig. 3 ). Removal of the 2β2–2β3 loop resulted in complete loss of cytoxicity, as did the incorporation of a dominant-negative double mutatation K397D, D425K (dubbed DNI, for dominant-negative inhibitor) [9] . The EC 50 of all of the aromatic and/or non-polar mutants that retained efficient pore-forming activity deviated less than 10-fold from the wild-type value under the conditions of our assay ( Table 1 and Fig. 3 ). Replacement of F313 and F314 with charged residues reduced LFn-DTA cytoxicity by at least 300 fold; mutation to two glycine residues resulted in complete ablation of cytotoxicity. 10.1371/journal.pone.0006280.g003 Figure 3 Cellular translocation assay of F313X/F314X mutants. (A) PA 83 was titrated, mixed with a fixed concentration of LF N -DTA (100 pM), added to CHO-K1 cells, and incubated at 37°C for 4 hours. The level of protein synthesis was measured by detecting [ 3 H] leucine incorporation and was expressed as fraction of that observed with PA-WT. Only WT, S1 (deletion of F313 and F314) and DNI (K397D, D425K) are shown. (B) EC 50 for all F313X/F314X mutants as calculated from sigmoidal fits to cytotoxicity experiments. PA carrying the F324A mutation was tested for activity in the K + release assay, in planar bilayers for translocation activity, and in cell culture for ability to mediate LF N -DTA-dependent cytotoxicity. No differences from wild-type PA were detected. Discussion According to our current model of the membrane-inserted PA pore, F313 and F314 lie in the turn region of the 14-strand β-barrel stem, at or near the aqueous interface of the trans leaflet of the bilayer [3] . In porins and many other membrane proteins, aromatic residues densely populate the boundary between the nonpolar and interfacial regions of the bilayer and are thought to help anchor these proteins in the membrane [10] , [11] . Crystal structures of β-barrel pore forming toxins like 〈 hemolysin and aerolysin have demonstrated that residues lining the trans leaflet of the bilayer in a rivet conformation must be hydrophobic in order to efficiently promote membrane insertion [12] , [13] . Our results demonstrate that the PA is very sensitive to changes in the hydophobicity of the residues at the trans leaflet anchoring position, supporting the hypothesis that two Phe residues alone comprise the rivet [3] . We showed that hydrophobic residues at positions 313 and 314 function well; however hydrophobic aromatic residues are optimal. While the His side chain contains six pi electrons capable of forming pi-stacking interactions it also becomes protonated at pH values below neutrality, and thus it is not surprising that mutation of F313 and F314 to His significantly attenuated PA channel insertion and intoxication. The model is consistent with the hypothesis that the side chains of both F313 and F314 serve to anchor the pore in the membrane. F313 and F314 may also facilitate insertion of the pore, presumably by generating a highly hydrophobic tip — a cluster of 14 Phe residues — that promotes partitioning into the bilayer. The location of F324 in the primary structure suggests that its side chain occupies an analogous location in the interface region of the cis leaflet of the bilayer. Thus, the F324 residues on the cis leaflet and the F313 and F314 residues in the trans leaflet most likely form aromatic girdles analogous to those seen in many integral membrane proteins. We detected no effect of replacing F324 with Ala, indicating that stable pore formation is primarily dependent on the residues at the cytosolic leaflet rather than those at the endosomal leaflet. The fact that single-channel conductance of the F313/F314 mutants examined remained unchanged from that of the wild-type protein in our experiments demonstrates that the passage of ions through the pore was unaffected by the side chains at these locations. Importantly, the half-time of translocation of LF N under the influence of a transmembrane proton gradient did not vary from that of the wild-type protein. These findings are consistent with the notion that the side chains of F313 and F314 are embedded in the membrane, and do not affect passage of monovalent ions or proteins through the pore. The effects of mutating these Phe residues differed strongly from effects of mutating F427, where major changes were seen in both single-channel conductance and protein translocation. The effects of F313/F314 mutations on delivery of LF N -DTA to the cytosol correlated well with the effects of these mutations on K + release. Replacing these residues with charged amino acids had large effects on cytotoxicity, K + release from liposomes, and formation of pores in planar bilayers, as would be predicted from the energetic barrier to membrane penetration by such residues. Deleting F313 and F314 presumably blocked membrane insertion and/or the stability of the pore in the membrane. Many explanations are possible for the smaller variations in activity seen among the other mutants, including, for example, effects on the kinetics of prepore-to-pore conversion resulting from altered side chain interactions with domains 2 and 4 surrounding the 2β2–2β3 loop in the prepore [6] .

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4640854/

Effect of ATM and HDAC Inhibition on Etoposide-Induced DNA Damage in Porcine Early Preimplantation Embryos

Oocyte maturation and embryonic development are sensitive to DNA damage. Compared with somatic cells or oocytes, little is known about the response to DNA damage in early preimplantation embryos. In this study, we examined DNA damage checkpoints and DNA repair mechanisms in parthenogenetic preimplantation porcine embryos. We found that most of the etoposide-treated embryos showed delay in cleavage and ceased development before the blastocyst stage. In DNA-damaged embryos, the earliest positive TUNEL signals were detected on Day 5 of in vitro culture. Caffeine, which is an ATM (ataxia telangiectasia mutated) and ATR (ataxia telangiectasia and Rad3-related protein) kinase inhibitor, and KU55933, which is an ATM kinase inhibitor, were equally effective in rescuing the etoposide-induced cell-cycle blocks. This indicates that ATM plays a central role in the regulation of the checkpoint mechanisms. Treating the embryos with histone deacetylase inhibitors (HDACi) increased embryonic development and reduced etoposide-induced double-strand breaks (DSBs). The mRNA expression of genes involved in non-homologous end-joining (NHEJ) or homologous recombination (HR) pathways for DSB repair was reduced upon HDACi treatment in 5-day-old embryos. Furthermore, HDACi treatment increased the expression levels of pluripotency-related genes ( OCT4 , SOX2 and NANOG ) and decreased the expression levels of apoptosis-related genes ( CASP3 and BAX ). These results indicate that early embryonic cleavage and development are disturbed by etoposide-induced DNA damage. ATMi (caffeine or KU55933) treatment bypasses the checkpoint while HDACi treatment improves the efficiency of DSB repair to increase the cleavage and blastocyst rate in porcine early preimplantation embryos. Introduction DNA double-strand breaks (DSBs) can be induced by endogenous metabolites or metabolic intermediates [ 1 ] as well as by exogenous factors such as UV [ 2 ], ionizing radiation (IR; including X-rays and γ-rays), chemical drugs (such as doxorubicin, bleomycin, MLN4924, and etoposide), and physical inducers (such as laser micro-beam dissection) [ 3 – 6 ]. The occurrence of DSBs activates DNA damage checkpoints and DNA repair mechanisms. DNA damage checkpoint mechanisms arrest cell division until all DSBs are repaired. However, if abnormal DNA DSBs, such as the programmed DSBs, are not repaired immediately they can induce chromatin remodeling, cell cycle arrest, cell cycle delay, apoptosis, or other forms of cell death [ 7 ]. DSBs are known to affect oocyte maturation in multiple species [ 4 – 6 , 8 ]. Compared with oocytes, little is known about the response to DSBs in early preimplantation embryos. In mammals, two molecular pathways are involved in DSBs repair: homologous recombination (HR) and non-homologous end-joining (NHEJ) [ 9 , 10 ]. In response to DNA damage, the protein kinases ATM (ataxia telangiectasia mutated) and ATR (ataxia telangiectasia and Rad3-related protein) phosphorylate checkpoint kinase 2 (CHK2). The activated CHK2 pathway then arrests the cell cycle [ 11 ]. ATM and ATR can also phosphorylate the histone H2AX (γH2AX) at the sites of DSB formation [ 12 , 13 ]. This phosphorylation involves large chromatin domains forming nuclear foci that are easily detected by immunostaining [ 14 ]. The phosphorylation of histone H2AX is critical for DSB repair because H2AX anchors some important initiator proteins that are required for both HR (e.g., RAD51 , MRE11A , BRCA1 ) and NHEJ (e.g., 53BP1 , PRKDC , XRCC6 ) pathways, which can be colocalized with γH2AX at the sites of DSBs [ 15 , 16 ]. After DSBs are repaired, the checkpoint proteins become inactivated to allow cell cycle progression to resume. Caffeine, a methylxanthine, has been widely used to study ATM and ATR signaling because it inhibits these kinases in vitro and allows DNA-damaged cells to bypass checkpoints in vivo [ 2 , 17 – 19 ]. Thus, caffeine is thought to overcome the checkpoint by preventing phosphorylation of ATM and ATR substrates. KU55933 is a potent inhibitor of ATM with a half maximal inhibitory concentration (IC 50 ) of 13 nmol/L, and is highly specific to ATM compared with other PIKKs (phosphoinositide-3-kinase-related protein kinases) [ 20 ]. It prevents the activation of selective ATM targets following DNA DSBs, and prevents the phosphorylation of ATM-dependent DNA damage checkpoints in response to ionizing radiation [ 20 ]. The relative roles of ATM (caffeine- and KU55933-sensitive) and ATR (caffeine-sensitive and KU55933-insensitive) in response to etoposide-induced DSBs in porcine embryos can be determined when KU55933 is used in conjunction with caffeine [ 2 ]. Scriptaid is a novel HDACi with low toxicity and enhanced transcriptional activity [ 21 ]. Scriptaid treatment significantly improves the development of cloned embryos and alleviates aberrant gene expression, as demonstrated in mice and cattle [ 22 , 23 ]. In somatic cells, histone modifications play important roles in DSBs repair. For instance, the acetylation of histone site H4K16 is critical for the DNA damage response and DSB repair by the HR and NHEJ pathways [ 24 ]. In addition, acetylation of histone sites H4K5 and H4K12 at the sites of DSBs facilitates the recruitment of the RAD51 protein [ 25 ]. It has also been shown that Gcn5, which is a histone acetyltransferase (HAT), can interact with γH2AX at the site of DSBs and then acetylate various lysine residues on histone H3 (including H3K9, H3K14, H3K18, and H3K23) [ 26 ]. H3K9Ac and H3K56Ac are reduced in response to DNA damage in human cells [ 27 ]. Based on these discoveries, we hypothesized that increasing histone acetylation by HDACi treatment could facilitate DNA damage repair in porcine early preimplantation embryos. Etoposide is a topoisomerase II (TOP2) inhibitor that causes DSB DNA damage in somatic cells as well as in oocytes, even at low concentrations (5 μg/mL) [ 4 , 28 ]. Exposure to etoposide has been used to study DNA damage signaling and repair mechanisms in both somatic cells and oocytes [ 4 ]. However, to our knowledge, there has been no research on the effect of etoposide in porcine embryos. Therefore, the objectives of this study were: (1) to evaluate the effect of DNA damage on the development of early embryos; and (2) to assess the effect of ATMi (caffeine or KU55933) and HDACi (scriptaid) treatment on the DNA damage response and development in preimplantation embryos. Materials and Methods Oocyte Collection and Culture All chemicals used in this study were purchased from Sigma-Aldrich, unless otherwise stated. All animal studies were performed after receiving the approval of the Institutional Animal Care and Use Committee (IACUC) of Chungbuk National University, Korea. Porcine ovaries were provided by the local slaughterhouse (Farm story dodarm B&F, Umsung, Chungbuk, Korea) and were transported to our laboratory at 25°C in Dulbecco's phosphate-buffered saline supplemented with 75 μg/L penicillin G and 50 μg/L streptomycin sulfate. Cumulus–oocyte complexes (COCs) were aspirated from follicles (diameter approximately 2–8 mm) and were washed three times with HEPES-buffered Tyrode's medium containing 0.1% (w/v) polyvinyl alcohol (PVA)(HEPES is 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid). The collected COCs were matured in tissue culture medium 199 (Gibco) supplemented with 0.1 g/L sodium pyruvate, 0.6 mM L-cysteine, 10 ng/mL epidermal growth factor, 10% porcine follicular fluid (v/v), 10 IU/mL luteinizing hormone, and 10 IU/mL follicle-stimulating hormone for 44 h at 38.5°C in 5% CO 2 and humidified air. After maturation, the cumulus cells were removed by pipetting in the presence of 0.1% hyaluronidase (w/v) for 2–3 min. Oocyte Activation and Embryo Culture The oocytes were activated parthenogenetically by a 5-min treatment with 50 μM Ca 2+ ionophore (A23187, Sigma-Aldrich). After 3 h of culture in porcine zygote medium 5 (PZM-5) supplemented with 7.5 μg/mL cytochalasin B (Sigma), the embryos were washed several times in PZM-5 supplemented with 0.4% (w/v) bovine serum albumin (BSA) and cultured in a humidified atmosphere of 5% CO 2 and 95% air at 38.5°C. Drug Treatments KU55933 (Selleckchem) was used as an ATM inhibitor, caffeine (Sigma) as an ATR and ATM inhibitor, and scriptaid (Sigma) as an HDAC inhibitor. To investigate DNA damage, etoposide (100 mg/mL stock in dimethyl sulfoxide; stored at 4°C until required) was added to the IVM or IVC medium to a final concentration of 0 (controls), 25, 50, and 100 μg/mL. To induce DSBs in oocytes, GV oocytes were cultured in IVM medium containing 25 μg/mL of etoposide for 5 h, after which the cumulus cells were removed to determine the levels of γH2AX or ATM-p. To induce DSBs in embryos, parthenogenic embryos were cultured in IVC medium containing 0 (controls), 25, 50, or 100 μg/mL of etoposide for 5 h starting after removal from cytochalasin B. Parthenogenic embryos were treated or not (control) with KU55933 (10 μM), caffeine (5 mM), or scriptaid (500 μM) for 20 h starting after Ca 2+ ionophore treatment. Immunofluorescence Analysis Oocytes or embryos were washed with phosphate-buffered saline (PBS), fixed in 3.7% paraformaldehyde (w/v) prepared in PBS containing 0.1% PVA, and permeabilized with 1% Triton X-100 (v/v) for 1 h at 37°C. The samples were blocked with 1% BSA (w/v) for 1 h, incubated overnight with different antibodies at 4°C in a blocking solution, and washed with 1% BSA. The oocytes or embryos were then incubated overnight with anti-γH2AX (pS139, 1:100; Cell Signaling Technology), anti-ATM (pS1981, 1:100; Cell Signaling Technology), and anti-Histone H3 (acetyl K9, Abcam, Cambridge) antibodies at 4°C overnight. The oocytes or embryos were washed three times with PBS containing 1% BSA and were labeled with FITC/TR-conjugated antibody (1:100) for 1 h at room temperature (FITC/TR is fluorescein isothiocyanate and Texas Red). The oocytes were then counterstained with 5 μg/mL Hoechst 33342 for 15 min, washed three times with PVA–PBS, mounted on a glass slide, and examined using an LSM 710 META confocal laser-scanning microscope (Zeiss, Jena, Germany). Fluorescence Intensity Analysis For fluorescence intensity analysis, samples of control oocytes/embryos and treated oocytes/embryos were mounted on the same glass slide. ImageJ software (v1.47) was used to define a region of interest (ROI), and the average fluorescence intensity per unit area within the ROI was determined. Independent measurements using identically sized ROIs were made for the cell nucleus. The average values of all measurements were used to compare the final average intensities between control and treated oocytes or embryos. Reverse Transcription Quantitative Polymerase Chain Reaction (RT-qPCR) with SYBR Green Total RNA was extracted from groups of 30 (Day 5) and 20 (Day 7) porcine embryos using the Dynabeads mRNA Direct Kit (Dynal Asa, Oslo, Norway) according to the manufacturer's instructions. First-strand complementary DNA (cDNA) was synthesized by reverse transcription of mRNA using the Oligo (dT) 12–18 primer and SuperScript TM III Reverse Transcriptase (Invitrogen Co., Grand Island, NY). Real-time PCR (also called quantitative PCR) was performed using a CFX96 Touch Real-time PCR Detection System (Bio-Rad) in a final reaction volume of 20 μL including SYBR Green, a fluorophore that binds to double-stranded DNA (qPCR kit from Finnzymes, Finland). The PCR conditions were as follows: 95°C for 10 min followed by 39 cycles of 95°C for 30 s, 60°C for 30 s, and 72°C for 25 s, and a final extension at 72°C for 5 min. Finally, gene expression was quantified using the 2 -ΔΔCt method, and normalized against the mRNA levels of glyceraldehyde 3-phosphate dehydrogenase (GAPDH). The primers used to amplify each gene are shown in Table 1 . 10.1371/journal.pone.0142561.t001 Table 1 Primers used for real-time reverse transcription-PCR. Gene Primer sequences (5'-3') Accession number BCLXL F: TTTTCTCCTTCGGTGGGG NM_214285.1 R: GCATTGTTTCCGTAGAGTTCC BAX F: T ACCTT ACCCTGGGAGTGGC XM_001928147.2 R: GGAAAACCTCCTCCGTGTC CASP3 F: GAAGACCATAGCAAAAGGAGCA NM_214131.1 R: TTTGGGTTTGCCAGTTAGAGTT NANOG F: TTCCTTCCTCCATGGATCTG R: NM_001129971.1 ATCTGCTGGAGGCTGAGGTA SOX2 F: CGCAGACCTACATGAACG NM_001123197.1 R: TCGGACTTGACCACTGAG OCT4 F: AAGCAGTGACTATTCGCAAC NM_001113060.1 R: CAGGGTGGTGAAGTGAGG ATM F: CCGGTGTTTTGGGAGAGTGT NM_001123080.1 R: CTTCCGACCAAACTCAGCGT ATR F: TGAGCTCCAGTGTTGGCATC XM_003132459.3 R: GCCAGTTCTCAGTGTGGTCA RAD51 F: CTTCGGTGGAAGAGGAGAGC NM_001123181.1 R: CGGTGTGGAATCCAGCTTCT BRCA1 F: TGCTAAATCCGGAACAAAACACA XM_003358030.1 R: CTGGTGGAACGATCCAGAGAT MRE11A F: GGAGGATGTTGTCCTGGCTG XM_003129789.2 R: AGACGTTCCCGTTCTGCATT XRCC6 F: ACGGAAGGTGCCCTTTACTG NM_001190185.1 R: TGCAGCACTGGGTTCTCAAA PRKDC F: ATTCTTTGTCGGGAGCAGCA XM_001925309.4 R: CCTAGCTGTGTGGCACATGA TP53BP1 F: GGGAAAGGGGGAGTTCGTG XM_001925938.4 R: CTCACGCTCGTGCTAGAGAT GAPDH F: GCCA TCACCA TCTTCCAGG R: NM_001206359.1 TCACGCCCATCACAAACAT F, forward; R, reverse. Terminal Deoxynucleotidyl Transferase (TdT)-Mediated dUTP-Digoxigenin Nick End Labeling (TUNEL) Assay Embryos were washed three times with PBS (pH 7.4) containing 1 mg/mL polyvinyl pyrrolidone (PBS/PVP). After fixation with 3.7% paraformaldehyde prepared in PBS for 1 h at room temperature (RT), the embryos were washed with PBS/PVP and permeabilized by incubation in 0.5% Triton X-100 for 1 h at RT. The embryos were then washed twice with PBS/PVP and incubated with fluorescein-conjugated dUTP and terminal deoxynucleotidyl transferase (In Situ Cell Death Detection Kit, Roche, Mannheim, Germany) in the dark for 1 h at 37°C. Blastocysts were counterstained with Hoechst 33342 (bis-benzimide H33342 trihydrochloride, Sigma Life Science) to label all nuclei, washed with PBS/PVP, mounted with slight coverslip compression, and examined under an Olympus fluorescence microscope. Statistical Analysis Each experiment was performed at least three times and at least 30 oocytes or 20 embryos were examined each time. Statistical analyses were performed with the SPSS software package (Version 11.5; SPSS). Analysis of variance (ANOVA) was adopted for statistical analyses. Differences between treated groups were evaluated with the Duncan multiple comparison test. Data are expressed as mean ± SD, and a p-value 0.05). (C) The levels of γH2AX in the nuclei of cleaved embryos. The concentration of etoposide used in this experiment was 25 μg/mL. Data are from 3 independent replicates with 30 embryos in each group. HDACi Treatment Leads to Increased Histone Acetylation To determine the level of histone acetylation in porcine embryos treated with HDACi for 20 h, the levels of the epigenetic marker H3K9ac were measured at the one- and two-cell stages of HDACi-treated and non-treated embryos. Under DNA-damaging conditions induced by etoposide-treatment (25 μg/mL), the fold change of the fluorescence signal for H3K9ac increased at both the one-cell (2.17 vs . 1.29) and two-cell (2.35 vs . 1.38) stages in embryos treated with HDACi, compared with non-treated embryos ( Fig 7 ). 10.1371/journal.pone.0142561.g007 Fig 7 HDACi treatment increased the levels of H3K9ac at 15 h (1-cell) and 28 h (2-cell) after activation. (A) The intensity of H3K9ac at 20 h (1-cell) and 38 h (2-cell) after activation. (B) Quantification of (A) to determine the relationship between HDACi and H3K9ac. Data from (A) are presented as normalized fold change. Fold change in H3K9ac was determined using the formula F/F con , where F is the mean value of each group and F con is the control value where no etoposide and HDACi are used. Error bars indicate SD. ETO, etoposide; * P 0.05). To investigate the effects of HDACi treatment on gene expression patterns during early embryonic development, the expression levels of three genes related to pluripotency ( OCT4 , SOX2 and NANOG ) and of three genes related to apoptosis ( BCLXL , CASP3 , and BAX ) were determined in 7-day-old embryos. The mRNA expression levels of CASP3 and BAX were significantly decreased in HDACi-treated embryos under DNA-damage conditions (25 μg/mL etoposide treatment) ( Fig 9A ). On the other hand, the mRNA expression levels of OCT4 , SOX2 and NANOG were significantly increased in HDACi-treated embryos under the same conditions ( Fig 9B ). 10.1371/journal.pone.0142561.g009 Fig 9 The relative expression patterns of apoptosis (A) and pluripotency (B) related genes in 7-day-old embryos. The concentration of etoposide used in this experiment was 25 μg/mL. The mRNA abundance was calculated relative to the reference gene GAPDH. Data are from 3 independent replicates with 20 embryos in each group. ETO, etoposide; * P 0.05). Early Embryonic Cleavage and Development Is Disturbed by Etoposide-Induced DNA DSBs In the non-exposed control group (0 μg/mL etoposide), no embryos had cleaved at 18 h after parthenogenetic activation, whereas 82.87% and 86.38% had cleaved at 28 h and 48 h, respectively ( Fig 1A ). However, only 56.41%, 42.20%, and 29.35%, respectively, of the 25, 50, or 100 μg/mL etoposide-exposed embryos had cleaved at 28 h. Forty-eight hours after treatment with 25, 50, or 100 μg/mL etoposide, the rate of cleavage increased to 73.09%, 70.62%, and 64%, respectively ( Fig 1A ). Although a significant proportion of etoposide-treated embryos could develop beyond the first cleavage stage, development to the blastocyst stage at Day 7 was dramatically reduced ( Fig 1B ). Indeed, development to blastocysts was almost completely blocked by 50 (7.83%) and 100 μg/mL (4.29%) of etoposide exposure, while it was approximately 50% reduced by 25 μg/mL treatment (17.41%), compared with the control embryos (34.77%) ( Fig 1B ). 10.1371/journal.pone.0142561.g001 Fig 1 Development of embryos treated with various concentrations of etoposide. Cleavage ( A ) and blastocyst ( B ) rates of porcine embryos that were unexposed (control, 0 μg/mL etoposide) or exposed to 25, 50, or 100 μg/mL etoposide for 5 h. Cleavage rates were determined at 28 h and 48 h after parthenogenetic activation. Blastocyst rates were detected at day 7 of embryonic culture. Data are from 3 independent replicates with a minimum of 30 embryos in each group. Asterisks denote a significant difference from the controls (* p 0.05). (C) The levels of γH2AX in the nuclei of cleaved embryos. The concentration of etoposide used in this experiment was 25 μg/mL. Data are from 3 independent replicates with 30 embryos in each group. HDACi Treatment Leads to Increased Histone Acetylation To determine the level of histone acetylation in porcine embryos treated with HDACi for 20 h, the levels of the epigenetic marker H3K9ac were measured at the one- and two-cell stages of HDACi-treated and non-treated embryos. Under DNA-damaging conditions induced by etoposide-treatment (25 μg/mL), the fold change of the fluorescence signal for H3K9ac increased at both the one-cell (2.17 vs . 1.29) and two-cell (2.35 vs . 1.38) stages in embryos treated with HDACi, compared with non-treated embryos ( Fig 7 ). 10.1371/journal.pone.0142561.g007 Fig 7 HDACi treatment increased the levels of H3K9ac at 15 h (1-cell) and 28 h (2-cell) after activation. (A) The intensity of H3K9ac at 20 h (1-cell) and 38 h (2-cell) after activation. (B) Quantification of (A) to determine the relationship between HDACi and H3K9ac. Data from (A) are presented as normalized fold change. Fold change in H3K9ac was determined using the formula F/F con , where F is the mean value of each group and F con is the control value where no etoposide and HDACi are used. Error bars indicate SD. ETO, etoposide; * P 0.05). To investigate the effects of HDACi treatment on gene expression patterns during early embryonic development, the expression levels of three genes related to pluripotency ( OCT4 , SOX2 and NANOG ) and of three genes related to apoptosis ( BCLXL , CASP3 , and BAX ) were determined in 7-day-old embryos. The mRNA expression levels of CASP3 and BAX were significantly decreased in HDACi-treated embryos under DNA-damage conditions (25 μg/mL etoposide treatment) ( Fig 9A ). On the other hand, the mRNA expression levels of OCT4 , SOX2 and NANOG were significantly increased in HDACi-treated embryos under the same conditions ( Fig 9B ). 10.1371/journal.pone.0142561.g009 Fig 9 The relative expression patterns of apoptosis (A) and pluripotency (B) related genes in 7-day-old embryos. The concentration of etoposide used in this experiment was 25 μg/mL. The mRNA abundance was calculated relative to the reference gene GAPDH. Data are from 3 independent replicates with 20 embryos in each group. ETO, etoposide; * P 0.05). Discussion In this study, we investigated the effect of etoposide-induced DNA DSBs on cell division and embryonic development in porcine early preimplantation embryos. The results indicate that etoposide-induced DNA DSBs in early embryos elicit DNA damage response, cause cell cycle delay, and lead to cell apoptosis in 5-day-old embryos. Here, we showed that caffeine (an ATM and ATR inhibitor) and KU55933 (an ATM inhibitor) effectively relieve the cell cycle blocks induced by etoposide-treatment in vitro . Furthermore, we found that treatment with HDACi promoted DNA damage repair and improved the development of porcine early embryos. The DNA damage checkpoint seems to function insufficiently in porcine one- and two-cell embryos. Parthenogenesis embryos that had been treated with etoposide divided after a temporal delay of the cell division and arrested at approximately the four- or eight-cell stage. It is likely that the DSBs had not been completely repaired during cell division arrest because the embryos ceased to develop before reaching the blastocyst stage. Therefore, their cleavage prior to the completion of DNA repair suggests that the DNA damage checkpoint mechanisms do not function sufficiently. γH2AX is required to maintain repair factors at the site of DNA DSBs [ 29 ]. Of the two DNA repair mechanisms, NHEJ is active throughout the cell cycle, while HR is restricted to the S and G2 phases of the cell cycle due to the requirement for the sister chromatid as a template [ 30 ]. γH2AX seems to play an important role in both HR and NHEJ mechanisms [ 31 ]. We evaluated the function of the DNA repair mechanisms by examining the phosphorylation of H2AX after etoposide treatment. The detection of phosphorylated H2AX in etoposide-treated embryos suggests that HR and NHEJ were activated. Based on these findings, we propose that the division of etoposide-treated embryonic cells is more likely due to a lack of efficient DNA damage checkpoint mechanisms, rather than a decreased capacity for DNA repair. The insufficient function of the DNA damage checkpoint in porcine one- and two-cell embryos may be caused by a low expression of the involved genes prior to zygotic gene activation (ZGA) [ 32 ]. Pig ZGA was confirmed to occur at the four-cell stage via genome-wide gene expression analysis, compensating for the loss of maternal transcripts [ 33 ]. Thus, if a DNA damage checkpoint gene is insufficiently transcribed, its transcripts would decrease to a low level at the one- or two-cell stage. These levels may increase again at the four-cell stage, allowing a functional DNA damage checkpoint to occur. This hypothesis is supported by the observation that etoposide-treated embryos arrest cell division at approximately the 4–8-cell stage. A major consequence of unrepaired DNA DSBs is cell apoptosis or death. Laser microbeam-induced DNA DSBs cause embryonic cell cycle arrest and finally cell apoptosis at the blastocyst stage in mouse embryos [ 8 ]. We used TUNEL staining as a means of detecting apoptosis in the etoposide-treated embryos. In the present study, the earliest positive TUNEL signals were detected in the etoposide-treated embryos on Day 5 of in vitro culture, while no apoptosis was observed on Days 3 and 4. This result is similar to the results of apoptosis assays previously performed in mouse and pig somatic cell nuclear transfer (SCNT) embryos [ 8 , 34 ]. It is not clear why no apoptosis is observed during the earliest stages. Although every cell has an apoptotic pathway, it is known that chromatin structures change during early development and that the major onset of gene expression begins only at the four-cell embryonic stage [ 33 ]. A previous study has shown that apoptosis depends on DNA methylation and histone acetylation in early embryos [ 34 ]. Furthermore, the propensity to apoptosis is continuously counterbalanced in the cell by genes that stimulate cell survival and proliferation [ 34 , 35 ]. These observations confirm that embryos exhibit dramatic changes in the expression pattern of several apoptosis-related genes when apoptosis is activated during the development of early implantation embryos. Caffeine is a well-established inhibitor of the activity of both ATR and ATM kinases, whereas KU55933 is the first potent and selective ATM inhibitor [ 18 , 20 , 36 ]. Interestingly, our results demonstrate that both caffeine (5 mM) and KU55933 (10 μM) were equally effective at reversing the cell-cycle block that was induced by etoposide treatment. These results hint that ATM plays a major role in mediating the cell-cycle block induced by etoposide treatment. A potential effector pathway is the ATM phosphorylation of CHK2 [ 37 ], which has a number of downstream target proteins capable of inducing a G1-S phase checkpoint. For instance, ATM and CHK2 cause inhibition of cell division cycle 25 homolog A (Cdc25A), which prevents the activation of Cdk2 and results in checkpoint activation [ 11 ]. However, the low levels of ATM activation in one- and two-cell embryos also indicate that the DNA damage checkpoints function insufficiently in these embryos. A possible reason for this limited ATM activity could be low ATM expression levels. This explanation was supported by a study that monitored the total levels of ATM in oocytes and blastocysts, the results of which indicated that the levels of ATM in fully grown oocytes were lower than those in growing oocytes and blastocysts [ 4 ]. These findings support the hypothesis that the lack of an efficient DNA damage checkpoint during early embryonic development is due to the ineffective activation of ATM kinase, the master regulator of the DNA damage response pathway. In addition, ATM activity is known to be influenced by chromatin structure [ 38 ], and early preimplantation embryos have a distinct chromatin configuration and histone modification [ 39 ], both of which may limit the response to DNA damage. Caffeine and KU55933 can reverse the cell-cycle block, but they do not seem to aid in DNA repair, as evidenced by the increased apoptosis in caffeine/KU55933- and etoposide-treated embryos. We next sought to determine if HDACi treatment affects the DNA damage repair in etoposide-treated embryos. Chromatin remodeling and epigenetic changes, including histone modifications, are important parameters for cell reprogramming and normal development of SCNT embryos [ 40 ]. It is known from studies in somatic cells that histone H4 acetylation at lysine 16 is critical for DNA damage response and DSBs repair [ 24 ]. It has also been shown that H3K9Ac and H3K56Ac are reduced in response to DNA damage in human cells [ 27 ]. The HAT Gcn5 can interact with γH2AX at the sites of DSBs and then acetylate various lysine residues on histone H3 (including H3K9, H3K14, H3K18, and H3K23) [ 26 ]. More recently, a study showed that HDACi treatment after nuclear transfer enhances DSB repair and development of SCNT embryos [ 41 ]. Thus, we hypothesized that increasing histone acetylation by HDACi treatment would facilitate DNA damage repair in etoposide-treated embryos. Indeed, we have found that HDACi treatment greatly increased cell division and embryonic development to the blastocyst stage in embryos treated with etoposide. Moreover, HDACi treatment consistently reduced the γH2AX in cleaved etoposide-treated embryos. These findings indicate that HDACi treatment can facilitate DNA damage repair in etoposide-treated embryos. Although HDACi may affect other DNA functions, including transcription and replication [ 42 ], its positive effect on DSB repair seems to be an important route by which HDACi treatment improves the development of etoposide-treated embryos. To further evaluate the effect of HDACi treatment on DSB repair, we assessed the expression of several genes involved in either the HR ( ATM , ATR , RAD51 , MRE11A , BRCA1 ) or the NHEJ ( 53BP1 , PRKDC , XRCC6 ) pathways in etoposide-treated embryos at Day 5 of development. Transcript levels of the genes involved in the two DNA damage repair pathways were increased in these embryos, confirming that both pathways are activated in etoposide-treated embryos. In this study, we observed an overall tendency for a decreasing expression of these genes when the embryos were exposed to HDACi. The lower expression of DNA-repair genes indicates that Day 5 etoposide-treated embryos treated with HDACi have fewer DSBs. In contrast to Day 5 embryos, we observed that none of the assessed genes were differentially expressed in 7-day-old blastocysts, even in those etoposide-treated embryos. This finding is in line with previous observations in embryos produced by somatic cell nuclear transfer, whereas the induction of DNA damage by UV treatment significantly reduced embryo development, but those that achieved the blastocyst stage had similar number of DSBs and levels of transcripts encoding DNA repair genes [ 41 ]. These findings support the hypothesis that only the embryos with less DNA DSBs and/or superior capacity for DNA repair are able to progress to the blastocyst stage. The fact that HDACi could rescue the development of a higher proportion of etoposide-treated embryos, and that important genes involved in apoptosis pathways were normally expressed in the developing blastocysts, further indicates that HDACi treatment promotes DSB repair during embryo development. In summary, the data obtained in this study indicate that etoposide-induced DNA DSBs can alter the kinetics of embryo cell cleavage and development to the blastocyst stage. However, DNA damage checkpoint function seems insufficient in one- and two-cell embryos, and this appears to be mainly due to limited ATM activation. The earliest TUNEL-positive cells were detected in embryos after 5 days of in vitro culture, which indicates that DNA DSBs cause embryonic cell cycle delay and finally cell apoptosis at a particular time-point. Caffeine and KU55933 were equally effective in reversing the etoposide-induced cell-cycle block and allowed porcine early embryos to progress through several further cell cycles. Moreover, we have shown that HDACi exposure can rescue the development of embryos with DSBs by improving DNA damage repair during early embryo development. In conclusion, the findings of this study revealed a negative correlation between the occurrence of DSBs and embryo cleavage kinetics, and the capacity of the embryo to develop to the blastocyst stage. Chromatin remodeling is important for the promotion of DNA damage repair and the preservation of genome integrity for normal embryo development. Supporting Information S1 Fig Relative mRNA abundance for genes involved in homologous recombination (A) and nonhomologous end-joining (B) pathways in day 7 embryos. Non-treated (-Etopodide; white bars) or etoposide-treated (+Etopodide; black bars) embryos were subjected to HDACi or vehicle (control). The mRNA abundance was calculated relative to the reference gene GAPDH. Data are from 3 independent replicates with 20 embryos in each group. ETO, etoposide. (TIFF) Click here for additional data file.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4105404/

Phylogeography of Bacillus anthracis in the Country of Georgia Shows Evidence of Population Structuring and Is Dissimilar to Other Regional Genotypes

Sequence analyses and subtyping of Bacillus anthracis strains from Georgia reveal a single distinct lineage (Aust94) that is ecologically established. Phylogeographic analysis and comparisons to a global collection reveals a clade that is mostly restricted to Georgia. Within this clade, many groups are found around the country, however at least one subclade is only found in the eastern part. This pattern suggests that dispersal into and out of Georgia has been rare and despite historical dispersion within the country, for at least for one lineage, current spread is limited. Introduction Bacillus anthracis , the causative agent of anthrax, continues to decimate livestock herds and cause concern over possible nefarious use. Source attribution of outbreaks is therefore essential for epidemiological and forensic investigations as well as control efforts. Accurate attribution requires determining evolutionary relationships among isolates and understanding how they fit into regional and global phylogeographic patterns. Genetic characterization of B. anthracis first gained considerable traction through multiple locus VNTR analysis (MLVA) [1] which identified major genetic groups. MLVA typing schemes have been expanded to include more loci; however, although these methods provide additional resolution, inferences on relationships among isolates are unreliable due to the prevalence of convergent alleles, negatively affecting the accuracy of downstream attribution conclusions. With the increased accessibility of whole genome sequencing, typing schemes based on single nucleotide polymorphisms (SNPs) provide both high resolution and highly accurate phylogenetic information [2] – [4] although phylogenetic discovery bias [2] must be taken into account. We and others have previously used MLVA and SNP typing to characterize regional and global phylogeographic patterns of B. anthracis . Although ungulate grazers certainly play a role in the dissemination of this species, human-mediated dispersal has contributed to both ancient [5] and recent [6] spread. Contaminated animal products are frequently identified during outbreaks of disease in humans [6] , [7] ; however, ecological establishment of exotic strains is rare [4] , [8] . None the less, distinguishing between endemic and non-indigenous strains is difficult and can obscure phylogeographic patterns and confuse attribution efforts. Intensive regional sampling studies are not only invaluable for identifying endemic strains and for defining regional patterns of dissemination and cycling, but also build the foundation for understanding global patterns of spread. We report here the phylogeographic patterns of B. anthracis samples collected within the country of Georgia. Located in a geographic bottleneck between the Black and Caspian Seas, and between Europe and Asia, this region has been impacted by ancient and modern human movement and trade. Such anthropogenic influences may have shaped the distribution of B. anthracis in this region and may prove to be important in understanding global phylogeographic patterns. Materials and Methods Phylogenetic placement To place the 272 Georgian isolates into the established global phylogeny [2] , [6] – [8] , we screened all isolates with previously described canSNP assays [8] . All isolates were assigned to the Aust94 genetic group. To obtain more detailed resolution within this genetic group, we genotyped the Georgian isolates and 225 additional Aust94 group isolates from our global collection with previously published Aust94 assays [9] ( Table S1 ) as well as with novel assays (see CanSNP Selection and Analysis below). DNA templates were extracted using either chloroform [10] , DNeasy blood and tissue kits (Qiagen, Valencia, CA) or heat soak. Whole Genome Sequencing To further resolve the genetic structure within Georgian Aust94 group, we sequenced the genomes of three Georgian isolates belonging to the Aust94 genetic group, using Illumina's Genome Analyzer II (San Diego, CA). Library preparation for this isolate involved sonication of 5 µg genomic DNA, obtained through a standard chloroform extraction protocol [10] and shearing the DNA to an average fragment size of 350 bp. The library was quantified using SYBR-based qPCR and primers modified from the adaptor sequence. Paired-end read lengths were ∼100 bp. The sequences of 52-G, 9080-G, and 8903-G were deposited into GenBank (PRJNA224563, PRJNA224558, and PRJNA224562, respectively) ( Table S2 ). SNP Discovery and Analysis To identify putative SNPs, homologous genomic regions were identified using MUMmer [11] and aligned to search for SNPs using SolSNP ( http://sourceforge.net/projects/solsnp/ ). We ensured site orthology by eliminating potential paralogs and requiring all genome alignments to include 100 bp flanking each side of the SNP. Furthermore, for analysis inclusion, all SNP loci were required to be present in all of the genomes analyzed. A maximum-parsimony tree was constructed by PAUP 4.0b10 software (Sinauer Associates, Inc., Sunderland, MA, USA) using all putative SNPs from this study and ten published genomes ( Figure 1 , panel A; Table S2 ). 10.1371/journal.pone.0102651.g001 Figure 1 Bacillus anthracis phylogenetics in Georgia. A) Established phylogeny of B. anthracis [2] , [6] – [8] . Terminal subgroups representing sequenced strains are shown as stars, and intervening nodes representing collapsed branches appear as circles. The highlighted yellow box (part of the Aust94 lineage) indicate the phylogenetic location of Georgian strains. Stars within the highlighted yellow area represent the three Georgian strains sequenced for this study. The number of Georgian strains is indicated in red. B) Expansion of the Aust94 group and canSNP subgroups within the Georgian lineage. The number and origin of isolates are shown for each node and stars or red arrows indicate the locations of the sequenced strains. Shapes of nodes correspond to geographic location. (C) Phylogeography of 272 B. anthracis isolates falling in the Aust94 group are mapped across the country of Georgia at a district level. The heat map legend indicates the number of isolates per subgroup found in a given district. CanSNP Selection and Analysis WGS comparisons between Georgian (52-G) and Aust94 (Aust94, AAES00000000) strains revealed 50 putative SNPs specific to 52-G. Of these, twenty-six were incorporated into melt-MAMA genotyping assays, as previously described [9] and eight were selected as canonical SNP assays ( Table 1 ). Allele-specific melt-MAMA primers were designed using Primer Express 3.0 software (Applied Biosystems, Foster City, CA). All other assay reagents and instrumentation were as previously described [9] . PCR reactions were first raised to 50°C for 2 min to activate the uracil glycolase, then raised to 95°C for 10 min to denature the DNA and then cycled at 95°C for 15 s and 55°C–60°C for 1 min for 33 cycles ( Table 1 ). Immediately after the completion of the PCR cycle, amplicon melt dissociation was measured by ramping from 60°C to 95°C in 0.2°C/min increments and recording the fluorescent intensity. 10.1371/journal.pone.0102651.t001 Table 1 Melt-MAMA primers targeting canonical SNPs for 8 new phylogenetic branches discovered in this study a . Branch AmesAnc position b Genome SNP state (D/A) c Melt-MAMA primer sequence d Conc, µM e Annealing temp °C A.Br.026 3,640,599 T/C A:CTTCTTTTAATACATCTAAGTAAGTAAGCGTTgC D:cggggcggggcggggcggggCTTCTTTTAATACATCTAAGTAAGTAAGCGTTcTC:ATTGACCCAACAGCTACGAAATAC 0.45 0.15 0.15 60 A.Br.027 4,355,524 A/G A:CCCATTCCAAGTGACACACTcG D:cggggcggggcggggcggggCCCATTCCAAGTGACACACTgA C:AGCACTTGCTTATCTTGGAGCTT 0.60 0.15 0.15 60 A.Br.028 791,256 A/G A:ACAGAGAAGGTTATAAGTCCAGAcGG D:cggggcggggcggggcggggACAGAGAAGGTTATAAGTCCAGAaGA C:CTCGCTTTTCCTGTTCTTTTATTCAC 0.15 0.15 0.15 60 A.Br.029 3,960,657 A/G A:AGTATTCCAACCATTACTATAGTCACTaG D:ggggcggggcggggcggggcggggcAGTATTCCAACCATTACTATAGTCACTcAC:GTACTTATTGGTGGTACTGCCAAATT 0.15 0.15 0.15 60 A.Br.030 3,528,668 A/G A:CAATCCCTCGATTTACATATAAATATAAaG D:ggggcggggcggggcggggcggggcCAATCCCTCGATTTACATATAAATATAAcAC:AGGTATGTATGAATTAGAAGGGAAGAA 0.15 0.15 0.15 60 A.Br.031 3,018,054 C/T A:ACTATCGCCAAAAGCAATTGaAT D:cggggcggggcggggcggggACTATCGCCAAAAGCAATTGtAC C:TATTTTAGACAAGTACGAACTAGATAAATCAA 0.15 0.15 0.15 55 A.Br.032 3,520,170 G/A A:CCACCAACAACGAATGGAAGtA D:cggggcggggcggggcggggCCACCAACAACGAATGGAAGaG C:AGCATTTAATGAACGGCGTAAGTAATA 0.45 0.15 0.15 60 A.Br.033 3,610,151 C/T A:TAAATAACCAAGGCGTCTTGCCAT D:ggggcggggcggggcggggcggggcCTAAATAACCAAGGCGTCTTGCtAC C:TGTAGGACGTAGTATGGTGAAAGTAGTAGAT 0.60 0.15 0.15 60 a Melt-MAMA, melt-mismatch amplification mutation assay; SNP, single nucleotide polymorphism; con, concentration. b Ames Ancestor reference genome (NC_006570). c SNP states are presented according to the top strand in the Ames ancestor AE017334 D: Derived SNP state; A: Ancestral SNP state. d Melt-mismatch amplification mutation assay (MAMA), A: Ancestral; D: Derived; C: Common. Primer tails and antepenultimate or penultimate mismatch bases are in lower case. e Final concentratinon of each primer in Melt-MAMA genotyping assays. Phylogenetic placement To place the 272 Georgian isolates into the established global phylogeny [2] , [6] – [8] , we screened all isolates with previously described canSNP assays [8] . All isolates were assigned to the Aust94 genetic group. To obtain more detailed resolution within this genetic group, we genotyped the Georgian isolates and 225 additional Aust94 group isolates from our global collection with previously published Aust94 assays [9] ( Table S1 ) as well as with novel assays (see CanSNP Selection and Analysis below). DNA templates were extracted using either chloroform [10] , DNeasy blood and tissue kits (Qiagen, Valencia, CA) or heat soak. Whole Genome Sequencing To further resolve the genetic structure within Georgian Aust94 group, we sequenced the genomes of three Georgian isolates belonging to the Aust94 genetic group, using Illumina's Genome Analyzer II (San Diego, CA). Library preparation for this isolate involved sonication of 5 µg genomic DNA, obtained through a standard chloroform extraction protocol [10] and shearing the DNA to an average fragment size of 350 bp. The library was quantified using SYBR-based qPCR and primers modified from the adaptor sequence. Paired-end read lengths were ∼100 bp. The sequences of 52-G, 9080-G, and 8903-G were deposited into GenBank (PRJNA224563, PRJNA224558, and PRJNA224562, respectively) ( Table S2 ). SNP Discovery and Analysis To identify putative SNPs, homologous genomic regions were identified using MUMmer [11] and aligned to search for SNPs using SolSNP ( http://sourceforge.net/projects/solsnp/ ). We ensured site orthology by eliminating potential paralogs and requiring all genome alignments to include 100 bp flanking each side of the SNP. Furthermore, for analysis inclusion, all SNP loci were required to be present in all of the genomes analyzed. A maximum-parsimony tree was constructed by PAUP 4.0b10 software (Sinauer Associates, Inc., Sunderland, MA, USA) using all putative SNPs from this study and ten published genomes ( Figure 1 , panel A; Table S2 ). 10.1371/journal.pone.0102651.g001 Figure 1 Bacillus anthracis phylogenetics in Georgia. A) Established phylogeny of B. anthracis [2] , [6] – [8] . Terminal subgroups representing sequenced strains are shown as stars, and intervening nodes representing collapsed branches appear as circles. The highlighted yellow box (part of the Aust94 lineage) indicate the phylogenetic location of Georgian strains. Stars within the highlighted yellow area represent the three Georgian strains sequenced for this study. The number of Georgian strains is indicated in red. B) Expansion of the Aust94 group and canSNP subgroups within the Georgian lineage. The number and origin of isolates are shown for each node and stars or red arrows indicate the locations of the sequenced strains. Shapes of nodes correspond to geographic location. (C) Phylogeography of 272 B. anthracis isolates falling in the Aust94 group are mapped across the country of Georgia at a district level. The heat map legend indicates the number of isolates per subgroup found in a given district. CanSNP Selection and Analysis WGS comparisons between Georgian (52-G) and Aust94 (Aust94, AAES00000000) strains revealed 50 putative SNPs specific to 52-G. Of these, twenty-six were incorporated into melt-MAMA genotyping assays, as previously described [9] and eight were selected as canonical SNP assays ( Table 1 ). Allele-specific melt-MAMA primers were designed using Primer Express 3.0 software (Applied Biosystems, Foster City, CA). All other assay reagents and instrumentation were as previously described [9] . PCR reactions were first raised to 50°C for 2 min to activate the uracil glycolase, then raised to 95°C for 10 min to denature the DNA and then cycled at 95°C for 15 s and 55°C–60°C for 1 min for 33 cycles ( Table 1 ). Immediately after the completion of the PCR cycle, amplicon melt dissociation was measured by ramping from 60°C to 95°C in 0.2°C/min increments and recording the fluorescent intensity. 10.1371/journal.pone.0102651.t001 Table 1 Melt-MAMA primers targeting canonical SNPs for 8 new phylogenetic branches discovered in this study a . Branch AmesAnc position b Genome SNP state (D/A) c Melt-MAMA primer sequence d Conc, µM e Annealing temp °C A.Br.026 3,640,599 T/C A:CTTCTTTTAATACATCTAAGTAAGTAAGCGTTgC D:cggggcggggcggggcggggCTTCTTTTAATACATCTAAGTAAGTAAGCGTTcTC:ATTGACCCAACAGCTACGAAATAC 0.45 0.15 0.15 60 A.Br.027 4,355,524 A/G A:CCCATTCCAAGTGACACACTcG D:cggggcggggcggggcggggCCCATTCCAAGTGACACACTgA C:AGCACTTGCTTATCTTGGAGCTT 0.60 0.15 0.15 60 A.Br.028 791,256 A/G A:ACAGAGAAGGTTATAAGTCCAGAcGG D:cggggcggggcggggcggggACAGAGAAGGTTATAAGTCCAGAaGA C:CTCGCTTTTCCTGTTCTTTTATTCAC 0.15 0.15 0.15 60 A.Br.029 3,960,657 A/G A:AGTATTCCAACCATTACTATAGTCACTaG D:ggggcggggcggggcggggcggggcAGTATTCCAACCATTACTATAGTCACTcAC:GTACTTATTGGTGGTACTGCCAAATT 0.15 0.15 0.15 60 A.Br.030 3,528,668 A/G A:CAATCCCTCGATTTACATATAAATATAAaG D:ggggcggggcggggcggggcggggcCAATCCCTCGATTTACATATAAATATAAcAC:AGGTATGTATGAATTAGAAGGGAAGAA 0.15 0.15 0.15 60 A.Br.031 3,018,054 C/T A:ACTATCGCCAAAAGCAATTGaAT D:cggggcggggcggggcggggACTATCGCCAAAAGCAATTGtAC C:TATTTTAGACAAGTACGAACTAGATAAATCAA 0.15 0.15 0.15 55 A.Br.032 3,520,170 G/A A:CCACCAACAACGAATGGAAGtA D:cggggcggggcggggcggggCCACCAACAACGAATGGAAGaG C:AGCATTTAATGAACGGCGTAAGTAATA 0.45 0.15 0.15 60 A.Br.033 3,610,151 C/T A:TAAATAACCAAGGCGTCTTGCCAT D:ggggcggggcggggcggggcggggcCTAAATAACCAAGGCGTCTTGCtAC C:TGTAGGACGTAGTATGGTGAAAGTAGTAGAT 0.60 0.15 0.15 60 a Melt-MAMA, melt-mismatch amplification mutation assay; SNP, single nucleotide polymorphism; con, concentration. b Ames Ancestor reference genome (NC_006570). c SNP states are presented according to the top strand in the Ames ancestor AE017334 D: Derived SNP state; A: Ancestral SNP state. d Melt-mismatch amplification mutation assay (MAMA), A: Ancestral; D: Derived; C: Common. Primer tails and antepenultimate or penultimate mismatch bases are in lower case. e Final concentratinon of each primer in Melt-MAMA genotyping assays. Results and Discussion Screening of 272 Georgian isolates with previously described canSNP assays [9] resulted in the assignment of all isolates to the group defined by the Australia94 (Aust94) genome [8] ( Figure 1 , panel A). The temporal, geographic and phylogenetic diversity of the Georgian Aust94 isolates, coupled with the detection of MLVA A3a [1] (roughly equivalent to the Aust94 lineage) isolates in the region by other researchers [12] – [15] is strong evidence for the ecological establishment of this clade ( Table S3 ). Reference isolates from Turkey in the MLVA A3a group reported in Keim et al. [1] include six genotypes (genotypes 33, 36, 37, 41–43) and thus show regional diversity that is also reflected by the SNP genotyping here. Members of the Aust94 group have been identified on five continents ( Figure 1 , panel B), suggesting extensive dispersal [8] , however detailed phylogeography of any part of this lineage has not been previously described. The resulting whole genome SNP phylogenetic tree ( Figure 1 , panel A) ( Table S2 ) drawn from 13 strains placed the three sequenced Georgian genomes within the Aust94 genetic group. Screening our global collection across previously published SNP assays (11) revealed five genetic groups along the Aust94 lineage ( Figure 1 , panel B) ( Table S1 and S3 ) and eight novel groups along the lineage terminating in the 52-G genome. The branches and topology leading to isolates that were genotyped, but not sequenced, remain unknown due to phylogenetic discovery bias and branch collapse, however clade membership is accurately estimated [2] , [4] . The node between branches A.Br.014 and A.Br.013 (named A.Br.014/013) forms the most basal subgroup and contains isolates collected from countries in five continents ( Figure 1 , panel B). The A.Br.013/015 subgroup contains isolates from Europe (n = 2) and USA (n = 8). It also gives rise to the lineages that contain all 272 Georgian isolates ( Figure 1 , panel B) as well as the lineage leading to the Aust94 genome. Despite the identification of isolates within this A.Br.013/015 node from Europe and the USA, this genetic group is not ecologically established in these regions and is thus not likely to be the source of the introduction into the Georgia/Turkey region. Rather, the presence of these isolates in Europe and the USA is likely to be due to the importation of contaminated animal products, possibly from the same geographic region responsible for the introduction into the Georgia/Turkey region. Additional phylogeography studies of these basal lineages are needed identify this source. To further resolve the A.Br.013/015 group to understand the Georgian population structure, we screened all members of this group (including the 272 Georgian isolates) across the twenty-six SNP assays leading out to the 52-G strain, resulting in the identification of eight new groups within this lineage. All Georgian isolates fell within one of the eight new groups ( Figure 1 , panel B)( Table S3 ); however, some isolates from Turkey and one from the USA were also placed within these groups. The single USA isolate is probably from a contaminated animal product imported from the Turkey/Georgia region. Most Turkish isolates included in this study are assigned to the basal node along the 52-G lineage. However, as we have no knowledge of the phylogenetic topology within this basal node, it is impossible to determine if this group was first introduced into Turkey and subsequently dispersed into Georgia or vice versa ; both scenarios are equally parsimonious. The presence of Turkish isolates in two more recent nodes preceded by exclusively Georgian nodes suggests at least two dispersal events from Georgia to Turkey ( Figure 1 , panel B). Further sampling from neighboring countries will provide more details on the geographic limits of this 52-G clade and the impact of national boundaries on limiting the dispersal of B. anthracis in the area. In any clade, members of the more ancient nodes have the greatest potential for widespread geographic dispersion. Indeed, along the lineage to the 52-G genome, members of the more basal nodes with multiple isolates from Georgia have been isolated across the country ( Figure 1 , panels B and C). However, without further sequencing of representative strains within each node, the phylogenetic topology cannot be known due to branch collapse [2] . Without phylogenetic knowledge, it is not possible to determine if phylogeographic clustering occurs at more recent evolutionary levels within these nodes. Conversely, isolates belonging to the clade after A.Br.030 which includes three nodes and the 52-G genome are only found in the southeast of Georgia ( Figure 1 , panel C), indicating more modern restrictions to dispersal. It is therefore likely that similar geographic structuring exists within the more basal nodes and would indicate that current dispersal of B. anthracis around Georgia is rare. Conclusions Our results are consistent with complex global dispersal patterns that have resulted in worldwide dispersal of the Aust94 group [6] – [8] . This work now provides additional resolution and detail within the Aust94 group and shows it to be highly geographically structured with a group of closely related isolates being largely restricted to the region in and around Georgia and Turkey. Even in Georgia there is evidence of geographic structuring within the one lineage we characterized in detail, suggesting that that although anthrax has been dispersed throughout the country, current dispersal may be limited. Supporting Information Table S1 Melt-MAMAs published in Birdsell et al 2012 and used in this study. (XLSX) Click here for additional data file. Table S2 Reference strains plus three new genomes used in this study and their NCBI accession numbers. (XLSX) Click here for additional data file. Table S3 Epidemiological and genotyping data for all B. anthracis strains included in this study. (XLSX) Click here for additional data file.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6042261/

A pragmatic harm reduction approach to manage a large outbreak of wound botulism in people who inject drugs, Scotland 2015

Background People who inject drugs (PWID) are at an increased risk of wound botulism, a potentially fatal acute paralytic illness. During the first 6 months of 2015, a large outbreak of wound botulism was confirmed among PWID in Scotland, which resulted in the largest outbreak in Europe to date. Methods A multidisciplinary Incident Management Team (IMT) was convened to conduct an outbreak investigation, which consisted of enhanced surveillance of cases in order to characterise risk factors and identify potential sources of infection. Results Between the 24th of December 2014 and the 30th of May 2015, a total of 40 cases were reported across six regions in Scotland. The majority of the cases were male, over 30 and residents in Glasgow. All epidemiological evidence suggested a contaminated batch of heroin or cutting agent as the source of the outbreak. There are significant challenges associated with managing an outbreak among PWID, given their vulnerability and complex addiction needs. Thus, a pragmatic harm reduction approach was adopted which focused on reducing the risk of infection for those who continued to inject and limited consequences for those who got infected. Conclusions The management of this outbreak highlighted the importance and need for pragmatic harm reduction interventions which support the addiction needs of PWID during an outbreak of spore-forming bacteria. Given the scale of this outbreak, the experimental learning gained during this and similar outbreaks involving spore-forming bacteria in the UK was collated into national guidance to improve the management and investigation of future outbreaks among PWID. Background People who inject drugs (PWID) are at an increased risk of wound botulism, a potentially fatal acute paralytic illness. During the first 6 months of 2015, a large outbreak of wound botulism was confirmed among PWID in Scotland, which resulted in the largest outbreak in Europe to date. Methods A multidisciplinary Incident Management Team (IMT) was convened to conduct an outbreak investigation, which consisted of enhanced surveillance of cases in order to characterise risk factors and identify potential sources of infection. Results Between the 24th of December 2014 and the 30th of May 2015, a total of 40 cases were reported across six regions in Scotland. The majority of the cases were male, over 30 and residents in Glasgow. All epidemiological evidence suggested a contaminated batch of heroin or cutting agent as the source of the outbreak. There are significant challenges associated with managing an outbreak among PWID, given their vulnerability and complex addiction needs. Thus, a pragmatic harm reduction approach was adopted which focused on reducing the risk of infection for those who continued to inject and limited consequences for those who got infected. Conclusions The management of this outbreak highlighted the importance and need for pragmatic harm reduction interventions which support the addiction needs of PWID during an outbreak of spore-forming bacteria. Given the scale of this outbreak, the experimental learning gained during this and similar outbreaks involving spore-forming bacteria in the UK was collated into national guidance to improve the management and investigation of future outbreaks among PWID. Background Wound botulism is a serious and potentially fatal acute paralytic illness caused by a highly potent neurotoxin produced by Clostridium botulinum [ 1 ]. Spores of C. botulinum are found in soil and can be introduced into illicit drugs such as heroin at any stage from point of production to use [ 1 , 2 ]. People who inject drugs (PWID) are at an increased risk of infection, particularly if they inject under the skin ("skin popping") or into muscle ("muscle popping") either deliberately or as a result of a "missed hit" (i.e. someone missing a vein when trying to inject intravenously). The necrotic devitalised tissue at the site of injection provides an ideal anaerobic environment for C. botulinum spores to germinate, multiply and produce toxin which then gets absorbed into the systemic circulation causing a neuroparalytic syndrome [ 3 – 7 ]. Although wound botulism in PWID have been noted in numerous countries across the world [ 4 , 8 – 12 ], the UK appears to be disproportionately affected with a number of outbreaks of wound botulism and other spore-forming bacteria such as Clostridium novyi , Bacillus anthracis and Clostridium tetani [ 5 , 13 – 17 ] . Between 2000 and 2014, there were a total of 170 reported cases of wound botulism among PWID in the UK. In 2015 alone, there were 47 reported cases, the largest outbreak of wound botulism seen among PWID in Europe, which was focused on Glasgow, Scotland [ 18 ]. Prior to the outbreak, the last probable (though unconfirmed) case in Scotland was reported in 2009. The outbreak began on Wednesday the 24th of December 2014, when NHS Greater Glasgow and Clyde (NHS GGC) health board Public Health Protection Unit (PHPU) were notified that a Glasgow resident in their thirties with a history of injecting drug use had been admitted to a hospital in Glasgow with neurological symptoms suggestive of botulism. Exactly 1 week later, NHS GGC were notified of a further suspected case, also in their thirties with a history of injecting drug use. The identification of two probable cases of wound botulism with a history of injecting drugs in a 1-week period suggested the contamination of a batch of heroin and/or cutting agent. In the ensuing weeks, further cases of suspected wound botulism were reported across six NHS health boards in Scotland. This report describes the epidemiological characteristics of what became the largest outbreak of wound botulism among PWID seen in Europe thus far [ 18 ], and the pragmatic harm reduction approach taken to limit morbidity and mortality associated with the infection during the outbreak. This manuscript aims to raise awareness of wound botulism and the importance and the need of harm reduction policies for the management of outbreaks among PWID. Methods Outbreak response Initially, the outbreak was managed by an Incident Management Team (IMT) based at NHS GGC. As the number of NHS health boards affected increased, a national IMT was convened and led by Health Protection Scotland (HPS). The multidisciplinary IMT consisted of representatives from NHS health boards, police (Police Scotland), non-governmental organisations (Scottish Drugs Forum), Public Health England (PHE), reference laboratories (Gastrointestinal Bacteria Reference Unit (GBRU)) and the Scottish Government. Epidemiologic investigation The IMT adapted the European Centre for Disease Control (ECDC) case definition of botulism for the outbreak, which included clinical, epidemiological and microbiological criteria (available from: http://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32012D0506&qid=1428573336660&from=EN#page=7 ) (Table 1 ). The case definition was categorised by the degree of certainty regarding the diagnosis as: Probable case —a high clinical suspicion and epidemiological evidence consistent with botulism Confirmed case —clinical, epidemiological and microbiological evidence consistent with botulism Table 1 Clostridium botulinum outbreak case definition Clinical Any person with at least one of bilateral cranial nerve impairment (e.g. diplopia, blurred vision, dysphagia, bulbar weakness) or peripheral symmetric paralysis. Epidemiological Use of illicit drugs by any route within the 2 weeks prior to onset of symptoms. Microbiological Isolation of Clostridium botulinum from infected wound and/or detection of botulinum toxin in a clinical specimen. Additionally, a "possible" case classification was used, when dealing with the initial report of suspected cases that merited further investigation. This provisional classification was given following review of the case by the Consultant Microbiologist of the GBRU and the treating physician. Cases were classified as "probable" or "confirmed" when further information became available. Botulism is legally notifiable under the Public Health Act (Scotland) 2008 (available from: http://www.gov.scot/Resource/0039/00398162.pdf ). For all cases identified, information on personal and medical details, source of drugs and drug use prior to the onset of illness, was collected using an established national enhanced surveillance form (available from: https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/577402/PWID_Botulism_Questionnaire_2006.1.6_14_December_2016.pdf ). Copies of the completed form were sent to HPS and collated on a copy of an Access database utilised by PHE for the UK surveillance of botulism associated with PWID. Laboratory investigations Microbiological confirmation of a clinical diagnosis of botulism requires the timely collection of appropriate clinical samples such as serum and tissue before administration of anti-toxin and commencement of antibiotics respectively. The early collection of samples following onset of symptoms maximises the opportunity for diagnosis. Confirmation was undertaken by either detection of toxin in serum by mouse bioassay (MBA), or by isolation of C lostridium botulinum from pus or wound tissue by culture and PCR detection of neurotoxin genes. Typing of C. botulinum isolates was performed by fAFLP as described previously [ 19 ]. Police investigation The public health investigation of this outbreak was supported by a parallel criminal investigation undertaken by Police Scotland. The objective of their investigation was to determine the source of the suspected contaminated heroin with a view to removing or reducing the amount of contaminated drugs in circulation, thereby maximising the safety of all individuals involved in the outbreak. Outbreak response Initially, the outbreak was managed by an Incident Management Team (IMT) based at NHS GGC. As the number of NHS health boards affected increased, a national IMT was convened and led by Health Protection Scotland (HPS). The multidisciplinary IMT consisted of representatives from NHS health boards, police (Police Scotland), non-governmental organisations (Scottish Drugs Forum), Public Health England (PHE), reference laboratories (Gastrointestinal Bacteria Reference Unit (GBRU)) and the Scottish Government. Epidemiologic investigation The IMT adapted the European Centre for Disease Control (ECDC) case definition of botulism for the outbreak, which included clinical, epidemiological and microbiological criteria (available from: http://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32012D0506&qid=1428573336660&from=EN#page=7 ) (Table 1 ). The case definition was categorised by the degree of certainty regarding the diagnosis as: Probable case —a high clinical suspicion and epidemiological evidence consistent with botulism Confirmed case —clinical, epidemiological and microbiological evidence consistent with botulism Table 1 Clostridium botulinum outbreak case definition Clinical Any person with at least one of bilateral cranial nerve impairment (e.g. diplopia, blurred vision, dysphagia, bulbar weakness) or peripheral symmetric paralysis. Epidemiological Use of illicit drugs by any route within the 2 weeks prior to onset of symptoms. Microbiological Isolation of Clostridium botulinum from infected wound and/or detection of botulinum toxin in a clinical specimen. Additionally, a "possible" case classification was used, when dealing with the initial report of suspected cases that merited further investigation. This provisional classification was given following review of the case by the Consultant Microbiologist of the GBRU and the treating physician. Cases were classified as "probable" or "confirmed" when further information became available. Botulism is legally notifiable under the Public Health Act (Scotland) 2008 (available from: http://www.gov.scot/Resource/0039/00398162.pdf ). For all cases identified, information on personal and medical details, source of drugs and drug use prior to the onset of illness, was collected using an established national enhanced surveillance form (available from: https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/577402/PWID_Botulism_Questionnaire_2006.1.6_14_December_2016.pdf ). Copies of the completed form were sent to HPS and collated on a copy of an Access database utilised by PHE for the UK surveillance of botulism associated with PWID. Laboratory investigations Microbiological confirmation of a clinical diagnosis of botulism requires the timely collection of appropriate clinical samples such as serum and tissue before administration of anti-toxin and commencement of antibiotics respectively. The early collection of samples following onset of symptoms maximises the opportunity for diagnosis. Confirmation was undertaken by either detection of toxin in serum by mouse bioassay (MBA), or by isolation of C lostridium botulinum from pus or wound tissue by culture and PCR detection of neurotoxin genes. Typing of C. botulinum isolates was performed by fAFLP as described previously [ 19 ]. Police investigation The public health investigation of this outbreak was supported by a parallel criminal investigation undertaken by Police Scotland. The objective of their investigation was to determine the source of the suspected contaminated heroin with a view to removing or reducing the amount of contaminated drugs in circulation, thereby maximising the safety of all individuals involved in the outbreak. Results Descriptive epidemiology A total of 47 individuals presented to hospital, between the 21st of December 2014 and the 29th of May 2015, with symptoms indicative of botulism. Seventeen of these were confirmed microbiologically and 23 cases were classed as probable (Fig. 1 ). Two cases were classified as (and remain) possible, with the remaining five cases discounted. The following analysis describes the 40 confirmed/probable cases only. Fig. 1 Date of hospital admission and case definitions of wound botulism infections among people who inject drugs, 21st of December 2014 to 29th of May 2015. N = 40 The majority of cases (98%) presented with classical symptoms of oculomotor and bulbar palsy but very few with descending limb weakness; over half (58%) were ventilated for respiratory paralysis although most did not require long-term support. All cases were promptly treated with antitoxin, 50% underwent wound debridement and all received antibiotics. There were four deaths; botulism was a contributory cause in two cases (case fatality rate (CFR) = 5%) [ 20 ]. Most of the cases were male (68%). The mean age among males and females was 44 and 38 years, respectively. All confirmed/probable cases were aged between 24 and 56 years old. The majority of cases were aged between 36 and 45 years old [ 20 ]. Cases were spread across the central belt of Scotland, with the majority residing in NHS GGC (63%), followed by NHS Lanarkshire (18%), NHS Forth Valley (15%), NHS Fife (2.5%) and NHS Ayrshire & Arran (2.5%) [ 20 ]. Risk information Detailed information on the history of drugs taken and route of administration was available for 34 of the 40 confirmed/probable cases. Where information was available, all cases reported using heroin, either alone (52%) or in combination with another drug. Of the 29 cases who reported their length of drug use, 50% had used drugs for over 10 years. Most of the cases (62%) reported using a combination of routes for administering their drugs in the month prior to illness (intravenous, skin/muscle popping, smoking/snorting) [ 20 ]. Microbiology A total of 17/40 (43%) cases were confirmed microbiologically: Clostridium botulinum type B was detected by PCR in wound pus or tissue from 13 patients, and botulism neurotoxin type B was detected in serum of 3 patients while in a fourth patient toxin detection was confirmed but not typed. For 2 other cases, the MBA result was unconfirmed due to insufficient serum to perform the neutralisation test. Molecular typing of the organism isolated from 11 patients gave an indistinguishable fAFLP profile indicating a common source of infection. All heroin seized by Police Scotland tested negative for Clostridium botulinum. Risk management and risk communication In the absence of intelligence on which (or how many) batches of heroin in circulation were potentially contaminated, all PWID were considered to be at risk of exposure and infection. Information from the initial cases on the source of their drugs used prior to symptoms suggested a focus on heroin obtained in or sourced via Glasgow city. Therefore, interventions to manage the incident were targeted in GGC and surrounding NHS boards. The IMT recognised that it was not a realistic expectation to be able to remove all contaminated drugs or cutting agents from the market, or to eliminate drug use among the susceptible cohort. The illicit consumption and supply of drugs mean that there is little intelligence on the source of contaminated drugs. Thus, a pragmatic harm reduction approach was adopted (Table 2 ). This approach, while recognising that the only way to eliminate risk of infection was to stop using drugs, also deployed interventions to reduce risk as far as possible for those who continued to inject and for those who were infected, reduce the risk of progression to serious illness. To raise awareness of the outbreak, communicate safer injection practices (for example, do not inject into the muscle or skin and promote smoking rather than injecting) and encourage early recognition of signs and symptoms, a postcard was created and made available to those at risk via all frontline and addiction services (mainly IEP services) across the NHS board areas involved (Fig. 2 ). The postcard was intended as an instrument for IEP staff to engage PWID in discussion, to raise awareness of botulism and safer injecting practices which reduce the risk of botulism. In addition, and in recognition that these services were key to identifying individuals at risk and filtering harm reduction messages, an information booklet, "Wound botulism and drug use: What workers need to know" composed by the Scottish Drug Forum (SDF) (available from: http://www.sdf.org.uk/wp-content/uploads/2017/03/Botulism_Booklet.pdf ), was developed and distributed to frontline staff with an opportunity to reinforce the learning through a workshop delivered by SDF. Furthermore, this ensured that staff were equipped to answer questions from PWID about the outbreak. Table 2 Hierarchy of objectives and advice/interventions Objective Advice/intervention Eliminate risk of infection • Advice on how to reduce or eliminate drug use altogether • Information on access to opiate substitution therapy services Reduce risk of infection • Advice on switching to safer route of drug use • Provision of foil to encourage smoking as an alternative to injecting • Advice on safer injecting behaviour (ensuring they inject into a vein) Reduce risk of severity • Education and awareness raising of the signs and symptoms of illness and where to seek help Fig. 2 Awareness-raising postcard distributed to people who inject drugs during the outbreak The IMT released three media statements and communicated regularly through the duration of the outbreak with GPs, hospitals and services for drug users to raise awareness of signs and symptoms, diagnostic procedures and how to obtain botulism antitoxin. Descriptive epidemiology A total of 47 individuals presented to hospital, between the 21st of December 2014 and the 29th of May 2015, with symptoms indicative of botulism. Seventeen of these were confirmed microbiologically and 23 cases were classed as probable (Fig. 1 ). Two cases were classified as (and remain) possible, with the remaining five cases discounted. The following analysis describes the 40 confirmed/probable cases only. Fig. 1 Date of hospital admission and case definitions of wound botulism infections among people who inject drugs, 21st of December 2014 to 29th of May 2015. N = 40 The majority of cases (98%) presented with classical symptoms of oculomotor and bulbar palsy but very few with descending limb weakness; over half (58%) were ventilated for respiratory paralysis although most did not require long-term support. All cases were promptly treated with antitoxin, 50% underwent wound debridement and all received antibiotics. There were four deaths; botulism was a contributory cause in two cases (case fatality rate (CFR) = 5%) [ 20 ]. Most of the cases were male (68%). The mean age among males and females was 44 and 38 years, respectively. All confirmed/probable cases were aged between 24 and 56 years old. The majority of cases were aged between 36 and 45 years old [ 20 ]. Cases were spread across the central belt of Scotland, with the majority residing in NHS GGC (63%), followed by NHS Lanarkshire (18%), NHS Forth Valley (15%), NHS Fife (2.5%) and NHS Ayrshire & Arran (2.5%) [ 20 ]. Risk information Detailed information on the history of drugs taken and route of administration was available for 34 of the 40 confirmed/probable cases. Where information was available, all cases reported using heroin, either alone (52%) or in combination with another drug. Of the 29 cases who reported their length of drug use, 50% had used drugs for over 10 years. Most of the cases (62%) reported using a combination of routes for administering their drugs in the month prior to illness (intravenous, skin/muscle popping, smoking/snorting) [ 20 ]. Microbiology A total of 17/40 (43%) cases were confirmed microbiologically: Clostridium botulinum type B was detected by PCR in wound pus or tissue from 13 patients, and botulism neurotoxin type B was detected in serum of 3 patients while in a fourth patient toxin detection was confirmed but not typed. For 2 other cases, the MBA result was unconfirmed due to insufficient serum to perform the neutralisation test. Molecular typing of the organism isolated from 11 patients gave an indistinguishable fAFLP profile indicating a common source of infection. All heroin seized by Police Scotland tested negative for Clostridium botulinum. Risk management and risk communication In the absence of intelligence on which (or how many) batches of heroin in circulation were potentially contaminated, all PWID were considered to be at risk of exposure and infection. Information from the initial cases on the source of their drugs used prior to symptoms suggested a focus on heroin obtained in or sourced via Glasgow city. Therefore, interventions to manage the incident were targeted in GGC and surrounding NHS boards. The IMT recognised that it was not a realistic expectation to be able to remove all contaminated drugs or cutting agents from the market, or to eliminate drug use among the susceptible cohort. The illicit consumption and supply of drugs mean that there is little intelligence on the source of contaminated drugs. Thus, a pragmatic harm reduction approach was adopted (Table 2 ). This approach, while recognising that the only way to eliminate risk of infection was to stop using drugs, also deployed interventions to reduce risk as far as possible for those who continued to inject and for those who were infected, reduce the risk of progression to serious illness. To raise awareness of the outbreak, communicate safer injection practices (for example, do not inject into the muscle or skin and promote smoking rather than injecting) and encourage early recognition of signs and symptoms, a postcard was created and made available to those at risk via all frontline and addiction services (mainly IEP services) across the NHS board areas involved (Fig. 2 ). The postcard was intended as an instrument for IEP staff to engage PWID in discussion, to raise awareness of botulism and safer injecting practices which reduce the risk of botulism. In addition, and in recognition that these services were key to identifying individuals at risk and filtering harm reduction messages, an information booklet, "Wound botulism and drug use: What workers need to know" composed by the Scottish Drug Forum (SDF) (available from: http://www.sdf.org.uk/wp-content/uploads/2017/03/Botulism_Booklet.pdf ), was developed and distributed to frontline staff with an opportunity to reinforce the learning through a workshop delivered by SDF. Furthermore, this ensured that staff were equipped to answer questions from PWID about the outbreak. Table 2 Hierarchy of objectives and advice/interventions Objective Advice/intervention Eliminate risk of infection • Advice on how to reduce or eliminate drug use altogether • Information on access to opiate substitution therapy services Reduce risk of infection • Advice on switching to safer route of drug use • Provision of foil to encourage smoking as an alternative to injecting • Advice on safer injecting behaviour (ensuring they inject into a vein) Reduce risk of severity • Education and awareness raising of the signs and symptoms of illness and where to seek help Fig. 2 Awareness-raising postcard distributed to people who inject drugs during the outbreak The IMT released three media statements and communicated regularly through the duration of the outbreak with GPs, hospitals and services for drug users to raise awareness of signs and symptoms, diagnostic procedures and how to obtain botulism antitoxin. Discussion A multi NHS board outbreak of Clostridium botulinum among PWID in Scotland resulted in 40 cases of botulism, representing the largest outbreak of wound botulism among PWID in Europe to date [ 18 ]. The epidemiology of the cases (predominantly male and over 30 years old) is representative of the PWID population in GGC [ 18 ]. Clostridium botulinum spores were not detected in samples of heroin which were obtained by Police Scotland, reflecting experiences from other outbreaks [ 8 , 21 , 22 ]. However, all epidemiological evidence suggested a contaminated batch of heroin or cutting agent as the source of the outbreak [ 20 ]. Given that the heroin in circulation in the UK is predominantly supplied through trafficking routes from Afghanistan, it is likely that a contamination event at some point higher up the distribution chain is the source of infections reported in the UK and Europe [ 2 , 17 ]. There are significant challenges associated with managing an outbreak of botulism among PWID. The only way to completely eliminate the risk of infection is to remove all contaminated drugs from the market, or to eliminate drug use among the susceptible population, neither of which are a realistic expectation. The recommendations and interventions to minimise the risk to PWID deployed during this outbreak were by necessity pragmatic, and focused on reducing the risk of infection for those who continued to inject drugs, and limiting the consequences for those who became infected. The overall strategy to manage the outbreak was predicated on augmenting the comprehensive harm reduction services currently deployed in Scotland [ 23 , 24 ], with enhanced interventions aimed at specifically reducing risk of exposure, e.g. the provision of foil to encourage smoking as an alternative to injecting, reviewing and expanding access to injecting equipment provision (IEP) and opiate substitution therapy (OST) services. While there is a lack of evidence that increasing injecting equipment provision reduces risk from bacterial infection, research suggests that it can attract and retain PWID in services [ 25 ] and therefore provide a vital window of opportunity to raise awareness of the outbreak and to filter harm reduction messages to those at ongoing risk. The communication strategy employed during the outbreak had two strands. The first strand involved communicating with health professionals. Many frontline healthcare workers (HCW) have limited practical experience of identifying cases of botulism, especially as signs of drug overdose or effects of multiple drugs can mask early signs of botulism [ 7 ]. As early diagnosis and treatment improves outcome and reduces risk of death [ 26 – 28 ], the IMT alerted frontline healthcare professionals to the presenting symptoms and provided guidance on the clinical management of botulism including how to obtain botulism antitoxin. The impact of ensuring awareness of the outbreak and quicker treatment initiation was reflected in shorter hospital stays as the outbreak continued and the majority of cases making a full recovery [ 29 ]. All information resources used during the outbreak were shared widely across Scotland. In addition, given the role Glasgow plays in the distribution of drugs throughout Scotland and the potential for contaminated heroin to be re-distributed across the country or elsewhere in the Europe, all resources developed to manage the incident were shared with colleagues across the country. Furthermore, an alert notice was posted through the European Union's (EU) Early Warning and Response System (EWRS) on 2 January 2015. The second strand included communicating with those at risk. Experience from previous outbreaks of spore-forming bacteria in PWID suggested that there was limited benefit of using conventional press or mass media as outlets for awareness-raising activities. Information was therefore targeted at those at risk through the development of a "postcard" early in the outbreak which contained information on risk reduction measures and signs and symptoms of infection (Fig. 2 ). Given the sudden onset of the outbreak and the need to respond quickly, peers were not involved in the development of the postcard but reviewed the postcard after it had been created. The impact of the postcard was evaluated via the Needle Exchange Surveillance Initiative [ 30 , 31 ] and qualitative interviews with persons who attended NHS GGC IEP services and had injected in the previous 6 months. The results suggested that the postcard may have improved awareness of the outbreak, the signs and symptoms of botulism and encouraged injecting behaviour change to reduce the risks of exposure, particularly when it was used as a prompt by frontline service providers for a focused discussion. The qualitative arm of the study revealed that the injection advice from the postcard which was most effective was to "ensure that they inject into a vein". Research suggests that PWID are grateful of pragmatic injecting advice [ 32 ] which recognises that PWID will continue to inject, despite risk of infection. Further research is needed, however, to establish the best mechanism for distribution and engagement with PWID, as many remained unaware of the risk they face during an outbreak [ 31 ]. Given the special role that frontline HCWs and IEP staff have as an access point to reach PWID at risk, their views and experiences of the response and communication of the outbreak should be explored to identify barriers and ensure best practice in future outbreaks. Improving communication and trust between these service providers and peers may improve the effectiveness of future outbreak alerts among PWID [ 33 ]. Research has highlighted that communication from peer-based social networks are an important source of information [ 34 ] and should be utilised more often. Thus, peer-delivered communication strategies and the involvement of peers and frontline staff in the development of interventions such as the postcard could be key to overcoming the communication barriers experienced during this outbreak. Officials in Glasgow have proposed the opening of a drug consumption room (DCR) and the establishment of a heroin-assisted treatment (HAT) service [ 35 ], in response to past outbreaks of spore-forming bacteria [ 13 – 15 ], high incidence of drug-related deaths [ 36 ] and an ongoing HIV outbreak [ 37 ]. The establishment of both these services may be an effective method of minimising the risk of another outbreak of spore-forming bacteria. HAT services would usually have stringent inclusion criteria [ 38 ]; however, these criteria could be relaxed during a severe outbreak. Furthermore, a DCR could promote safer injecting practices [ 39 ], potentially reducing the probability of soft tissue infections and therefore reducing the risk of wound botulism. Additionally, the presence of HCW in the DCR may help the detection and early recognition of spore-forming bacterial infections. Conclusion The management of this outbreak, the largest to date in Europe [ 18 ], highlights the importance and need for greater recognition of a pragmatic harm reduction approach that supports the addiction needs of those vulnerable to infection during an outbreak of spore-forming bacteria. The approach adopted took onboard significant experiential learning that had been gained from similar outbreaks of spore-forming bacterial infection in the UK since 2000 [ 5 , 13 – 17 ]. Alongside expert knowledge and published evidence where available, the experimental learning has been collated into a single guidance document published by the Scottish Health Protection Network "Guidelines for the public health management of tetanus, botulism or anthrax among people who use drugs" [ 40 ], to facilitate an efficient and effective response to future outbreaks. The response to this outbreak could have been further improved through deeper collaborations with peer and frontline staff, particularly by considering the use of peer-based social networks to filter harm reduction communications. Availability of data and materials The data supporting the conclusions of this article are included within the article. This article was based on the Health Protection Scotland Incident Management Team Report [ 17 ]: http://www.hps.scot.nhs.uk/resourcedocument.aspx?id=6002 . Ethics approval and consent to participate Not applicable Consent for publication Not applicable Competing interests The authors declare that they have no competing interests. Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4639457/

Advax™, a novel microcrystalline polysaccharide particle engineered from delta inulin, provides robust adjuvant potency together with tolerability and safety

There is an ongoing need for new adjuvants to facilitate development of vaccines against HIV, tuberculosis, malaria and cancer, amongst many others. Unfortunately, the most potent adjuvants are often associated with toxicity and safety issues. Inulin, a plant-derived polysaccharide, has no immunological activity in its native soluble form but when crystallized into a stable microcrystalline particulate from (delta inulin) acquires potent adjuvant activity. Delta inulin has been shown to enhance humoral and cellular immune responses against a broad range of co-administered viral, bacterial, parasitic and toxin antigens. Inulin normally crystallizes as large heterogeneous particles with a broad size distribution and variable solubility temperatures. To ensure reproducible delta inulin particles with a consistent size distribution and temperature of solubility, a current Good Manufacturing Practice (cGMP) process was designed to produce Advax™ adjuvant. In its cCMP form, Advax™ adjuvant has proved successful in human trials of vaccines against seasonal and pandemic influenza, hepatitis B and insect sting anaphylaxis, enhancing antibody and T-cell responses while being safe and well tolerated. Advax™ adjuvant represents a novel human adjuvant that enhances both humoral and cellular immunity. This review describes the discovery and development of Advax™ adjuvant and research into its unique mechanism of action. 1 Inulin: historical background Inulin is a simple plant-based fructan polysaccharide comprising a family of linear β(2 → 1) polyfructofuranosyl α- d -glucose polymer chains in which an unbranched chain of fructose rings is terminated with a single glucose ( Fig. 1 ). Inulin is utilized as a storage carbohydrate by plants of the Compositae family that include dahlias, chicory, artichoke, onions, and garlic (reviewed in [1] ). Inulin's medicinal uses date back to ancient times. Pedanios Dioscoride, a physician with the Roman army, in 100 AD reported the beneficial effects of chicory root extract (∼40% inulin by weight) for treatment of stomach, liver and kidney complaints [2] . In 1804, inulin was extracted by a German scientist from a boiling water extract of Inula helenium [3] . Based on its Inula source, Thomson in 1818 gave this extract the name, inulin [4] . Oral inulin's beneficial effects on metabolic disorders was re-discovered in 1874 with a report that diabetic patients lost their glycosuria when put on a diet of 100 g of inulin per day [3] . Dietary inulin not only has a favourable effect on blood sugar levels in patients with type 2 diabetes but also reduces fasting insulin concentrations, triacylglycerol levels [5] and hepatic lipogenesis [6] . Inulin has also been shown to reduce atherosclerotic lesions in Apo E-deficient mice [7] . Furthermore, short chain fatty acids including butyrate and propionate formed by inulin fermentation in the gut have anti-proliferative effects [8] and stimulate apoptosis of colon cancer cells, enhancing expression of enzymes including histone deacetylases involved in detoxification of carcinogens [9] . Inulin fermentation products in the gut also have anti-inflammatory effects and may help suppress autoimmune disease [10] . Fig. 1 Structure of a single inulin polymer. This shows inulin comprises a variable-length chain of fructose rings terminated with a single glucose ring. In general, inulin polymers isolated from plant sources range in length from 3 to 100 fructose units. 2 Use of inulin for measurement of human glomerular filtration Although inulin can be metabolized by gut flora, the inability of mammals to metabolize injected inulin led Shannon and Smith in the 1920s to explore inulin's use for measurement of glomerular filtration (GFR) [11] . They confirmed the nontoxicity of inulin by intravenously injecting themselves with 160 g of soluble inulin thereby confirming inulin's exceptional safety [11] . Injection of soluble inulin subsequently became the gold standard for GFR measurement, with intravenous injection of a loading dose of 50 mg/kg body weight followed by an infusion to maintain steady plasma levels. Measurement of urinary inulin then provides an estimate of GFR. Many thousands of patients have safely had this GFR test including pregnant women and newborn babies, with no known adverse effects attributable to inulin other than a mild osmotic diuresis [12] . In the early days hospitals made their own inulin formulations and patients receiving inulin infusions occasionally developed transient hypotensive symptoms. This was found to reflect complement activation that was independent of the only then known (now called classical) complement pathway. A search for the mechanism of complement activation by inulin led to discovery of the alternative complement pathway (ACP) [13] . ACP activation was found to occur when injected inulin was not entirely dissolved with microscopic crystalline inulin contaminants responsible for ACP activation. Hence ACP activation is a unique feature of crystalline but not soluble inulin, a feature critical to inulin's use as a vaccine adjuvant as described below. 3 Development of gamma inulin as an anti-cancer agent Complement activation has anti-cancer effects in animals and humans. An early demonstration of this anti-cancer effect utilized the complement activator, Staphylococcus aureus protein A Unfortunately, this also had major side effects including pyrexia, nausea, vomiting, and cardiopulmonary toxicity [14] . To avoid the problems of using S. aureus for complement activation, an alternative approach to cancer therapy was developed by Cooper et al. with the idea of instead using inulin crystals for complement activation [15] . This led to the development of a crystalline isoform of inulin given the name gamma inulin to distinguish it from highly soluble alpha and beta inulin isoforms [16] . Gamma inulin was potent at ACP activation in human plasma reaching a maximum at ∼20 μg/ml [17] . In mice, intraperitoneal administration of gamma inulin at doses of 0.2–0.5 mg induced greater than 50% serum complement depletion within 30 min with complement levels returning to normal within 16–24h [18] . Despite potent ACP activation, intraperitoneal or subcutaneous injections of gamma inulin at doses up to 10 mg were well tolerated with mice not exhibiting any adverse effects. Intraperitoneal administration of gamma inulin to mice implanted with a B16 melanoma cell line prolonged survival times and depletion of complement factor 3 (C3) abrogated this protective effect, consistent with gamma inulin's anti-tumour effects being mediated by ACP activation [19] . Gamma inulin induced tumour regression when injected into squamous cell carcinomas in sheep, an effect enhanced by combination therapy with cyclophosphamide [18] . Intralesional injection of gamma inulin into equine sarcoids or spontaneous tumours in dogs similarly resulted in tumour regression (P. Cooper, personal communication). Gamma inulin also potentiated anti-cancer effects of photodynamic therapy in an ACP-dependent manner [20] . 4 Development of microcrystalline inulin as an immune adjuvant Complement also plays an important role in generation of adaptive immunity. This effect of complement has been exploited in vaccine design. For example, covalent coupling of antigens to complement factor C3d was shown to enhance their immunogenicity [21] . Given its potent ability to activate complement, we asked whether co-formulation of antigens with gamma inulin might similarly enhance their immunogenicity. When tested in this way, gamma inulin proved to have a beneficial effect on adaptive immune responses. Many of the early studies were performed with keyhole limpet hemocyanin (KLH) and when KLH was co-administered with gamma inulin this resulted in enhanced anti-KLH antibody responses in mice with similar results in guinea pigs and rabbits [22] . Gamma inulin was subsequently shown to improve the immunogenicity of a range of vaccine antigens (reviewed in [23] ). The mechanism of adjuvant action of gamma inulin was shown to involve increased C3 deposition on the surface of macrophages leading to enhanced T-cell activation [24] . Although it did not induce macrophage activation as measured by chemiluminescent respiratory burst, gamma inulin primed macrophages for a greater respiratory burst in response to phorbol ester. Incubation of human peripheral blood mononuclear cells with tetanus toxoid in the presence of gamma inulin resulted in enhanced IL-2 secretion when compared to toxoid alone, consistent with augmented antigen processing and presentation [18] . Cobra venom depletion of serum complement in mice prior to immunization blocked gamma inulin's adjuvant effects, consistent with adjuvant activity being dependent upon ACP activation [24] . In an attempt to enhance its adjuvant potency gamma inulin was co-crystallized with aluminium hydroxide to form Algammulin, resulting in further enhancement of antibody responses [17] . Nevertheless, Algammulin had significant drawbacks that prevented its development as a human vaccine adjuvant including potential toxicity of the aluminium adjuvant component [25] , [26] , Th2-immune bias [27] , [28] , and high batch-to-batch variability in potency. To overcome these problems, studies were undertaken into the structure and function of inulin with the aim of developing a more stable and consistent inulin adjuvant platform. This led to development of the delta inulin isoform that not only had much greater and more consistent adjuvant activity than gamma inulin but, most importantly, was able to be manufactured as a highly consistent cGMP product [29] . Delta inulin subsequently formed the basis of the Advax™ adjuvant platform. 5 Structural basis of delta inulin Inulin has a hydrophobic polyoxyethylene-like backbone that is critical to its structure in solution and when crystallized. Inulin particles obtained by precipitation from cold water are referred to as alpha inulin and the particles obtained by precipitation from ethanol as beta inulin; dried commercial inulin is almost all in alpha or beta forms. Both alpha and beta isoforms are highly soluble and readily dissolved in water at 25 °C, gamma inulin starts to become soluble in water at 37 °C, and delta inulin is insoluble at temperatures below 40 °C, an important distinction as this means that delta inulin particles are insoluble at body temperatures. Hence, the key factor behind understanding of delta inulin is the fact that inulin polymers can be crystallized into discrete isoforms characterized by different solubility rates in aqueous media [29] , [30] . In essence, each inulin particle is made up of individual inulin polymers each formed into an antiparallel double helix, with these helixes then being aggregated together through lateral hydrogen bonding to form lamellar sheets [29] , [31] , [32] . The greater stability of delta inulin when compared to earlier inulin isoforms reflects greater aggregate strength of H-bonding between each of the individual inulin polymers forming the crystal structure [30] . Transmission and scanning electron microscopy and atomic force microscopy studies of delta inulin particles show consistent spherulite-like discoid particles 1–2 μm in diameter made up of a series of lamellar sheets [33] . Better knowledge of the structural basis of delta inulin crystallization assisted design of a cGMP production process, thereby ensuring ability to produce Advax™ adjuvant particles with a highly consistent size and reproducible adjuvant properties, as shown in Fig. 2 . This was in marked contrast to the highly irregular shape, size and inconsistent adjuvant properties of gamma inulin and initial non-GMP delta inulin formulations. Fig. 2 Microscopic structures of crystalline inulin particles. Transmission electron microscopy (TEM) reveals gamma inulin particles to have highly variable shape, size and structure (A), potentially explaining their large batch-to-batch variation in adjuvant activity. Initial delta inulin particles had better defined structure on TEM but still exhibited considerable size variation (B). Optimization and control of crystallization parameters allowed reproducible production of delta inulin particles of consistent size and shape (C) that subsequently formed the basis of Advax™ adjuvant. Freeze-fracture scanning EM of Advax™ particles suspended in water reveals their discoid shape and formation from lamellar sheets (D and E). Atomic force microscopy of Advax™ particles during their formation confirms their discoid character and also reveals fine filament precursors from which the final delta inulin particles are assembled (F). 6 Alternative complement pathway activation Gamma and delta inulin both activate the ACP [29] . It is not known exactly why these inulin structures are such potent complement activators, although they do have a high number of surface hydroxyl groups recognized as important to complement activation. Virtually any biological surface can activate the ACP provided no sialic acid moieties are present with sialic acid functioning as a discriminator to protect vertebrate surfaces from self-complement attack [34] . Complement factor C3b has only a moderate preference for spontaneous reaction with glucose or fructose moieties [35] . Hence, inulin polymers themselves have no particular features to suggest high C3 convertase activation activity. Indeed, soluble inulin polymers have no ability to activate complement. For ACP initiation to lead to the rapid amplification of complement activation, the labile C3 convertase products produced by ACP activation must be localized on a surface site relatively protected from circulating complement inhibitors and degrading enzymes [35] . Therefore, it is possible that the surface of gamma and delta inulin crystal packing structures is able to protect activated complement components. Despite potent ACP activation capability, both gamma and delta inulin are extremely well tolerated even when injected at high doses. This suggests that the complement activation they induce occurs in a controlled fashion that avoids production of toxic downstream mediators. We hypothesize that delta inulin may activate complement regulators such as clusterin thereby blocking production of downstream anaphylatoxins and complement membrane attack complexes [36] . If so, this could explain how delta inulin is able to separate the beneficial effects on adaptive immunity from the harmful downstream effects of complement membrane attack complexes. 7 Preclinical vaccine studies Delta inulin has multiple advantages as an adjuvant over gamma inulin including greater temperature stability, ability to be produced as a monodisperse particle population under cGMP ( Fig. 2 ) and improved adjuvant potency in animal models. Delta inulin is also resistant to damage from sterilizing doses of gamma irradiation with gamma inulin being highly susceptible [37] . Consequently, development of gamma inulin was ceased more than decade ago and efforts instead refocused on delta inulin, in the form of Advax™ adjuvant. Whilst Advax™ adjuvant is comprised of delta inulin, the terms Advax™ and delta inulin are not synonymous, as Advax™ adjuvant comprises delta inulin of highly specific particle size and morphology, thereby distinguishing it from delta inulin more generally. Early studies showed that immunization of mice with hepatitis B surface antigen (HBsAg) plus Advax™ adjuvant provided at least 4-fold antigen-dose sparing when compared to a standard alum-adjuvanted vaccine [29] . The Advax™-immunized mice also showed a more balanced Th1 and Th2 immune response by contrast to the extreme Th2-bais seen in mice immunized with alum adjuvant. Use of Advax™ adjuvant was associated with increased CD4+ and CD8+ T-cell proliferative responses and increased antigen-stimulated production of IFNγ, TNFα, IL-4, IL-5, IL-6, IL-10, IL-13, IL-17 and GM-CSF [38] , consistent with broad based enhancement of all T-cell subsets, rather than the strong Th2- or Th1-bias of alum or CpG adjuvants, respectively [39] . IL-1α production was not increased by Advax™ adjuvant [38] suggesting that, unlike alum [40] , Advax™ does not induce inflammasome activation. Its adjuvant effect was not dependent on antigen absorption as antibody responses were enhanced even when Advax™ was injected 1 day prior to the antigen [38] . Advax's effects on influenza vaccines have been extensively studied. Formulation of inactivated influenza vaccine with Advax™ enhanced serum neutralizing antibody titres in mice and provided up to 100-fold antigen sparing [41] . This correlated with a higher frequency of influenza-specific IgM- and IgG-secreting B cells in the bone marrow and spleen of immunized mice. As seen for HBsAg immunization, use of Advax™ was associated with an increase in influenza-specific CD4+ and CD8+ T-cell proliferation and increased IFN-γ, IL-2, IL-5, IL-6, and GM-CSF production, which translated into enhanced protection against lethal influenza challenge [41] . Immunity was long-lived with high antibody titres, T-cell responses and influenza protection still evident 1 year post-immunization with Advax™. Injection of mice with Advax™ alone in the absence of antigen provided no protection against influenza infection, indicating that it does not induce non-specific innate immune protection. Advax™ has also been shown to boost the immunogenicity of inactivated influenza vaccine even when delivered via the intrapulmonary route [42] . Advax™ adjuvant was also tested for its ability to help overcome the effects of pregnancy-associated immune suppression on influenza vaccine responses. Advax™ adjuvant was well tolerated by pregnant dams and was not associated with any adverse reproductive or developmental effects in mothers or pups [43] . Pregnant dams that received a single intramuscular injection of inactivated influenza antigen with Advax™ adjuvant had increased serum anti-influenza IgG titres and this translated into higher anti-influenza IgG titres in their breast milk. This in turn resulted in higher anti-influenza IgG titres in the serum of their suckling pups and these pups were completely protected when challenged with influenza at 4 weeks-of-age. By contrast, no survival was seen in pups of mothers immunized with a standard unadjuvanted influenza vaccine [43] . In a ferret challenge model Advax™ adjuvant enhanced protection conferred by a licensed vaccine against high-pathogenicity avian (H5N1) influenza. Two immunizations with inactivated H5N1 antigen plus Advax™ adjuvant with or without the addition of CpG oligonucleotide induced high serum H5N1 neutralizing antibody titres whereas antibody was barely detectable in ferrets that received standard vaccine [44] . Enhanced survival, reduced clinical disease and absence of brain invasion by virus was seen in the Advax™-adjuvanted groups with the virus being completely cleared from the nasal wash by day 4 post-challenge. Advax™ adjuvant was similarly shown to enhance vaccine protection against pulmonary anthrax. Advax™ enhanced antibody responses to recombinant protective antigen (rPA) resulting in enhanced protection against a lethal anthrax challenge even after just a single immunization [45] . A synergistic effect was seen when murabutide, a NOD2 agonist, was co-formulated with the Advax™ adjuvant. As shown by in vivo imaging of cathepsin cleavage of ProSense 750, Advax™ plus murabutide induced significantly less local injection site inflammation than the alum adjuvant. Flaviviruses represent another major vaccine challenge. Although there are human vaccines against Japanese encephalitis virus (JEV) [46] , there are currently no licensed vaccines against West Nile virus or dengue. Inclusion of Advax™ adjuvant in an inactivated JEV antigen not only enhanced neutralizing antibody responses in immunized mice and horses but was also associated with induction of cross-neutralizing antibody responses against Murray Valley encephalitis virus (MVEV) and West Nile virus (WNV) [47] . It was subsequently shown that the Advax™ adjuvant enhanced production of memory B cells able to transfer protection against JEV and related flaviviruses to naïve mice [48] . Cross-protection against WNV was achieved after just a single dose of JEV vaccine formulated with Advax™ adjuvant and was maintained out to at least 1 year post-immunization [49] . Advax™-adjuvanted JEV vaccine was found to be safe and effective when administered to pregnant horses and newborn foals in which it similarly induced broadly cross-protective anti-flavivirus antibodies [50] . Advax™ adjuvant enhanced neutralizing antibody responses and protection in a murine model of severe acute respiratory syndrome (SARS) coronavirus. Formulations of either inactivated SARS virus or recombinant spike protein with Advax™ adjuvant protected mice and reduced lung virus titres when compared to immunization with either antigen alone [51] . One year post-immunization, animals immunized with Advax™-adjuvanted vaccine had higher SARS-specific CD4+ and CD8+ T-cell proliferation responses and IFNγ, IL2, IL4, IL6, IL10, IL17 and TNF α production. Whereas alum-formulated SARS antigens were associated with increased eosinophilic lung immunopathology in response to challenge with SARS virus, lung immunopathology was absent in mice immunized with delta inulin-adjuvanted vaccine. The absence of lung immunopathology in the Advax™ adjuvant groups correlated with a higher frequency of SARS-specific IFNγ-producing T cells, consistent with Advax™ adjuvant driving a more balanced Th1 and Th2 response. Inhibition of eosinophilic lung immunopathology was further assisted by co-formulation of Advax™ adjuvant with CpG oligonucleotide [51] . Advax™ adjuvant alone or in combination with CpG oligonucleotide enhanced the immunogenicity in mice of an HIV envelope protein boost following a DNA plasmid intramuscular prime, thereby inducing persistent gp120-specific mucosal IgA, serum IgG and memory T- and B-cell responses [52] . Similarly, in rabbits and nonhuman primates Advax™ adjuvant enhanced production of simian immunodeficiency virus cross-neutralizing antibodies when combined with an envelope protein boost following a DNA prime [53] . Advax™ adjuvant enhanced protection afforded by a Listeria vaccine based on T-cell peptide epitopes conjugated to gold nanoparticles [54] , [55] . Enhanced protection by the Advax™ adjuvant correlated with an increased frequency of splenic CD4+ and CD8+ T cells, NK cells, CD8α+ DC and Th1 cytokine production (IL12, IFNγ, TNFα and MCP1) and was associated with increased T-cell epitope spreading following live Listeria challenge [56] . Examples of veterinary vaccines where Advax™ adjuvant has been shown to have beneficial effects on neutralizing antibody responses and/or protection include vaccines against Peste de petit ruminants in mice and goats [57] and against African Horse Sickness and Melioidosis in camels [58] . 8 Administration routes, species-specificity and mechanism of action In most of the above-described preclinical studies, Advax™ adjuvant was administered intramuscularly or subcutaneously to mimic its intended human use. However, Advax™ adjuvant is also effective if given via other routes, including intradermal, intraperitoneal or intranasal delivery (unpublished data). It was recently shown to have adjuvant activity even when administered directly into the murine lung. Intratracheal administration of a dry powder inactivated influenza vaccine together with Advax™ adjuvant induced higher influenza-specific intranasal IgA titres and a more balanced Th1/Th2 antibody isotype response with no observed adverse effects [42] . Whilst the vast majority of immunogenicity studies of Advax™ adjuvant have been undertaken in mice, adjuvant activity has been demonstrated across a broad range of other animal species including guinea pigs [38] , ferrets [44] , goats [57] , horses [47] and camels [58] . To date, no species limitation in Advax™ adjuvant action has been seen with adjuvant activity also seen in rats, rabbits, dogs, cats and sheep (unpublished data). It is also effective across different life stages, being efficacious and safe when administered in combination with inactivated influenza vaccine to pregnant mice [43] or 7-day-old mouse pups [59] or in combination with inactivated Japanese encephalitis vaccine to pregnant mares or newborn foals [50] . Most adjuvants have been shown to work via danger signals. For example, aluminium salts inducing cell necrosis resulting in inflammasome activation [60] and also triggering toll-like receptor (TLR)-9 activation through DNA released from lysed cells [61] . Similarly, monophosphoryl lipid A activates TLR4, flagellin activates TLR5 and CpG oligonucleotides activate TLR9. The final common pathways for all these adjuvants working via danger signals is activation of the transcription factor, nuclear factor-kappa B (NFkB), the master regulator of inflammatory responses [62] . Advax™ adjuvant is an exception as it fails to induce NFkB activation in human monocytes or murine splenocytes and similarly does not activate the inflammasome or induce inflammatory cytokine production (unpublished data). The mechanism underlying the unique ability of Advax™ adjuvant to separate innate immune activation and inflammation from enhancement of adaptive immunity is the subject of intense ongoing study by our group. 9 Clinical development A first-in-man Phase 1 clinical trial was conducted to asses the safety and tolerability of three intramuscular doses of HBsAg formulated with Advax™ adjuvant in healthy adult subjects was undertaken following successful preclinical acute and repeat dose toxicology studies in which no safety problems were identified [63] . Advax™ adjuvant was well tolerated by the trial subjects with injection site pain scores not significantly different to subjects receiving HBsAg alone. Advax™ enhanced anti-HBsAg antibody titres and seroprotection rates when compared to administration of HBsAg alone. The proportion of subjects with positive anti-HBsAg CD4+ T-cell responses was also significantly higher in subjects that received Advax™ adjuvant [63] . A Phase 1/2 study was undertaken in adult subjects to assess the ability of Advax™ adjuvant to enhance the immunogenicity of a pandemic influenza A/H1N1/2009pdm vaccine made from recombinant hemagglutinin [64] . Advax™ increased seroprotection rates by 1.9 times after the first, and 2.5 times after the second, immunization, when compared to immunization with the recombinant hemagglutinin alone. Advax™ adjuvant was well tolerated with no adjuvant-associated adverse reactions observed. Advax™ adjuvant has also been found to be safe and effective when combined with a bee venom-based immunotherapy administered to human subjects with bee-sting anaphylaxis. It induced more rapid and higher titres of venom-specific IgG4, a marker of successful immunotherapy [65] . Other clinical trials where Advax™ adjuvant has been successfully tested include in combination with seasonal trivalent inactivated influenza vaccines and in combination with a universal T-cell vaccine against influenza based on synthetic peptide epitopes (unpublished data). 10 Conclusions and future prospects Advax™ delta inulin adjuvant successfully enhanced vaccine immunogenicity across a broad range of antigen types including whole inactivated viruses, recombinant proteins, synthetic peptides, toxins and venoms. It has been shown effective across all animal species tested, and during pregnancy and early neonatal life. It consistently enhances serum antibody titres and CD4+ and CD8+ T-cell immunity and acts synergistically with traditional innate immune activators such as murabutide or CpG oligonucleotide. Early promise in preclinical studies has been supported by successful human clinical trials confirming its safety, tolerability and efficacy. Ongoing priorities are to further characterize delta inulin's mechanism of action, develop it as a mucosal adjuvant, explore synergistic combinations with other immune activators and extend human clinical studies to paediatric populations.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6587716/

Omadacycline Enters the Ring: A New Antimicrobial Contender

Omadacycline is a novel aminomethylcycline approved for the treatment of community‐acquired bacterial pneumonia and acute bacterial skin and skin structure infections. This article reviews existing data pertaining to the biochemistry, mechanism of action, pharmacokinetics/pharmacodynamics, in vitro activity, and current progress with omadacycline in clinical trials. Omadacycline inhibits protein synthesis by binding to the 30S subunit of the bacterial ribosome at the tetracycline‐binding site with an affinity similar to glycylcyclines. It is able to bypass older tetracycline resistance mechanisms and demonstrates activity against bacterial strains that are tetracycline resistant. In addition, omadacycline displays broad‐spectrum activity against gram‐positive organisms (including methicillin‐resistant Staphylococcus aureus and vancomycin‐resistant enterococci), gram‐negative organisms, atypical organisms, and anaerobes. It has been evaluated against infections in adults both intravenously and orally. Dosage adjustments are not required for patients with renal impairment. Omadacycline displays a comparable efficacy and safety profile to standard‐of‐care agents, with the most common side effects observed being gastrointestinal. Currently available data for omadacycline suggest that this is a promising agent added to our antimicrobial armamentarium. Data Sources A literature search was performed using the PubMed electronic database with the following search terms: omadacycline and PTK 0796. A secondary search for relevant clinical trials was completed using Google Scholar and ClinicalTrials.gov. Additional information was gathered from abstracts from Infectious Diseases Week 2017, the European Congress of Clinical Microbiology and Infectious Diseases conference from 2018, and the Paratek Pharmaceuticals website. Chemistry, Structure, and Function Omadacycline is a semisynthetic aminomethylcycline derived from minocycline (Figure 1 ). It has a novel modification with an aminomethyl group present at the C9 position of the basic tetracycline structure. This modification allows omadacycline to overcome bacterial resistance mechanisms commonly used against doxycycline and minocycline, including tetracycline efflux and ribosomal protection. 4 In addition, the structure modification increases antimicrobial potency, as well as limits unwanted side effects, such as nausea and emesis, commonly seen with tigecycline. 5 , 6 Figure 1 (a) Chemical structure of omadacycline. (b) Chemical structure of minocycline. Like other tetracyclines, omadacycline inhibits protein synthesis of bacteria without having a significant impact on synthesis of DNA, RNA, or peptidoglycan. It binds to the 30S subunit of the bacterial ribosome at the tetracycline‐binding site with an affinity similar to glycylcyclines. 6 , 7 Microbiologic Activity Similar to other tetracyclines, omadacycline is a broad‐spectrum antimicrobial with activity against aerobic and anaerobic gram‐positive and gram‐negative bacteria, as well as atypical bacteria (Table 1 ). In addition, omadacycline demonstrates activity against organisms with multidrug resistance, including tetracycline resistance. Breakpoints of ≤4, ≤0.5mg/L, ≤0.25mg/L, and ≤0.25mg/L have been defined for Enterobacteriaceae, Staphylococcus aureus, Enterococcus faecalis , and Streptococcus species , respectively. Table 1 In Vitro Susceptibility of Various Organisms to Omadacycline and Comparator Agents Organism (# of Isolates) Omadacycline Doxycycline Minocycline Tetracycline MIC 50 (mg/L) MIC 90 (mg/L) MIC Range (mg/L) MIC 50 (mg/L) MIC 90 (mg/L) MIC Range (mg/L) MIC 50 (mg/L) MIC 90 (mg/L) MIC Range (mg/L) MIC 50 (mg/L) MIC 90 (mg/L) MIC Range (mg/L) Gram‐positive organisms MRSA (n=39) 4 0.25 0.5 0.125–1 0.125 8 ≤ 0.06–8 0.25 8 ≤ 0.06–16 0.25 64 ≤ 0.06–64 MRSA (n=1438) 8 0.12 0.25 0.03–8 – – – – – – – – – S. pneumoniae (n=41) 4 ≤ 0.06 0.125 ≤ 0.06–0.25 2 – ≤ 0.06–4 2 8 ≤ 0.06–8 16 32 ≤ 0.06–64 E. faecalis (n=31) 4 0.25 0.5 0.125–1 4 16 ≤ 0.06–16 8 16 0.125–16 32 64 0.125– > 64 E. faecium (n=24) 4 0.25 0.5 0.125–0.5 2 8 ≤ 0.06–16 8 16 0.125–32 32 64 0.125– > 64 E. faecium (n=390) 8 0.06 0.12 ≤ 0.015–8 – – – – – – – – – Gram‐negative organisms H. influenzae (n=53) 4 1 2 0.5–8 0.5 4 0.125–8 – – – 2 32 0.125–64 M. catarrhalis (n=408) 8 0.25 0.25 0.06–0.5 – – – – – – – – – K. pneumoniae (n=14) 4 2 4 1–8 2 32 1–64 2 64 2– > 64 2 > 64 0.5– > 64 K. pneumoniae (n=1771) 8 2 8 0.25– > 32 – – – – – – – – – Enterobacter cloacae species complex (n=572) 8 2 4 0.25– > 32 – – – – – – – – – A. baumannii (n=441) 8 4 8 0.06– > 32 – – – – – – – – – S. maltophilia (n=315) 8 2 8 0.25– > 32 – – – – – – – – – Atypical organisms Mycoplasma pneumoniae (n=20) 9 0.125 0.25 0.125–0.25 0.25 0.5 0.125–0.5 – – – 0.5 0.5 0.25–0.5 Ureaplasma spp. (n=20) 9 1 2 0.25–2 0.25 4 0.06–4 – – – 1 16 0.25–16 MIC = minimum inhibitory concentration; MIC 50 = minimum concentration of antibiotic that inhibits 50% of the isolates; MIC 90 = minimum concentration of antibiotic that inhibits 90% of the isolates; MRSA = methicillin‐resistant Staphylococcus aureus . John Wiley & Sons, Ltd Gram‐Positive Aerobic Coverage Omadacycline provides activity against many gram‐positive organisms. In an evaluation of 339 gram‐positive isolates, omadacycline activity was compared to that of tetracycline, minocycline, doxycycline, vancomycin, and other agents with clinically relevant gram‐positive coverage. 4 Against S. aureus isolates, including methicillin‐susceptible, methicillin‐resistant, and multidrug‐resistant strains, omadacycline MIC at which 90% of isolates were inhibited (MIC 90 ) values were 0.5 mg/L. In S. aureus strains possessing tetracycline resistance, omadacycline produced MICs ranging from 0.125 to 1 mg/L. In both Enterococcus faecalis and E. faecium strains, including those resistant to vancomycin or tetracycline, omadacycline displayed activity with an MIC 90 value of 0.5 mg/L. All streptococcal strains were inhibited by omadacycline at concentrations of 0.5 mg/L. Similarly, in a global surveillance study including roughly 70,000 isolates, 99.9% of all S. aureus and Enterococcus spp. were inhibited by omadacycline concentrations less than or equal to 2 mg/L. 10 Gram‐Negative Aerobic Coverage Gram‐negative coverage for omadacycline includes many Enterobacteriaceae. Against clinical pathogens, omadacycline was compared to standard‐of‐care agents. For E. coli and Klebsiella spp., respectively, MIC 90 values of 2 and 4 mg/L were observed. Omadacycline also displays activity against Haemophilus influenzae and Moraxella catarrhalis . 8 , 11 In 3383 H. influenzae isolates, including 736 β‐lactamase positive, 99% of the organisms were inhibited by omadacycline concentrations of less than or equal to 2 mg/L. Against M. catarrhalis , 100% of the 1126 isolates were inhibited by concentrations of less than or equal to 1 mg/L. Omadacycline also demonstrates antimicrobial inhibition against multidrug‐resistant organisms. Activity against Acinetobacter baumannii‐Acinetobacter calcoaceticus spp. complex (n=2101) and other Acinetobacter spp. (n=292) was assessed, and omadacycline inhibited growth of 91.5% and 95.5% of the isolates at less than or equal to 4 mg/L, respectively. 10 Against 315 Stenotrophomonas maltophilia isolates, omadacycline inhibited growth of 82.2% of the organisms evaluated. 10 Against other multidrug‐resistant Enterobacteriaceae, omadacycline inhibited 85.3% of non–ceftazidime‐susceptible (n=1439) and 52.7% of non–imipenem‐susceptible isolates (n=277). 8 Anaerobic, Atypical, and Other Coverage Similar to other tetracyclines, omadacycline displays activity against a variety of other organisms. Susceptibility of omadacycline was evaluated against 186 anaerobic organisms. 12 Against Bacteroides fragilis , Prevotella spp., Clostridium difficile , Clostridium perfringens , and Peptostreptococcus spp., MIC 90 values were 4, 2, 0.5, 16, and 1 mg/L, respectively. These values were equivalent or within 1‐dilution difference compared to tigecycline. Omadacycline displayed comparable susceptibility to doxycycline, tetracycline, clindamycin, azithromycin, and moxifloxacin against Mycoplasma spp. and Ureaplasma spp. with MIC 90 values less than or equal to 2 mg/L. 9 A total of 125 dog and cat bite infection isolates were tested for omadacycline susceptibility. 13 All isolates, excluding Eikenella corrodens , had omadacycline MICs less than 1 mg/L. Reduced susceptibility to all tetracyclines was observed for Eikenella spp. Omadacycline activity has also been evaluated against bioterrorism pathogens, including Bacillus anthracis and Yersinia pestis . 14 The MIC 90 value observed for omadacycline for B. anthracis was 0.06 mg/L compared to 0.06 mg/L for doxycycline and 0.12 mg/L for both ciprofloxacin and tetracycline. For Y. pestis , MIC 90 values for omadacycline, doxycycline, tetracycline, and ciprofloxacin were 1, 1, 2, and 0.03 mg/L, respectively. Gram‐Positive Aerobic Coverage Omadacycline provides activity against many gram‐positive organisms. In an evaluation of 339 gram‐positive isolates, omadacycline activity was compared to that of tetracycline, minocycline, doxycycline, vancomycin, and other agents with clinically relevant gram‐positive coverage. 4 Against S. aureus isolates, including methicillin‐susceptible, methicillin‐resistant, and multidrug‐resistant strains, omadacycline MIC at which 90% of isolates were inhibited (MIC 90 ) values were 0.5 mg/L. In S. aureus strains possessing tetracycline resistance, omadacycline produced MICs ranging from 0.125 to 1 mg/L. In both Enterococcus faecalis and E. faecium strains, including those resistant to vancomycin or tetracycline, omadacycline displayed activity with an MIC 90 value of 0.5 mg/L. All streptococcal strains were inhibited by omadacycline at concentrations of 0.5 mg/L. Similarly, in a global surveillance study including roughly 70,000 isolates, 99.9% of all S. aureus and Enterococcus spp. were inhibited by omadacycline concentrations less than or equal to 2 mg/L. 10 Gram‐Negative Aerobic Coverage Gram‐negative coverage for omadacycline includes many Enterobacteriaceae. Against clinical pathogens, omadacycline was compared to standard‐of‐care agents. For E. coli and Klebsiella spp., respectively, MIC 90 values of 2 and 4 mg/L were observed. Omadacycline also displays activity against Haemophilus influenzae and Moraxella catarrhalis . 8 , 11 In 3383 H. influenzae isolates, including 736 β‐lactamase positive, 99% of the organisms were inhibited by omadacycline concentrations of less than or equal to 2 mg/L. Against M. catarrhalis , 100% of the 1126 isolates were inhibited by concentrations of less than or equal to 1 mg/L. Omadacycline also demonstrates antimicrobial inhibition against multidrug‐resistant organisms. Activity against Acinetobacter baumannii‐Acinetobacter calcoaceticus spp. complex (n=2101) and other Acinetobacter spp. (n=292) was assessed, and omadacycline inhibited growth of 91.5% and 95.5% of the isolates at less than or equal to 4 mg/L, respectively. 10 Against 315 Stenotrophomonas maltophilia isolates, omadacycline inhibited growth of 82.2% of the organisms evaluated. 10 Against other multidrug‐resistant Enterobacteriaceae, omadacycline inhibited 85.3% of non–ceftazidime‐susceptible (n=1439) and 52.7% of non–imipenem‐susceptible isolates (n=277). 8 Anaerobic, Atypical, and Other Coverage Similar to other tetracyclines, omadacycline displays activity against a variety of other organisms. Susceptibility of omadacycline was evaluated against 186 anaerobic organisms. 12 Against Bacteroides fragilis , Prevotella spp., Clostridium difficile , Clostridium perfringens , and Peptostreptococcus spp., MIC 90 values were 4, 2, 0.5, 16, and 1 mg/L, respectively. These values were equivalent or within 1‐dilution difference compared to tigecycline. Omadacycline displayed comparable susceptibility to doxycycline, tetracycline, clindamycin, azithromycin, and moxifloxacin against Mycoplasma spp. and Ureaplasma spp. with MIC 90 values less than or equal to 2 mg/L. 9 A total of 125 dog and cat bite infection isolates were tested for omadacycline susceptibility. 13 All isolates, excluding Eikenella corrodens , had omadacycline MICs less than 1 mg/L. Reduced susceptibility to all tetracyclines was observed for Eikenella spp. Omadacycline activity has also been evaluated against bioterrorism pathogens, including Bacillus anthracis and Yersinia pestis . 14 The MIC 90 value observed for omadacycline for B. anthracis was 0.06 mg/L compared to 0.06 mg/L for doxycycline and 0.12 mg/L for both ciprofloxacin and tetracycline. For Y. pestis , MIC 90 values for omadacycline, doxycycline, tetracycline, and ciprofloxacin were 1, 1, 2, and 0.03 mg/L, respectively. Resistance Tetracycline resistance is common in both gram‐positive and gram‐negative organisms 10 , 15 The resistance mechanisms against tetracyclines can be stratified based on frequency into major and minor mechanisms. 6 , 7 , 16 The two major mechanisms of resistance are increased number of efflux pumps and production of ribosomal protection proteins. The two minor mechanisms of resistance include modification of the ribosomal target and enzymatic inactivation. 6 , 7 , 16 Omadacycline retains activity for organisms with these resistance genes and does not appear to be affected by resistance to other antimicrobials. 7 , 8 , 10 No strains with induced resistance have been reported. The primary advantage of omadacycline is its ability to bypass certain resistance mechanisms that affect older tetracycline antibiotics, including efflux pumps and ribosomal protection proteins. 7 , 16 When exposed to the ribosomal protection protein Tet(O), omadacycline continues to inhibit protein synthesis despite tetracycline becoming inactive in the presence of Tet(O). 6 Tet(B) produces an efflux protein in gram‐negative bacteria that causes resistance in the tetracycline class, but minocycline, glycylcyclines, and omadacycline are not affected by this gene. 6 In gram‐positive bacteria, the gene Tet(K) is responsible for tetracycline efflux, and omadacycline remains active despite its presence. 6 In addition to evading the mechanisms of resistance above, the resistance mechanisms active against omadacycline have not been found to be clinically relevant. 16 An analysis on omadacycline binding sites was performed on E. coli isolates. Omadacycline was found to have the same binding site as tetracycline and tigecycline and is susceptible to the same 16S rRNA mutations that confer binding‐site alterations. Two mutations to the 16S rRNA are required to affect the primary binding site, but when these mutations occur, tetracycline resistance results in a 4‐ to 8‐fold increase in MIC. However, this mechanism causes low level resistance and decreases the fitness of the organisms by impairing cell growth. 16 Pharmacokinetics Pharmacokinetics of both oral and IV omadacycline have been evaluated in several clinical studies. Absolute bioavailability of omadacycline is 34.5%, leading to an oral dose of 300 mg versus a 100 mg IV dose. 17 Omadacycline displays linear pharmacokinetics, with higher area under the curve (AUC) and maximum observed plasma concentrations ( C max ) with increasing dosages. 17 A mean plasma half‐life ( t 1/2 ) of 17 hours was observed, which was independent of formulation (tablet, solution, IV) for single‐dose administration. 18 In a multiple‐dose evaluation, including dosages up to 600 mg, a 13‐hour t 1/2 was observed after single dose but was comparable to the previous study on day 5 (16 hrs). 17 Protein binding of omadacycline was low (20%) and nonspecific. 19 Omadacycline undergoes minimal hepatic metabolism and is neither a substrate, inducer, nor inhibitor of the cytochrome P450 system. In addition, it is not a potential transport substrate or inhibitor of clinically relevant drug transporters. Omadacycline is eliminated predominately in the feces (81.1%), with some renal elimination (14.4%). 19 A phase I evaluation of 16 subjects, eight with end‐stage renal disease (ESRD) and eight healthy, assessed the pharmacokinetics of omadacycline. 20 The AUC values observed were comparable between ESRD and healthy subjects, irrespective of whether the omadacycline dose was administered before or after dialysis. In addition, volume of distribution and overall clearance were comparable between the two groups, suggesting that dosage adjustments are not required for omadacycline in patients with renal dysfunction. The effect of food on omadacycline bioavailability was assessed in a phase I, randomized, four‐period, crossover study. 21 The four periods consisted of overnight fasting with a standard high‐fat (nondairy) meal 3 hours after dosing, standard high‐fat (nondairy) meal 4 hours before dosing, standard high‐fat (nondairy) meal 2 hours before dosing, and standard high‐fat meal 2 hours before dosing. Substantial effects on omadacycline exposure were observed when administered in varying fed states and even greater effect when dairy was included compared to fasted states. Area under the curve reductions of 17%, 42%, and 63% occurred for nondairy meal 4 hours prior, nondairy meal 2 hours prior, and meal with dairy 2 hours prior to omadacycline receipt. These results suggest omadacycline should be taken in a fasted state with avoidance of dairy or other multivalent cations. Finally, omadacycline penetrated well into the epithelial lining fluid (ELF), suggesting that this agent may be an option for treatment of lower respiratory tract bacterial infections. In a phase I assessment of 58 healthy adult subjects, omadacycline and tigecycline plasma, ELF, and alveolar cell concentrations were compared. 22 Systemic exposure, based upon AUC 0–24 and AUC 0–12 values, was 3‐fold higher in the plasma, ELF, and alveolar cells for omadacycline versus tigecycline. There are no currently published data on omadacycline pharmacodynamics against gram‐negative organisms. Pharmacodynamics Like other tetracyclines, area under the unbound concentration‐time curve to the MIC ( f AUC/MIC) is the antimicrobial activity predictor for omadacycline. 23 Omadacycline pharmacodynamics were assessed in a murine pneumonia model. 24 In this model, omadacycline activity was examined against four S. pneumoniae isolates. Bactericidal (≥ 3‐log kill) activity was observed in three of the four strains. Approximately 100% of the drug in plasma penetrated into the ELF. The plasma f AUC/MIC values for stasis and 2‐log 10 kill were 15.79–19.83 and 18.65–56.2, respectively. Similarly, the ELF f AUC/MIC values for stasis and 2‐log 10 kill were 14.18–17.80 and 17.26–47.27, respectively. Animal Efficacy Models Two distinct animal efficacy models of omadacycline have been described. 14 , 24 First, omadacycline was evaluated against four strains of S. pneumoniae (MIC 0.03–0.125 mg/L) in a murine pneumonia model. 24 Omadacycline was administered subcutaneously at doses ranging from 0.1 to 25.6 mg/kg every 12 hours and was bactericidal at all doses in two strains. For the other two strains, omadacycline induced stasis at 0.92–1.28 mg/kg every 24‐hour doses and 1‐log 10 kill at 1.26–18.24 mg/kg every 24‐hour dose, respectively. In the second study, omadacycline was evaluated against B. anthracis and Y. pestis in postexposure prophylaxis (PEP) and in a delayed‐treatment model of inhalational anthrax in murine models. 14 In the Bacillus PEP arm, animals were exposed to four separate doses of B. anthracis (mean 30.5 × 50% lethal dose [LD 50 ]) by whole‐body aerosol. 14 Omadacycline was administered 24 hours postexposure at doses ranging from 0.75 to 15 mg/kg IP twice/day and was compared to ciprofloxacin 30 mg/kg twice/day, doxycycline 0.75–15 mg/kg twice/day, and saline. Omadacycline‐ and doxycycline‐matched doses (2.5–15 mg/kg) demonstrated similar survival rates, but omadacycline 0.75 mg/kg‐dose survival was significantly better (p=0.0125). All 10 saline animals died (median survival 2.25 days), compared to six of ten animals in the omadacycline 0.75 mg/kg group (median survival 4.75 days), eight of ten animals in the doxycycline 0.75 mg/kg group, and two of nine animals in the ciprofloxacin group. In the delayed‐treatment anthrax model, omadacycline 15 mg/kg IP twice daily was administered 48 hours postexposure and compared to ciprofloxacin 30 mg/kg, doxycycline 15 mg/kg, and saline. No significant differences were seen between omadacycline, ciprofloxacin, and doxycycline. All 10 animals died in the saline arm (median survival 2.25 days), whereas four, two, and three animals died in the omadacycline, ciprofloxacin, and doxycycline arms, respectively. In the Yersinia arm, animals were exposed to three separate doses of Y. pestis (mean 29.4x LD 50 ) by whole‐body aerosol. 14 Omadacycline (5–40 mg/kg IP twice/day) for 7 days was compared to doxycycline (5–40 mg/kg twice/day), ciprofloxacin (15 mg/kg twice/day), and saline. In mice treated with omadacycline 40 mg/kg or doxycycline 40 mg/kg, 90% survived (compared to 100% of ciprofloxacin‐treated mice and no (0%) saline‐treated mice). In omadacycline 40 mg/kg‐treated and ciprofloxacin‐treated mice, no viable bacteria were recovered. Clinical Efficacy Trials Community‐Acquired Bacterial Pneumonia The Omadacycline for Pneumonia Treatment In the Community (OPTIC) trial was a phase III, randomized, double‐blind, multicenter study comparing the safety and efficacy of omadacycline to moxifloxacin for the treatment of adults with CABP. 25 Patients were randomized to IV omadacycline 100 mg twice/day for two doses followed by 100 mg IV/day or IV moxifloxacin 400 mg/day for 3 days, with the option to switch to oral therapy or continue IV for a total of 7–14 days. Patients were included if they had clinical evidence of CABP, signs of infection or systemic inflammatory response, radiographically confirmed acute bacterial pneumonia, and CABP categorized as Pneumonia Patient Outcomes Research Team Score (PORT) risk class II, III, or IV. Patients were excluded if they received other effective antibacterial treatment within 72 hours, had hospital‐acquired or health care–associated pneumonia, empyema, lung abscess, septic shock, or end‐stage liver disease. The primary end points were early clinical response (ECR) at 72–120 hours (FDA primary end point) and clinical success at the posttreatment evaluation (PTE, occurred 5–10 days following the last treatment dose) in both the intention‐to‐treat (ITT) population and the clinically evaluable (CE) populations (European Medicines Agency [EMA] primary end point). In the ITT population at ECR evaluation, omadacycline performed similarly to moxifloxacin (81.1% vs 82.7%, treatment difference [TD] −1.6 [95% confidence interval (CI) −7.1 to 3.8]) (Table 2 ). At the PTE, results for omadacycline versus moxifloxacin were similar in both the ITT (87.6% vs 85.1%; TD 2.5 [−2.4 to 7.4]) and the CE (92.9% vs 90.4%; TD 2.5 [95% CI −1.7 to 6.8]) populations, respectively. By pathogen, omadacycline had similar rates of clinical success as moxifloxacin against atypical pathogens (92.4% vs 91.5%), gram‐negative bacteria (84.8% vs 80.9%), and S. pneumoniae (86% vs 91.2%), respectively. Clinical success rates for omadacycline were slightly less with S. aureus compared to moxifloxacin (72.7% vs 81.8%, respectively). Based on these results, omadacycline was non‐inferior for both the FDA and EMA end points. Table 2 Summary of Clinical Response in Patients Treated with Omadacycline in Phase II and III trials Population Rate of Clinical Response, n (%) % Difference (95% CI) Phase II cSSSI trial 26 Omadacycline Linezolid TOC (ITT) 98/111 (88.3) 82/108 (75.9) 12.4 (1.9 to 22.9) TOC (MITT) 75/84 (89.3) 59/78 (75.6) 13.6 (1.4 to 25.9) TOC (CE) 98/100 (98) 82/88 (93.2) 4.8 (−1.7 to 11.3) TOC (ME) 75/77 (97.4) 59/63 (93.7) 3.8 (−4.0 to 11.5) Phase III ABSSSI trial (OASIS) 27 ECR (MITT) 268/316 (84.8) 267/311 (85.8) −0.7 (−6.3 to 4.9) PTE (MITT) 272/316 (86.1) 260/311 (83.6) 2.5 (−3.2 to 8.2) PTE (CE) 259/269 (96.3) 243/260 (93.5) 2.8 (−1.0 to 6.9) Phase III ABSSSI trial (OASIS‐2) 28 ECR (MITT) 315/360 (87.5) 297/360 (82.5) 5.0 (−0.2 to 10.3) PTE (MITT) 303/360 (84.2) 291/360 (80.8) 3.3 (−2.2 to 9.0) PTE (CE) 278/284 (97.9) 279/292 (95.5) 2.3 (−0.5 to 5.8) Phase III CABP trial (OPTIC) 25 Omadacycline Moxifloxacin ECR (ITT) 313/386 (81.1) 320/388 (82.7) −1.6 (−7.1 to 3.8) PTE (ITT) 338/386 (87.6) 330/388 (85.1) 2.5 (−2.4 to 7.4) PTE (CE) 316/340 (92.9) 312/345 (90.4) 2.5 (−1.7 to 6.8) ABSSSI = acute bacterial skin and skin structure infection; CABP = community‐acquired bacterial pneumonia; cSSSI = complicated skin and skin structure infection; CE = clinically evaluable; ECR = early clinical response; ITT = intent to treat; MITT = modified intent to treat; ME = microbiologically evaluable; PTE = post‐treatment evaluation; TOC = test of cure. John Wiley & Sons, Ltd Skin and Skin Structure Infections (SSSI) Phase II Trial A phase II, randomized, controlled, investigator‐blind, multicenter study compared omadacycline monotherapy to linezolid with or without aztreonam for treatment of complicated SSSI (cSSSI). 26 Included patients were ≥ 18 years old and had one of four cSSSI (wound infection, major abscess, infected ulcer in the lower extremity, or cellulitis). Any patient with an infection treatable with surgical intervention alone was excluded. Both omadacycline and linezolid were initially given as IV (100 mg IV daily vs 600 mg IV twice/day, respectively) with the option to transition to oral (200 mg orally/day vs 600 mg orally twice/day, respectively). Patients were evaluated at four specified time points: baseline, end of IV treatment, end of treatment, and 10–17 days after the last dose of treatment (test of cure [TOC] evaluation). The primary hypothesis was that safety and tolerability of omadacycline was comparable with linezolid. The secondary hypothesis was that omadacycline was noninferior to linezolid for the rate of successful clinical response at TOC. The majority of patients in the omadacycline (65.8%) and linezolid (66.7%) groups had major abscesses. The most common organism identified was S. aureus, with the majority being MRSA. Patients received an equivalent mean duration of IV treatment (4.3 days) for both drugs, and the overall mean treatment duration was similar between the omadacycline and linezolid groups (10 vs 9.6 days, respectively). In each of the four populations assessed, omadacycline‐treated patients had higher rates of treatment success compared with patients receiving linezolid. For patients treated with omadacycline versus linezolid, clinical cure rates were 88.3% versus 75.9% (TD 12.4 [95% CI 1.9–22.9]) for the ITT population; 89.3% versus 75.6% (TD 13.6 [95% CI 1.4–25.9]) in the MITT population; 98% versus 93.2% (TD 4.8 [95% CI −1.7 to 11.3]) in the CE population; and 97.4% versus 93.7% (TD 3.8 [95% CI −4.0 to 11.5]) in the microbiologically evaluable (ME) population, respectively (Table 2 ). When evaluating the ME population by organism, both treatment arms achieved 100% clinical response for gram‐negative and gram‐positive bacteria other than S. aureus . For S. aureus, omadacycline had higher success rates compared with linezolid (97.2% vs 92.7%), including MRSA (97.7% vs 93.8%). A retrospective analysis of the ITT patients was performed to evaluate clinical response during the first 72 hours of starting therapy to align with the 2010 FDA guidance for the development of drugs for treatment of ABSSSI. 29 Clinical response at 72 hours was similar between the omadacycline‐ and linezolid‐treated patients (96.8% vs 94.4%, respectively). These ECR data were primarily from patients transitioning from IV to oral medications in the first 72 hours of treatment. 26 Based on the clinical response results for this phase II trial for cSSSI, omadacycline was deemed non‐inferior to linezolid. Phase III Trials The Omadacycline in Acute Skin and Skin Infections Study (OASIS) and OASIS‐2 are phase III, randomized, double‐blind, multicenter studies comparing the safety and efficacy of omadacycline to linezolid for treating adult subjects with ABSSSI. 27 , 28 , 30 , 31 , 32 In the OASIS, patients were randomized 1:1 to receive IV omadacycline 100 mg twice/day for two doses followed by 100 mg IV/day or IV linezolid 600 mg twice/day with the option to continue IV or switch to oral omadacycline 300 mg/day or oral linezolid 600 mg twice/day for a total of 7–14 days. 27 The primary end points were ECR at 48–72 hours (FDA primary end point) and clinical response at the PTE in the MITT population and CE population, which was 7–14 days after the last dose of treatment (EMA co‐primary end points). Included patients had a qualifying ABSSSI (≥ 75 cm 2 total surface area of contiguous involved tissue) and evidence of systemic inflammatory response within 24 hours prior to randomization. Patients were excluded if ≥ 1 dose of a potentially effective antibiotic were given within 72 hours prior to the first dose of the study drug or if the patient was taking medications known to interact with linezolid, such as monoamine oxidase inhibitors (MAOIs). Unlike the phase II cSSSI study, infection type was well distributed in the OASIS population with similar percentages of patients with wound infections, cellulitis/erysipelas, and major abscesses. For ECR in the MITT population, omadacycline (84.8%) performed similarly to linezolid (85.8%; TD −0.7 [−6.3 to 4.9]) (Table 2 ). Omadacycline performed comparatively to linezolid at the PTE for MITT (86.1% vs 83.6%; TD 2.5 [−3.2 to 8.2]) and CE populations (96.3% vs 93.5%; TD 2.8 [−1.0 to 6.9]), respectively. Based on these results, omadacycline met noninferiority for both the FDA and EMA efficacy end points. Clinical response by pathogen at PTE in the micro‐MITT population showed comparable clinical responses for omadacycline and linezolid. Clinical success rates were high for both omadacycline and linezolid for S. aureus (83.3% vs 83.4%), including MRSA (82.6% vs 86%), respectively. Success rates were lower for both omadacycline (74.5%) and linezolid (70.3%) for Streptococcus anginosus group. On the other hand, omadacycline had lower rates of clinical success compared with linezolid for Streptococcus pyogenes (72.7% vs 88.9%, respectively). For vancomycin‐sensitive E. faecalis, clinical success rates were higher for both omadacycline and linezolid (90.0% vs 92.3%, respectively). Multiple subgroup analyses from the OASIS trial have been performed, including patients with high body mass index (BMI), diabetes, chronic kidney disease (CKD), intravenous drug use (IVDU), and hepatitis C positive (HCV+). 30 , 31 , 32 Patients with normal BMI compared with high BMI (i.e., overweight and obese) had similar rates of clinical success at ECR for omadacycline (84.7% vs 84.8%, respectively) and linezolid (84.9% vs 85.9%, respectively). 30 Clinical success rates at PTE were similar in the MITT and CE populations regardless of treatment or BMI, and no significant differences were found between the subgroups. For patients with CKD, clinical success at ECR was similar between omadacycline and linezolid regardless of CKD staging (CKD stage 0/1 vs 2/3). 31 Omadacycline performed comparatively at PTE for both the MITT and CE populations compared with linezolid. Clinical success was high in both groups across all CKD stages, and no statistical differences were found. In the IVDU subgroup analysis, no significant difference between treatment groups was found in clinical success rates at ECR for the MITT population. 32 Similarly, clinical success rates were comparable between IVDU patients, IVDU/HCV+ patients, and non‐IVDU/HCV− patients at PTE in the MITT and CE populations. A significantly higher clinical success rate at PTE was found with omadacycline compared to linezolid in non‐IVDU patients in both the MITT and CE populations. In the OASIS‐2 study, patients were randomized 1:1 to receive oral omadacycline or linezolid for a total duration of 7–14 days. 28 Patients in the oral omadacycline treatment arm received 450 mg/day for the first 2 days, then 300 mg/day thereafter. Patients in the oral linezolid treatment arm received 600 mg twice/day from day 1 of treatment onwards. The same primary end points were evaluated in OASIS and OASIS‐2 (i.e., ECR at 48–72 hrs for the MITT population and PTE 7–14 days after last treatment dose for the MITT and CE populations). Primary infection type was most commonly wound infections in the omadacycline and linezolid groups (58.3% vs 59.4%), followed by cellulitis/erysipelas and major abscesses. Omadacycline achieved all primary end points for noninferiority for the FDA and EMA. For the MITT population, ECR was similar for omadacycline and linezolid (87.5% vs 82.5%; TD 5.0 [−0.2 to 10.3], respectively) (Table 2 ). Clinical success rates were also comparable at PTE for the MITT population (84.2% vs 80.8%; TD 3.3 [−2.2 to 9.0]) and CE population (97.9% vs 95.5%; TD 2.3 [−0.5 to 5.8]), respectively. Omadacycline maintained high rates of clinical success for all gram‐positive pathogens in the study and performed favorably over linezolid for S. aureus (82.7% vs 79.8%; n=453), MRSA (85.6% vs 79.4%; n=211), S. pyogenes (69% vs 56.3%; n=45), S. anginosus group (86% vs 73.3%; n=47), and vancomycin‐sensitive E. faecalis (100% vs 70%; n=17), respectively. Community‐Acquired Bacterial Pneumonia The Omadacycline for Pneumonia Treatment In the Community (OPTIC) trial was a phase III, randomized, double‐blind, multicenter study comparing the safety and efficacy of omadacycline to moxifloxacin for the treatment of adults with CABP. 25 Patients were randomized to IV omadacycline 100 mg twice/day for two doses followed by 100 mg IV/day or IV moxifloxacin 400 mg/day for 3 days, with the option to switch to oral therapy or continue IV for a total of 7–14 days. Patients were included if they had clinical evidence of CABP, signs of infection or systemic inflammatory response, radiographically confirmed acute bacterial pneumonia, and CABP categorized as Pneumonia Patient Outcomes Research Team Score (PORT) risk class II, III, or IV. Patients were excluded if they received other effective antibacterial treatment within 72 hours, had hospital‐acquired or health care–associated pneumonia, empyema, lung abscess, septic shock, or end‐stage liver disease. The primary end points were early clinical response (ECR) at 72–120 hours (FDA primary end point) and clinical success at the posttreatment evaluation (PTE, occurred 5–10 days following the last treatment dose) in both the intention‐to‐treat (ITT) population and the clinically evaluable (CE) populations (European Medicines Agency [EMA] primary end point). In the ITT population at ECR evaluation, omadacycline performed similarly to moxifloxacin (81.1% vs 82.7%, treatment difference [TD] −1.6 [95% confidence interval (CI) −7.1 to 3.8]) (Table 2 ). At the PTE, results for omadacycline versus moxifloxacin were similar in both the ITT (87.6% vs 85.1%; TD 2.5 [−2.4 to 7.4]) and the CE (92.9% vs 90.4%; TD 2.5 [95% CI −1.7 to 6.8]) populations, respectively. By pathogen, omadacycline had similar rates of clinical success as moxifloxacin against atypical pathogens (92.4% vs 91.5%), gram‐negative bacteria (84.8% vs 80.9%), and S. pneumoniae (86% vs 91.2%), respectively. Clinical success rates for omadacycline were slightly less with S. aureus compared to moxifloxacin (72.7% vs 81.8%, respectively). Based on these results, omadacycline was non‐inferior for both the FDA and EMA end points. Table 2 Summary of Clinical Response in Patients Treated with Omadacycline in Phase II and III trials Population Rate of Clinical Response, n (%) % Difference (95% CI) Phase II cSSSI trial 26 Omadacycline Linezolid TOC (ITT) 98/111 (88.3) 82/108 (75.9) 12.4 (1.9 to 22.9) TOC (MITT) 75/84 (89.3) 59/78 (75.6) 13.6 (1.4 to 25.9) TOC (CE) 98/100 (98) 82/88 (93.2) 4.8 (−1.7 to 11.3) TOC (ME) 75/77 (97.4) 59/63 (93.7) 3.8 (−4.0 to 11.5) Phase III ABSSSI trial (OASIS) 27 ECR (MITT) 268/316 (84.8) 267/311 (85.8) −0.7 (−6.3 to 4.9) PTE (MITT) 272/316 (86.1) 260/311 (83.6) 2.5 (−3.2 to 8.2) PTE (CE) 259/269 (96.3) 243/260 (93.5) 2.8 (−1.0 to 6.9) Phase III ABSSSI trial (OASIS‐2) 28 ECR (MITT) 315/360 (87.5) 297/360 (82.5) 5.0 (−0.2 to 10.3) PTE (MITT) 303/360 (84.2) 291/360 (80.8) 3.3 (−2.2 to 9.0) PTE (CE) 278/284 (97.9) 279/292 (95.5) 2.3 (−0.5 to 5.8) Phase III CABP trial (OPTIC) 25 Omadacycline Moxifloxacin ECR (ITT) 313/386 (81.1) 320/388 (82.7) −1.6 (−7.1 to 3.8) PTE (ITT) 338/386 (87.6) 330/388 (85.1) 2.5 (−2.4 to 7.4) PTE (CE) 316/340 (92.9) 312/345 (90.4) 2.5 (−1.7 to 6.8) ABSSSI = acute bacterial skin and skin structure infection; CABP = community‐acquired bacterial pneumonia; cSSSI = complicated skin and skin structure infection; CE = clinically evaluable; ECR = early clinical response; ITT = intent to treat; MITT = modified intent to treat; ME = microbiologically evaluable; PTE = post‐treatment evaluation; TOC = test of cure. John Wiley & Sons, Ltd Skin and Skin Structure Infections (SSSI) Phase II Trial A phase II, randomized, controlled, investigator‐blind, multicenter study compared omadacycline monotherapy to linezolid with or without aztreonam for treatment of complicated SSSI (cSSSI). 26 Included patients were ≥ 18 years old and had one of four cSSSI (wound infection, major abscess, infected ulcer in the lower extremity, or cellulitis). Any patient with an infection treatable with surgical intervention alone was excluded. Both omadacycline and linezolid were initially given as IV (100 mg IV daily vs 600 mg IV twice/day, respectively) with the option to transition to oral (200 mg orally/day vs 600 mg orally twice/day, respectively). Patients were evaluated at four specified time points: baseline, end of IV treatment, end of treatment, and 10–17 days after the last dose of treatment (test of cure [TOC] evaluation). The primary hypothesis was that safety and tolerability of omadacycline was comparable with linezolid. The secondary hypothesis was that omadacycline was noninferior to linezolid for the rate of successful clinical response at TOC. The majority of patients in the omadacycline (65.8%) and linezolid (66.7%) groups had major abscesses. The most common organism identified was S. aureus, with the majority being MRSA. Patients received an equivalent mean duration of IV treatment (4.3 days) for both drugs, and the overall mean treatment duration was similar between the omadacycline and linezolid groups (10 vs 9.6 days, respectively). In each of the four populations assessed, omadacycline‐treated patients had higher rates of treatment success compared with patients receiving linezolid. For patients treated with omadacycline versus linezolid, clinical cure rates were 88.3% versus 75.9% (TD 12.4 [95% CI 1.9–22.9]) for the ITT population; 89.3% versus 75.6% (TD 13.6 [95% CI 1.4–25.9]) in the MITT population; 98% versus 93.2% (TD 4.8 [95% CI −1.7 to 11.3]) in the CE population; and 97.4% versus 93.7% (TD 3.8 [95% CI −4.0 to 11.5]) in the microbiologically evaluable (ME) population, respectively (Table 2 ). When evaluating the ME population by organism, both treatment arms achieved 100% clinical response for gram‐negative and gram‐positive bacteria other than S. aureus . For S. aureus, omadacycline had higher success rates compared with linezolid (97.2% vs 92.7%), including MRSA (97.7% vs 93.8%). A retrospective analysis of the ITT patients was performed to evaluate clinical response during the first 72 hours of starting therapy to align with the 2010 FDA guidance for the development of drugs for treatment of ABSSSI. 29 Clinical response at 72 hours was similar between the omadacycline‐ and linezolid‐treated patients (96.8% vs 94.4%, respectively). These ECR data were primarily from patients transitioning from IV to oral medications in the first 72 hours of treatment. 26 Based on the clinical response results for this phase II trial for cSSSI, omadacycline was deemed non‐inferior to linezolid. Phase III Trials The Omadacycline in Acute Skin and Skin Infections Study (OASIS) and OASIS‐2 are phase III, randomized, double‐blind, multicenter studies comparing the safety and efficacy of omadacycline to linezolid for treating adult subjects with ABSSSI. 27 , 28 , 30 , 31 , 32 In the OASIS, patients were randomized 1:1 to receive IV omadacycline 100 mg twice/day for two doses followed by 100 mg IV/day or IV linezolid 600 mg twice/day with the option to continue IV or switch to oral omadacycline 300 mg/day or oral linezolid 600 mg twice/day for a total of 7–14 days. 27 The primary end points were ECR at 48–72 hours (FDA primary end point) and clinical response at the PTE in the MITT population and CE population, which was 7–14 days after the last dose of treatment (EMA co‐primary end points). Included patients had a qualifying ABSSSI (≥ 75 cm 2 total surface area of contiguous involved tissue) and evidence of systemic inflammatory response within 24 hours prior to randomization. Patients were excluded if ≥ 1 dose of a potentially effective antibiotic were given within 72 hours prior to the first dose of the study drug or if the patient was taking medications known to interact with linezolid, such as monoamine oxidase inhibitors (MAOIs). Unlike the phase II cSSSI study, infection type was well distributed in the OASIS population with similar percentages of patients with wound infections, cellulitis/erysipelas, and major abscesses. For ECR in the MITT population, omadacycline (84.8%) performed similarly to linezolid (85.8%; TD −0.7 [−6.3 to 4.9]) (Table 2 ). Omadacycline performed comparatively to linezolid at the PTE for MITT (86.1% vs 83.6%; TD 2.5 [−3.2 to 8.2]) and CE populations (96.3% vs 93.5%; TD 2.8 [−1.0 to 6.9]), respectively. Based on these results, omadacycline met noninferiority for both the FDA and EMA efficacy end points. Clinical response by pathogen at PTE in the micro‐MITT population showed comparable clinical responses for omadacycline and linezolid. Clinical success rates were high for both omadacycline and linezolid for S. aureus (83.3% vs 83.4%), including MRSA (82.6% vs 86%), respectively. Success rates were lower for both omadacycline (74.5%) and linezolid (70.3%) for Streptococcus anginosus group. On the other hand, omadacycline had lower rates of clinical success compared with linezolid for Streptococcus pyogenes (72.7% vs 88.9%, respectively). For vancomycin‐sensitive E. faecalis, clinical success rates were higher for both omadacycline and linezolid (90.0% vs 92.3%, respectively). Multiple subgroup analyses from the OASIS trial have been performed, including patients with high body mass index (BMI), diabetes, chronic kidney disease (CKD), intravenous drug use (IVDU), and hepatitis C positive (HCV+). 30 , 31 , 32 Patients with normal BMI compared with high BMI (i.e., overweight and obese) had similar rates of clinical success at ECR for omadacycline (84.7% vs 84.8%, respectively) and linezolid (84.9% vs 85.9%, respectively). 30 Clinical success rates at PTE were similar in the MITT and CE populations regardless of treatment or BMI, and no significant differences were found between the subgroups. For patients with CKD, clinical success at ECR was similar between omadacycline and linezolid regardless of CKD staging (CKD stage 0/1 vs 2/3). 31 Omadacycline performed comparatively at PTE for both the MITT and CE populations compared with linezolid. Clinical success was high in both groups across all CKD stages, and no statistical differences were found. In the IVDU subgroup analysis, no significant difference between treatment groups was found in clinical success rates at ECR for the MITT population. 32 Similarly, clinical success rates were comparable between IVDU patients, IVDU/HCV+ patients, and non‐IVDU/HCV− patients at PTE in the MITT and CE populations. A significantly higher clinical success rate at PTE was found with omadacycline compared to linezolid in non‐IVDU patients in both the MITT and CE populations. In the OASIS‐2 study, patients were randomized 1:1 to receive oral omadacycline or linezolid for a total duration of 7–14 days. 28 Patients in the oral omadacycline treatment arm received 450 mg/day for the first 2 days, then 300 mg/day thereafter. Patients in the oral linezolid treatment arm received 600 mg twice/day from day 1 of treatment onwards. The same primary end points were evaluated in OASIS and OASIS‐2 (i.e., ECR at 48–72 hrs for the MITT population and PTE 7–14 days after last treatment dose for the MITT and CE populations). Primary infection type was most commonly wound infections in the omadacycline and linezolid groups (58.3% vs 59.4%), followed by cellulitis/erysipelas and major abscesses. Omadacycline achieved all primary end points for noninferiority for the FDA and EMA. For the MITT population, ECR was similar for omadacycline and linezolid (87.5% vs 82.5%; TD 5.0 [−0.2 to 10.3], respectively) (Table 2 ). Clinical success rates were also comparable at PTE for the MITT population (84.2% vs 80.8%; TD 3.3 [−2.2 to 9.0]) and CE population (97.9% vs 95.5%; TD 2.3 [−0.5 to 5.8]), respectively. Omadacycline maintained high rates of clinical success for all gram‐positive pathogens in the study and performed favorably over linezolid for S. aureus (82.7% vs 79.8%; n=453), MRSA (85.6% vs 79.4%; n=211), S. pyogenes (69% vs 56.3%; n=45), S. anginosus group (86% vs 73.3%; n=47), and vancomycin‐sensitive E. faecalis (100% vs 70%; n=17), respectively. Phase II Trial A phase II, randomized, controlled, investigator‐blind, multicenter study compared omadacycline monotherapy to linezolid with or without aztreonam for treatment of complicated SSSI (cSSSI). 26 Included patients were ≥ 18 years old and had one of four cSSSI (wound infection, major abscess, infected ulcer in the lower extremity, or cellulitis). Any patient with an infection treatable with surgical intervention alone was excluded. Both omadacycline and linezolid were initially given as IV (100 mg IV daily vs 600 mg IV twice/day, respectively) with the option to transition to oral (200 mg orally/day vs 600 mg orally twice/day, respectively). Patients were evaluated at four specified time points: baseline, end of IV treatment, end of treatment, and 10–17 days after the last dose of treatment (test of cure [TOC] evaluation). The primary hypothesis was that safety and tolerability of omadacycline was comparable with linezolid. The secondary hypothesis was that omadacycline was noninferior to linezolid for the rate of successful clinical response at TOC. The majority of patients in the omadacycline (65.8%) and linezolid (66.7%) groups had major abscesses. The most common organism identified was S. aureus, with the majority being MRSA. Patients received an equivalent mean duration of IV treatment (4.3 days) for both drugs, and the overall mean treatment duration was similar between the omadacycline and linezolid groups (10 vs 9.6 days, respectively). In each of the four populations assessed, omadacycline‐treated patients had higher rates of treatment success compared with patients receiving linezolid. For patients treated with omadacycline versus linezolid, clinical cure rates were 88.3% versus 75.9% (TD 12.4 [95% CI 1.9–22.9]) for the ITT population; 89.3% versus 75.6% (TD 13.6 [95% CI 1.4–25.9]) in the MITT population; 98% versus 93.2% (TD 4.8 [95% CI −1.7 to 11.3]) in the CE population; and 97.4% versus 93.7% (TD 3.8 [95% CI −4.0 to 11.5]) in the microbiologically evaluable (ME) population, respectively (Table 2 ). When evaluating the ME population by organism, both treatment arms achieved 100% clinical response for gram‐negative and gram‐positive bacteria other than S. aureus . For S. aureus, omadacycline had higher success rates compared with linezolid (97.2% vs 92.7%), including MRSA (97.7% vs 93.8%). A retrospective analysis of the ITT patients was performed to evaluate clinical response during the first 72 hours of starting therapy to align with the 2010 FDA guidance for the development of drugs for treatment of ABSSSI. 29 Clinical response at 72 hours was similar between the omadacycline‐ and linezolid‐treated patients (96.8% vs 94.4%, respectively). These ECR data were primarily from patients transitioning from IV to oral medications in the first 72 hours of treatment. 26 Based on the clinical response results for this phase II trial for cSSSI, omadacycline was deemed non‐inferior to linezolid. Phase III Trials The Omadacycline in Acute Skin and Skin Infections Study (OASIS) and OASIS‐2 are phase III, randomized, double‐blind, multicenter studies comparing the safety and efficacy of omadacycline to linezolid for treating adult subjects with ABSSSI. 27 , 28 , 30 , 31 , 32 In the OASIS, patients were randomized 1:1 to receive IV omadacycline 100 mg twice/day for two doses followed by 100 mg IV/day or IV linezolid 600 mg twice/day with the option to continue IV or switch to oral omadacycline 300 mg/day or oral linezolid 600 mg twice/day for a total of 7–14 days. 27 The primary end points were ECR at 48–72 hours (FDA primary end point) and clinical response at the PTE in the MITT population and CE population, which was 7–14 days after the last dose of treatment (EMA co‐primary end points). Included patients had a qualifying ABSSSI (≥ 75 cm 2 total surface area of contiguous involved tissue) and evidence of systemic inflammatory response within 24 hours prior to randomization. Patients were excluded if ≥ 1 dose of a potentially effective antibiotic were given within 72 hours prior to the first dose of the study drug or if the patient was taking medications known to interact with linezolid, such as monoamine oxidase inhibitors (MAOIs). Unlike the phase II cSSSI study, infection type was well distributed in the OASIS population with similar percentages of patients with wound infections, cellulitis/erysipelas, and major abscesses. For ECR in the MITT population, omadacycline (84.8%) performed similarly to linezolid (85.8%; TD −0.7 [−6.3 to 4.9]) (Table 2 ). Omadacycline performed comparatively to linezolid at the PTE for MITT (86.1% vs 83.6%; TD 2.5 [−3.2 to 8.2]) and CE populations (96.3% vs 93.5%; TD 2.8 [−1.0 to 6.9]), respectively. Based on these results, omadacycline met noninferiority for both the FDA and EMA efficacy end points. Clinical response by pathogen at PTE in the micro‐MITT population showed comparable clinical responses for omadacycline and linezolid. Clinical success rates were high for both omadacycline and linezolid for S. aureus (83.3% vs 83.4%), including MRSA (82.6% vs 86%), respectively. Success rates were lower for both omadacycline (74.5%) and linezolid (70.3%) for Streptococcus anginosus group. On the other hand, omadacycline had lower rates of clinical success compared with linezolid for Streptococcus pyogenes (72.7% vs 88.9%, respectively). For vancomycin‐sensitive E. faecalis, clinical success rates were higher for both omadacycline and linezolid (90.0% vs 92.3%, respectively). Multiple subgroup analyses from the OASIS trial have been performed, including patients with high body mass index (BMI), diabetes, chronic kidney disease (CKD), intravenous drug use (IVDU), and hepatitis C positive (HCV+). 30 , 31 , 32 Patients with normal BMI compared with high BMI (i.e., overweight and obese) had similar rates of clinical success at ECR for omadacycline (84.7% vs 84.8%, respectively) and linezolid (84.9% vs 85.9%, respectively). 30 Clinical success rates at PTE were similar in the MITT and CE populations regardless of treatment or BMI, and no significant differences were found between the subgroups. For patients with CKD, clinical success at ECR was similar between omadacycline and linezolid regardless of CKD staging (CKD stage 0/1 vs 2/3). 31 Omadacycline performed comparatively at PTE for both the MITT and CE populations compared with linezolid. Clinical success was high in both groups across all CKD stages, and no statistical differences were found. In the IVDU subgroup analysis, no significant difference between treatment groups was found in clinical success rates at ECR for the MITT population. 32 Similarly, clinical success rates were comparable between IVDU patients, IVDU/HCV+ patients, and non‐IVDU/HCV− patients at PTE in the MITT and CE populations. A significantly higher clinical success rate at PTE was found with omadacycline compared to linezolid in non‐IVDU patients in both the MITT and CE populations. In the OASIS‐2 study, patients were randomized 1:1 to receive oral omadacycline or linezolid for a total duration of 7–14 days. 28 Patients in the oral omadacycline treatment arm received 450 mg/day for the first 2 days, then 300 mg/day thereafter. Patients in the oral linezolid treatment arm received 600 mg twice/day from day 1 of treatment onwards. The same primary end points were evaluated in OASIS and OASIS‐2 (i.e., ECR at 48–72 hrs for the MITT population and PTE 7–14 days after last treatment dose for the MITT and CE populations). Primary infection type was most commonly wound infections in the omadacycline and linezolid groups (58.3% vs 59.4%), followed by cellulitis/erysipelas and major abscesses. Omadacycline achieved all primary end points for noninferiority for the FDA and EMA. For the MITT population, ECR was similar for omadacycline and linezolid (87.5% vs 82.5%; TD 5.0 [−0.2 to 10.3], respectively) (Table 2 ). Clinical success rates were also comparable at PTE for the MITT population (84.2% vs 80.8%; TD 3.3 [−2.2 to 9.0]) and CE population (97.9% vs 95.5%; TD 2.3 [−0.5 to 5.8]), respectively. Omadacycline maintained high rates of clinical success for all gram‐positive pathogens in the study and performed favorably over linezolid for S. aureus (82.7% vs 79.8%; n=453), MRSA (85.6% vs 79.4%; n=211), S. pyogenes (69% vs 56.3%; n=45), S. anginosus group (86% vs 73.3%; n=47), and vancomycin‐sensitive E. faecalis (100% vs 70%; n=17), respectively. Safety and Tolerability Omadacycline's safety and tolerability were tested in three phase III trials: OPTIC, OASIS, and OASIS‐2. In OPTIC, there were 170 (44.5%) patients in the omadacycline group and 200 (51.5%) patients in the moxifloxacin group who experienced any adverse event during the study period. 25 Treatment‐emergent adverse events (TEAEs) occurred in 157 (41.1%) and 188 (48.5%) patients in the omadacycline and moxifloxacin groups, respectively (Table 3 ), with only 39 (10.2%) omadacycline patients and 69 (17.8%) moxifloxacin patients experiencing drug‐related TEAEs. Of those patients, only two (0.5%) patients in each group experienced a serious drug‐related TEAE. Most of these TEAEs were gastrointestinal (GI) related (vomiting and nausea) and did not cause discontinuation of the study drugs. In addition, there were no reported cases of C. difficile due to omadacycline; however, eight (2%) patients developed a C. difficile infection or complication in the moxifloxacin group. Overall, drug discontinuation due to any adverse event was low in both groups (4.4% omadacycline vs 7.2% moxifloxacin). 25 Table 3 Treatment‐Emergent Adverse Events (TEAEs) in at Least 5% of Patients Treated with Omadacycline in Phase III Study Populations 25 , 27 , 28 Adverse Events, n (%) CABP ABSSSI OPTIC OASIS OASIS‐2 Omadacycline IV/PO (n=382) Moxifloxacin IV/PO (n=388) Omadacycline IV/PO (n=323) Linezolid IV/PO (n=322) Omadacycline PO (n=368) Linezolid PO (n=367) Any TEAE 157 (41.1) 188 (48.5) 156 (48.3) 147 (45.7) 197 (53.5) 137 (37.3) Nausea 9 (2.4) 21 (5.4) 10 (12.4) 32 (9.9) 111 (30.2) 28 (7.6) Infusion site extravasation 4 (1.0) 9 (2.3) 28 (8.7) 19 (5.9) – – Subcutaneous abscess – – 17 (5.3) 19 (5.9) – – Vomiting 10 (2.6) 6 (1.5) 17 (5.3) 16 (5.0) 62 (16.8) 11 (3) Wound Infection 1 (0.3) – 15 (4.6) 15 (4.7) 22 (6.0) 17 (4.6) ALT Increased 14 (3.7) 18 (4.6) 9 (2.8) 14 (4.3) 19 (5.2) 11 (3.0) ALT = alanine transaminase; ABSSSI = acute bacterial skin and skin structure infection; CABP = community‐acquired bacterial pneumonia; IV = intravenous; OASIS = Omadacycline in Acute Skin and Skin Infections Study; OPTIC = Omadacycline for Pneumonia Treatment In the Community; PO = oral; TEAE = treatment‐emergent adverse event. John Wiley & Sons, Ltd In the OASIS trial, TEAEs occurred in 156 (48.3%) patients receiving omadacycline, which was slightly higher than those in the linezolid group (147 (45.7%) patients). 27 Similar to the OPTIC trial, most TEAEs were GI related (nausea, vomiting, diarrhea) in both groups, and study drug discontinuation due to any adverse event was low (1.8% omadacycline vs 2.1% linezolid) and likely attributed to only mild or moderate nausea observed. In a subgroup analysis of patients with CKD stage 0/1, a TEAE was observed in 129 (50.6%) patients receiving omadacycline compared to 130 (48.7%) patients receiving linezolid. 31 In patients with CKD stage 2/3, 26 (38.8%) patients in the omadacycline group and 16 (30.8%) in the linezolid group experienced a TEAE. Omadacycline appears to be safe in patients with both normal and impaired renal function. Another subgroup analysis examined the impact of BMI on rates of drug‐related TEAE development. 30 Approximately 15 (12.5%) patients receiving omadacycline and 16 (14.8%) patients receiving linezolid experienced a drug‐related TEAE in the normal BMI group; whereas 24 (11.8%) omadacycline patients and 26 (12.1%) linezolid patients experienced a drug‐related TEAE in the high BMI group. In the OASIS‐2 trial, 201 (54.6%) patients in the omadacycline group and 140 (38.1%) patients in the linezolid group experienced at least one adverse event. 28 Of those patients who experienced an adverse event, 139 (37.8%) omadacycline patients and 52 (14.2%) linezolid patients experienced a drug‐related TEAE, with only one linezolid patient experiencing a serious drug‐related TEAE. No participants in the omadacycline group experienced a serious drug‐related TEAE. In total, six (1.6%) omadacycline patients and three (0.8%) linezolid patients discontinued the study drug due to an adverse event. Liver function tests (LFTs) for omadacycline‐treated patients were examined during all three phase III clinical trials. In the OPTIC study, 14 (3.7%) and 8 (2.1%) patients had alanine aminotransferase (ALT) and aspartate aminotransferase (AST) elevations, respectively; in the OASIS trial, 2.8% and 2.5% of patients had ALT and AST elevations, respectively; in the OASIS‐2 trial, 5.2% and 4.6% of patients had ALT and AST elevations, respectively. 25 , 27 , 28 Overall, few patients had LFT elevations greater than 3 times the upper limit of normal (ALT 3.2%; AST 2.8%) or total bilirubin elevations greater than 1.5 times the upper limit of normal (1.3%) in all three studies. 25 , 27 , 28 Omadacycline's effect in patients with ESRD was also examined. Five of 16 (31.3%) patients with ESRD experienced a TEAE, with only one (6.3%) patient experiencing a TEAE that was considered to be due to the study drug. 19 Further safety testing was performed to determine the effect of omadacycline on QTc prolongation. In the OPTIC trial, 42 (11.9%) patients treated with omadacycline experienced an increase in QTc greater than 450 milliseconds, with 24 (6.8%) patients experiencing an increase greater than 450 milliseconds corresponding to a greater than 30 milliseconds change from their baseline QTc interval. 33 In addition, six (1.7%) patients experienced an increase in QTc greater than 500 milliseconds, with five (1.4%) patients experiencing an increase greater than 500 milliseconds corresponding to a greater than 30 milliseconds change from their baseline QTc interval. Within the OASIS trial, 23 (7.1%) patients treated with omadacycline experienced an increase in their QTc interval greater than 450 milliseconds, with only three (0.9%) patients in this group having a greater than 30 milliseconds change from baseline. 33 There was one (0.3%) patient who experienced a greater than 30 milliseconds change from baseline that resulted in a QTc interval greater than 500 milliseconds. In the OASIS‐2 trial, only four (1.1%) omadacycline‐treated patients experienced an increase in QTc greater than 450 milliseconds, and no patients had a recorded QTc greater than 500 milliseconds. 33 Conclusion Omadacycline is a novel aminomethylcycline with potent broad‐spectrum activity against infectious pathogens that frequently cause community‐acquired infections, including ABSSSI and CABP, with the potential for additional coverage against organisms displaying multidrug resistance. Omadacycline displays favorable pharmacokinetic profiles allowing for once daily dosing, penetration into the ELF, and lack of renal dosing adjustments. Results from phase III studies for CABP and ABSSSI are promising, demonstrating comparable efficacy to standard‐of‐care agents. In addition, omadacycline provides a similar safety profile to comparators with GI‐related side effects representing the most common adverse drug event with mild nausea and no C. difficile –associated diarrhea observed in clinical trials. In patients with recurrent C. difficile infections or β‐lactam allergies, omadacycline is a non fluoroquinolone option that should be considered for treatment of CABP or ABSSSIs.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2732510/

Use of Automated Ambulatory-Care Encounter Records for Detection of Acute Illness Clusters, Including Potential Bioterrorism Events

The advent of domestic bioterrorism has emphasized the need for enhanced detection of clusters of acute illness. We describe a monitoring system operational in eastern Massachusetts, based on diagnoses obtained from electronic records of ambulatory-care encounters. Within 24 hours, ambulatory and telephone encounters recording patients with diagnoses of interest are identified and merged into major syndrome groups. Counts of new episodes of illness, rates calculated from health insurance records, and estimates of the probability of observing at least this number of new episodes are reported for syndrome surveillance. Census tracts with unusually large counts are identified by comparing observed with expected syndrome frequencies. During 1996–1999, weekly counts of new cases of lower respiratory syndrome were highly correlated with weekly hospital admissions. This system complements emergency room- and hospital-based surveillance by adding the capacity to rapidly identify clusters of illness, including potential bioterrorism events. Methods The utility of diagnoses from automated ambulatory encounter data for detecting respiratory disease clusters has been described (8) . In this report, we extend the use of encounter data to produce daily surveillance summary reports covering a broad range of syndromes for use by public health officials and health-care providers. The encounter data come from an electronic medical record system used by Harvard Vanguard Medical Associates, a large multispecialty group practice, to record all ambulatory-care encounters, including telephone contacts, regular visits, and urgent-care encounters, but not emergency room visits. The practice serves approximately 250,000 members, representing approximately 10% of the population of eastern Massachusetts. The automated record system is a commercial product (Epicare; Epic Systems Corporation, Madison, Wisconsin; available from URL: http://www.epicsys.com ) used by many large medical groups. It represents a valuable source of surveillance data because it operates in real time (i.e., records are updated as information is entered). Additionally, to the extent that practices engage in some form of prepaid care, the population served can be explicitly enumerated; the surveillance report described below is restricted to approximately 175,000 members of Harvard Pilgrim Health Care, a principal health maintenance organization in the region. These persons constitute a defined population that receives essentially all its ambulatory care in this practice. Demographic information and addresses are available for all these persons. At the time of consultation, clinical diagnoses are assigned for each encounter by the clinician, who chooses from lists of terms on the encounter screen; essentially all episodes are coded by the end of the same day on which care is given. Although an unlimited number of codes can be chosen, approximately 90% of encounters have three or fewer codes assigned (8) , stored as ICD-9 codes. Each night, an extract is created of all encounters recorded in the previous 24 hours with any of >1,500 ICD-9 codes in any of the syndrome categories. The patient's temperature is also recorded along with the ICD-9 codes. Demographic data are merged with each record through a link to the patient's membership record. As a way of grouping insured persons into neighborhoods, the addresses of the insured plan members, obtained from the HMO's data, have been coded by Geographic Information System (Mapping Analytics, Rochester, NY) to determine the census tracts of their residences ( 9 , 10 ). Developing and Defining Syndromes Patient encounters are categorized into syndrome groups according to the ICD-9 codes assigned at the time of consultation. The surveillance software considers each encounter record in turn and merges related ICD-9 diagnosis codes into syndrome groups by using a modification of a provisional classification scheme developed as part of the ESSENCE project (5) . This scheme reduces the complexity of the ICD-9 into eight syndrome categories: coma/shock, neurologic, upper gastrointestinal, lower gastrointestinal, upper respiratory, lower respiratory, dermatologic, and sepsis/fever. We made two major modifications of the syndrome definitions: the number of ambulatory episodes in the coma/shock category was almost zero in 4 years of data we examined, so it was combined with the neurologic syndrome category. A new syndromic category representing diagnoses of Centers for Disease Control and Prevention (CDC) bioterrorism category A agents (11) (anthrax, botulism, plague, smallpox, tularemia, and hemorrhagic fever) is reported separately. We also added an additional influenza-like illness category, defined by the CDC sentinel surveillance definition of fever >37.8 o C plus cough and/or sore throat in the absence of a known cause (12) . An individual patient may have multiple encounters associated with a single episode of illness (e.g., initial consultation, consultation 1–2 days later for laboratory results, and follow-up consultation a few weeks later) (8) . To avoid double counting from this common pattern of ambulatory care, the first encounter for each patient within any single syndrome group is reported, but subsequent encounters with the same syndrome are not reported as new episodes until > 6 weeks has elapsed since the most recent encounter in the same syndrome. We have reported that grouping respiratory illness visits into episodes reduces the total number of events by 38% in this clinical setting (8) . This practice of grouping clinical encounters into episodes of illness occurs independently for different syndromes. For example, a patient could qualify for two different syndromes on a single visit if codes for cough (lower respiratory syndrome) and diarrhea (lower gastrointestinal syndrome) are assigned at the same visit; or a lower gastrointestinal syndrome episode could begin a few days after the start of a lower respiratory syndrome episode. Reporting Results A daily surveillance summary report ( Tables 1 and 2 ; Figure 1 ) was designed in collaboration with staff from the Massachusetts Department of Public Health and the medical group's administration, which has operated since it was implemented on October 25, 2001. The aim of the report was to identify any unusually large numbers of episodes of illness within the ambulatory-care system. The current version ( Table 1 ) shows new episode counts and rates (per 1,000 insured persons) for all syndromes combined and for each individual syndrome, during the previous day. Mean rates are presented for the same day of the week in the same month of the previous 2 years, as well as the statistical probability associated with these counts derived from a generalized linear mixed model (described in the Models and Analysis section) for the four most common syndromes. Table 1 Daily public health surveillance report of office visits with diagnoses corresponding to infection syndromes: summary report for Monday, March 4, 2002, Massachusetts Syndrome Rate/1,000 health plan members (no. of visits) a Probability b 1999 average rates for this weekday in the same month 2000 average rates for this weekday in the same month All combined 2.015 (328) 1.918 2.123 Upper respiratory 1.087 (177) 0.999 1.151 1.251 Lower respiratory 0.405 (66) 0.999 0.369 0.474 Upper gastrointestinal 0.166 (27) 0.999 0.094 0.110 Lower gastrointestinal 0.227 (37) 0.999 0.221 0.173 CNS/neurologic c 0.000 (0) 0.003 0.007 Dermatologic 0.012 (2) 0.023 0.022 Sepsis/fever 0.000 (0) 0.057 0.086 Influenza-like illness 0.117 (19) — — CDC bioterrorism category A Agents d 0.000 (0) 0 0 a Repeated visits within 6 weeks excluded. b Probability of at least this many episodes occurring at least once per year, when the data are adjusted for month, day of week, holidays, secular trend, and variability among census tracts c CNS, central nervous system; CDC, Centers for Disease Control and Prevention. d Anthrax, botulism, plague, smallpox, tularemia, and hemorrhagic fever. Table 2 Lower respiratory syndrome by census tract, Massachusetts: sample small area report for March 4, 2002 a Population center Census tract code Cases in tract Denominator in this tract No. of days between counts this extreme b Randolph 250214202 4 1,232 1 Brookline 250214006 2 730 1 Boston 250250902 1 136 1 Somerville 250173507 2 918 1 Boston 250250304 1 225 1 a No census tract had an unusual number of new lower respiratory syndrome episodes on that day. The five most extreme tracts are shown, plus all with counts not expected to occur more than once per month. Tracts with most extreme counts are compared with their own history. b Estimated number of days between counts this extreme in any of the 529 census tracts, when data are adjusted for this tract's unique characteristics, as well as month, day of week, holidays, and secular trend. Figure 1 Map of sample small area syndrome counts for Monday, March 4, 2002, showing the five census tracts with the most extreme probability values. Labels show name of town, census tract code (state and county prefixes have been removed), and number of cases for the 24 hours included in the report. Each day's report also includes a list and maps of the residence locations of cases with respiratory and gastrointestinal syndromes ( Tables 1 and 2 ; Figure 1 ). The list and the map both show the five census tracts in the region with the most improbably large number of new episodes, based on the statistical model described. Daily updates are disseminated to authorized persons through a password-protected area on a Secure Sockets Layer (SSL) ( 13 , 14 ) encrypted website. Models and Analysis For each syndrome, we used a generalized linear mixed model (GLMM) ( 15 – 17 ) to model the daily counts from local neighborhoods over a 4-year historical period. In our model, census tracts (CT) form the neighborhoods, but this unit can be extended easily to larger or smaller geographic units if desired. Sample SAS code is provided in Appendix 1. The model closely resembles logistic regression, so that the logit, log( ), is modeled as a linear function of some covariates: where i indexes units of analysis, and are covariates or predictors, and is often denoted as In the GLMM version of logistic regression, where is the binomial-distributed number of visits in CT i on day t , is the number of members living in that CT on that day, is the probability that any patient has had a visit with a diagnosis in the syndrome, is a set of covariates measured on CT i at time t , is a vector of fixed effects, and is a random effect distributed with mean 0 and variance . The model can be used to generate an estimate of by inverting the logit. In the model now in use, we include in an intercept, an indicator for 6 days of the week, indicators for 11 months of the year, an indicator of the day as regular or a national holiday, and a linear term for the secular time trend. In each case, the terms contribute significantly to the fit of the model (p 100 lower respiratory syndrome episodes were identified in health plan members during 1996–1999; these cases accounted for approximately 70% of all ambulatory lower respiratory syndrome episodes recorded in health plan members. Figure 3 Health plan membership by census tract in eastern Boston. Each census tract contains approximately 4,000 residents. The weekly numbers of these episodes in health plan members were compared with weekly hospital admissions for all residents (not limited to health plan members) of the same 120 zip codes. Hospital admission data with personal identifiers removed were obtained from the Massachusetts Division of Health Care Finance and Policy for the 3 years ending September 30, 1999. These records included only patients discharged from the hospital, so the final 3 weeks of the hospital admission data were truncated to minimize the "edge effect" from the period when patients may have been admitted but not yet discharged and thus were not included in the available data. Using the same procedure to group hospital discharge ICD-9 codes as was used for the ambulatory data, we identified all admissions from residents of the 120 zip codes who had a discharge diagnosis in the lower respiratory syndrome group. Hospitalizations were assigned to the date of admission. We compiled the number of ambulatory lower respiratory syndrome episodes and the number of hospital admissions for lower respiratory syndrome for each week for the 3 years ending September 30, 1999. Time-series plots were prepared to compare seasonal patterns in the two independent data sources, and Spearman rank correlations were calculated between weekly hospital admission counts and ambulatory care episodes in the same week, the previous week, and so on up to 6 weeks, by using SAS Proc CORR (18) . Developing and Defining Syndromes Patient encounters are categorized into syndrome groups according to the ICD-9 codes assigned at the time of consultation. The surveillance software considers each encounter record in turn and merges related ICD-9 diagnosis codes into syndrome groups by using a modification of a provisional classification scheme developed as part of the ESSENCE project (5) . This scheme reduces the complexity of the ICD-9 into eight syndrome categories: coma/shock, neurologic, upper gastrointestinal, lower gastrointestinal, upper respiratory, lower respiratory, dermatologic, and sepsis/fever. We made two major modifications of the syndrome definitions: the number of ambulatory episodes in the coma/shock category was almost zero in 4 years of data we examined, so it was combined with the neurologic syndrome category. A new syndromic category representing diagnoses of Centers for Disease Control and Prevention (CDC) bioterrorism category A agents (11) (anthrax, botulism, plague, smallpox, tularemia, and hemorrhagic fever) is reported separately. We also added an additional influenza-like illness category, defined by the CDC sentinel surveillance definition of fever >37.8 o C plus cough and/or sore throat in the absence of a known cause (12) . An individual patient may have multiple encounters associated with a single episode of illness (e.g., initial consultation, consultation 1–2 days later for laboratory results, and follow-up consultation a few weeks later) (8) . To avoid double counting from this common pattern of ambulatory care, the first encounter for each patient within any single syndrome group is reported, but subsequent encounters with the same syndrome are not reported as new episodes until > 6 weeks has elapsed since the most recent encounter in the same syndrome. We have reported that grouping respiratory illness visits into episodes reduces the total number of events by 38% in this clinical setting (8) . This practice of grouping clinical encounters into episodes of illness occurs independently for different syndromes. For example, a patient could qualify for two different syndromes on a single visit if codes for cough (lower respiratory syndrome) and diarrhea (lower gastrointestinal syndrome) are assigned at the same visit; or a lower gastrointestinal syndrome episode could begin a few days after the start of a lower respiratory syndrome episode. Reporting Results A daily surveillance summary report ( Tables 1 and 2 ; Figure 1 ) was designed in collaboration with staff from the Massachusetts Department of Public Health and the medical group's administration, which has operated since it was implemented on October 25, 2001. The aim of the report was to identify any unusually large numbers of episodes of illness within the ambulatory-care system. The current version ( Table 1 ) shows new episode counts and rates (per 1,000 insured persons) for all syndromes combined and for each individual syndrome, during the previous day. Mean rates are presented for the same day of the week in the same month of the previous 2 years, as well as the statistical probability associated with these counts derived from a generalized linear mixed model (described in the Models and Analysis section) for the four most common syndromes. Table 1 Daily public health surveillance report of office visits with diagnoses corresponding to infection syndromes: summary report for Monday, March 4, 2002, Massachusetts Syndrome Rate/1,000 health plan members (no. of visits) a Probability b 1999 average rates for this weekday in the same month 2000 average rates for this weekday in the same month All combined 2.015 (328) 1.918 2.123 Upper respiratory 1.087 (177) 0.999 1.151 1.251 Lower respiratory 0.405 (66) 0.999 0.369 0.474 Upper gastrointestinal 0.166 (27) 0.999 0.094 0.110 Lower gastrointestinal 0.227 (37) 0.999 0.221 0.173 CNS/neurologic c 0.000 (0) 0.003 0.007 Dermatologic 0.012 (2) 0.023 0.022 Sepsis/fever 0.000 (0) 0.057 0.086 Influenza-like illness 0.117 (19) — — CDC bioterrorism category A Agents d 0.000 (0) 0 0 a Repeated visits within 6 weeks excluded. b Probability of at least this many episodes occurring at least once per year, when the data are adjusted for month, day of week, holidays, secular trend, and variability among census tracts c CNS, central nervous system; CDC, Centers for Disease Control and Prevention. d Anthrax, botulism, plague, smallpox, tularemia, and hemorrhagic fever. Table 2 Lower respiratory syndrome by census tract, Massachusetts: sample small area report for March 4, 2002 a Population center Census tract code Cases in tract Denominator in this tract No. of days between counts this extreme b Randolph 250214202 4 1,232 1 Brookline 250214006 2 730 1 Boston 250250902 1 136 1 Somerville 250173507 2 918 1 Boston 250250304 1 225 1 a No census tract had an unusual number of new lower respiratory syndrome episodes on that day. The five most extreme tracts are shown, plus all with counts not expected to occur more than once per month. Tracts with most extreme counts are compared with their own history. b Estimated number of days between counts this extreme in any of the 529 census tracts, when data are adjusted for this tract's unique characteristics, as well as month, day of week, holidays, and secular trend. Figure 1 Map of sample small area syndrome counts for Monday, March 4, 2002, showing the five census tracts with the most extreme probability values. Labels show name of town, census tract code (state and county prefixes have been removed), and number of cases for the 24 hours included in the report. Each day's report also includes a list and maps of the residence locations of cases with respiratory and gastrointestinal syndromes ( Tables 1 and 2 ; Figure 1 ). The list and the map both show the five census tracts in the region with the most improbably large number of new episodes, based on the statistical model described. Daily updates are disseminated to authorized persons through a password-protected area on a Secure Sockets Layer (SSL) ( 13 , 14 ) encrypted website. Models and Analysis For each syndrome, we used a generalized linear mixed model (GLMM) ( 15 – 17 ) to model the daily counts from local neighborhoods over a 4-year historical period. In our model, census tracts (CT) form the neighborhoods, but this unit can be extended easily to larger or smaller geographic units if desired. Sample SAS code is provided in Appendix 1. The model closely resembles logistic regression, so that the logit, log( ), is modeled as a linear function of some covariates: where i indexes units of analysis, and are covariates or predictors, and is often denoted as In the GLMM version of logistic regression, where is the binomial-distributed number of visits in CT i on day t , is the number of members living in that CT on that day, is the probability that any patient has had a visit with a diagnosis in the syndrome, is a set of covariates measured on CT i at time t , is a vector of fixed effects, and is a random effect distributed with mean 0 and variance . The model can be used to generate an estimate of by inverting the logit. In the model now in use, we include in an intercept, an indicator for 6 days of the week, indicators for 11 months of the year, an indicator of the day as regular or a national holiday, and a linear term for the secular time trend. In each case, the terms contribute significantly to the fit of the model (p 100 lower respiratory syndrome episodes were identified in health plan members during 1996–1999; these cases accounted for approximately 70% of all ambulatory lower respiratory syndrome episodes recorded in health plan members. Figure 3 Health plan membership by census tract in eastern Boston. Each census tract contains approximately 4,000 residents. The weekly numbers of these episodes in health plan members were compared with weekly hospital admissions for all residents (not limited to health plan members) of the same 120 zip codes. Hospital admission data with personal identifiers removed were obtained from the Massachusetts Division of Health Care Finance and Policy for the 3 years ending September 30, 1999. These records included only patients discharged from the hospital, so the final 3 weeks of the hospital admission data were truncated to minimize the "edge effect" from the period when patients may have been admitted but not yet discharged and thus were not included in the available data. Using the same procedure to group hospital discharge ICD-9 codes as was used for the ambulatory data, we identified all admissions from residents of the 120 zip codes who had a discharge diagnosis in the lower respiratory syndrome group. Hospitalizations were assigned to the date of admission. We compiled the number of ambulatory lower respiratory syndrome episodes and the number of hospital admissions for lower respiratory syndrome for each week for the 3 years ending September 30, 1999. Time-series plots were prepared to compare seasonal patterns in the two independent data sources, and Spearman rank correlations were calculated between weekly hospital admission counts and ambulatory care episodes in the same week, the previous week, and so on up to 6 weeks, by using SAS Proc CORR (18) . Models and Analysis For each syndrome, we used a generalized linear mixed model (GLMM) ( 15 – 17 ) to model the daily counts from local neighborhoods over a 4-year historical period. In our model, census tracts (CT) form the neighborhoods, but this unit can be extended easily to larger or smaller geographic units if desired. Sample SAS code is provided in Appendix 1. The model closely resembles logistic regression, so that the logit, log( ), is modeled as a linear function of some covariates: where i indexes units of analysis, and are covariates or predictors, and is often denoted as In the GLMM version of logistic regression, where is the binomial-distributed number of visits in CT i on day t , is the number of members living in that CT on that day, is the probability that any patient has had a visit with a diagnosis in the syndrome, is a set of covariates measured on CT i at time t , is a vector of fixed effects, and is a random effect distributed with mean 0 and variance . The model can be used to generate an estimate of by inverting the logit. In the model now in use, we include in an intercept, an indicator for 6 days of the week, indicators for 11 months of the year, an indicator of the day as regular or a national holiday, and a linear term for the secular time trend. In each case, the terms contribute significantly to the fit of the model (p 100 lower respiratory syndrome episodes were identified in health plan members during 1996–1999; these cases accounted for approximately 70% of all ambulatory lower respiratory syndrome episodes recorded in health plan members. Figure 3 Health plan membership by census tract in eastern Boston. Each census tract contains approximately 4,000 residents. The weekly numbers of these episodes in health plan members were compared with weekly hospital admissions for all residents (not limited to health plan members) of the same 120 zip codes. Hospital admission data with personal identifiers removed were obtained from the Massachusetts Division of Health Care Finance and Policy for the 3 years ending September 30, 1999. These records included only patients discharged from the hospital, so the final 3 weeks of the hospital admission data were truncated to minimize the "edge effect" from the period when patients may have been admitted but not yet discharged and thus were not included in the available data. Using the same procedure to group hospital discharge ICD-9 codes as was used for the ambulatory data, we identified all admissions from residents of the 120 zip codes who had a discharge diagnosis in the lower respiratory syndrome group. Hospitalizations were assigned to the date of admission. We compiled the number of ambulatory lower respiratory syndrome episodes and the number of hospital admissions for lower respiratory syndrome for each week for the 3 years ending September 30, 1999. Time-series plots were prepared to compare seasonal patterns in the two independent data sources, and Spearman rank correlations were calculated between weekly hospital admission counts and ambulatory care episodes in the same week, the previous week, and so on up to 6 weeks, by using SAS Proc CORR (18) . Results Data from an example of the summary report, one of the syndrome census tract reports, and the corresponding map for March 4, 2002, are shown in Table 1 , Table 2 and Figure 1 , respectively. The overall counts were all well within model-based expectations for this time of year, so the associated probabilities were all close to 1; at the level of census tracts, all counts are common enough to be expected daily. Figure 2 shows daily rates of new episodes of influenza-like illness and lower respiratory syndromes. Day-to-day variation is marked, especially on weekends, as is the expected winter increase in rates. Holidays such as New Year's Day have the lowest rates of reported illness. Figure 2 Daily incidence rates of lower respiratory and influenza-like illness after December 17, 2001, showing that within-week variation is substantially greater than seasonal variation. The sensitivity of the statistical model in the face of this extreme day-to-day and seasonal variation is illustrated in Table 3 . As few as three cases among health plan members may constitute an event predicted by the GLMM to occur less often than once per year, depending on the day of week and the month of the year. Table 3 Number of episodes of lower respiratory syndrome that would be expected to occur only once a month and once a year, based on a generalized linear mixed model (GLMM), in a representative eastern Massachusetts census tract a Month Day of week No. needed for once per month event No. needed for once per year event January Monday 5 6 January Tuesday 5 6 January Wednesday 5 6 January Thursday 5 6 January Friday 5 5 January Saturday 4 4 January Sunday 4 4 April Monday 4 5 April Tuesday 4 5 April Wednesday 4 5 April Thursday 4 5 April Friday 4 5 April Saturday 3 4 April Sunday 3 4 July Monday 4 5 July Tuesday 4 4 July Wednesday 4 4 July Thursday 4 4 July Friday 4 4 July Saturday 3 4 July Sunday 3 4 October Monday 5 6 October Tuesday 4 5 October Wednesday 4 5 October Thursday 4 5 October Friday 4 5 October Saturday 4 4 October Sunday 4 4 a This census tract has 491 health plan members and a random effect of 0.083, illustrating the effect of day of week and month of year for 2002. Visual inspection of the weekly counts of episodes in the ambulatory setting compared with hospital admissions shows congruent patterns, including pronounced winter peaks ( Figure 4 ). The data for admissions appear to lag behind data for ambulatory-care visits, most obviously for the winters of 1997 and 1999. Overall, weekly ambulatory-care episodes for lower respiratory illness were highly correlated with hospital admissions over the 3 years examined. The Spearman rank correlation between hospital admissions and ambulatory-care visits during the same week was 0.89. Correlating hospital admissions with ambulatory encounters from the previous week yielded a value of 0.90. Repeating this analysis, increasing the lag by 1 week at a time up to 6 weeks, yielded correlations of 0.92 at 2 weeks, 0.89 at 3 weeks, 0.85 at 4 weeks, 0.80 at 5 weeks, and 0.76 at 6 weeks. Figure 4 Weekly total ambulatory-care episodes of lower respiratory syndrome (broken line) and hospital admissions for lower respiratory syndrome (solid line) in Massachusetts for the 3 years from September 9, 1996, through September 9, 1999. The eligible population for the hospital data was the entire population of each zip code; the ambulatory care data came from a variable subset of each zip code. As a result, the number of hospital admissions was higher than the number of ambulatory-care episodes for parts of the period shown. Discussion The approach we have taken in the syndrome reporting system is to try to maximize the probability that any "signal" from the earliest stages of a bioterrorism or other public health event can be detected above the "noise" of normal clinical practice. The principal value of a syndromic surveillance system like the one described here is its ability to identify clusters of illness manifest by an unusual number of events, none of which individually differs appreciably from common respiratory, gastrointestinal, or other illnesses. Such nonspecific presentations might be the first sign of a widespread bioterrorism attack. They may also be the only routinely available clinical evidence of other important illness clusters, such as influenza or cryptosporidiosis, for which specific diagnostic tests are typically not performed. Even commonly available tests, such as x-rays, leukocyte counts, and sputum cultures are often not performed for lower respiratory illness with fever in an otherwise healthy patient, so syndromic surveillance can complement surveillance for individual cases of severe or unanticipated illness, which depend on detailed information about history, signs, symptoms, and diagnostic testing. Both syndromic and disease-specific surveillance systems are important components of any complete public health system. To provide the best possible opportunities for effective intervention, an ideal surveillance system should gather timely, valid, and inexpensive data from a sufficiently large proportion of the population to detect events of interest in the region, and then process and present it to public health personnel in a form that enables efficient decision making. Important elements of such a syndromic surveillance system exist in the automated data generated by health plans and other parts of the health-care delivery system as part of routine operations. The system described here meets many of these criteria because it results from collaboration between academic investigators, health-care providers, and public health officials. The automated medical records used here are well suited for surveillance of ambulatory-care encounters because the system is deeply integrated into the daily work of all clinicians and it is linked to both the provider payment and the membership systems. Although the data used in this system originate in a complete electronic medical record system, most of the syndromes are defined by diagnosis codes that are also available in other automated systems, including nurse hot lines and increasingly common same-day financial claims processing systems. Thus, several different kinds of data sources could contribute to an integrated surveillance network. Time-series plots ( Figure 4 ) provide some evidence that the data have validity as a measure of illness in the community, since the seasonal pattern is comparable with that in independently collected and validated hospital admission records for the same geographic region. The highest correlation, 0.92 at 2 weeks lag, implies that up to 85% of the variability of weekly hospital admission rates is predicted by variation in ambulatory-care admission levels 2 weeks earlier. Although the principal focus of this system is identifying unusual patterns of apparently common conditions, it also ensures prompt reporting of any encounter with a diagnosis suggestive of a CDC category A bioterrorism agent. In practice, any clinician making one of these diagnoses would be likely to report such a case separately, but there is almost no marginal cost to implement or run this additional surveillance component. The unadjusted counts and rates for each syndrome ( Table 1 ) may be most useful in responding to a very large and widespread bioterrorism event or identifying expected events such as the advent of influenza in a community. In these cases, statistical refinement is unnecessary because those monitoring the system will see substantially elevated rates. We believe statistical inference will be most useful when the signal from an event is weak or restricted to a small geographic region. Many syndromes have large seasonal fluctuations, such as the well-known winter peak for lower respiratory disease. Individual census tracts also show substantial variability in daily syndrome episode rates, possibly associated with demographic and socioeconomic differences. The statistical model adjusts daily expectations to account for important sources of variation, so those parts of the report based on statistical models take large "expected" seasonal increases in illness into account (e.g., Figure 4 ). The sensitivity of the resulting system ( Tables 1 and 2 ) in the face of expected variability appears to be much higher than more commonly advocated time-series based analytic approaches for public health surveillance (21) . Daily counts for each syndrome within single census tracts are usually zero, and as few as three to five health plan members affected would be unusual in a typical tract, depending on the month and day of week ( Tables 1 and 2 ). To allow a rapid assessment of the distribution of illness in the region, we highlight the five extreme census tract counts for each syndrome in our daily reports, even though there is nothing unusual in any census tract on most days. An alerting system could easily be triggered when there is a sufficiently unusual cluster for any syndrome. The thresholds can be different for different syndromes, and they can be adjusted to accommodate any desired frequency of alerts. For example, in the absence of a period of heightened alert, public health authorities may wish to be notified when the daily count of syndrome episodes within any census tract attains a level that would only be expected to occur within the entire catchment area once every three or more months. Thus, users will be able to adjust the notification system to suit their needs in terms of the preferred balance of false-positive alerts against the risk of false negatives (no alert in the presence of an actual event of interest). This kind of information, which is being developed as part of this project, could be a useful supplementary source for other public health surveillance systems. This reporting system includes strong protections of the privacy of individual patients' health records, since routine reports contain only aggregated information. Existing clinical and administrative security protocols that control statutory or other authorized access to confidential patient data will apply when follow-up is requested by public health authorities. One advantage of this system is that it takes advantage of the experience of ambulatory-care clinicians, who are likely to be among the first to encounter patients during the prodrome of any potential bioterrorism-related or other acute illness. In addition, the system imposes no additional reporting burden on clinicians, thus ensuring unbiased ascertainment of syndromes of interest that come to the attention of the practice. The data used here are already being gathered as part of the day-to-day practice of all participating clinicians. There is an initial cost for a system of this type because obtaining the data in a suitable form requires initial programming and testing, but subsequent processing requires relatively little additional expenditure and adds substantial value. All the technology used is widely available and inexpensive. While any simplification inevitably hides some potentially important detail, we believe that in addition to making the reports more comprehensible, grouping the ICD-9 codes decreases the impact of variation in coding practices. This effect is particularly important since the earliest manifestations of an outbreak may be nonspecific. The fact that syndromic surveillance focuses on unusual counts of common events means that detection of a signal may not be greatly influenced by intensity of diagnostic testing performed, completeness of documentation in the medical record, or variation between physicians or health-care systems in the use of diagnostic terminology or assignment of ICD-9 codes. For example, we have shown that >90% of lower respiratory illness episodes are represented by only three of the 119 ICD-9 diagnosis codes included in the lower respiratory illness syndrome (8) . As new ICD coding schemes are adopted, changes to the mapping used to translate code into syndrome will be required, but variation among tens of thousands of discrete individual codes is unlikely to have any major impact at the level of the broad syndromes used in our system. This emphasis on broad groupings of diagnoses also supports the notion that different data sources, including automated medical records, nurse call centers, and transaction data, might be combined into an integrated surveillance system. Because the focus is on the acute illness that prompts a medical encounter, we expect that the performance characteristics will not be seriously affected by differences between automated data systems, for instance, in the number of diagnoses captured or in the method of assigning diagnosis codes. However, experience with additional systems will be required to elucidate these issues. To the extent that different systems yield similar discrimination of events of interest, it will be possible to integrate them at the regional level, to improve overall sensitivity, and at the national level, to allow coherent surveillance of the entire population. While many types of data systems can contribute valuable surveillance information, appreciating the added value of more sophisticated data sources is also important. For instance, the availability of temperatures in the automated medical record system described here allows automated surveillance for influenza-like illness. The availability of automated laboratory test results and free text also provides opportunities to detect a wider array of conditions and to improve the specificity of detection of acute illness clusters. For example, anthrax surveillance might be limited to patients with fever and a lower respiratory illness syndrome. An additional noteworthy feature of surveillance systems such as this one is the fact that they need not cover the entire population to identify at least some clusters of interest. The minimum proportion of the population that must be under surveillance to detect clusters of different sizes has not been determined, but our coverage of 5%-10% of the population of the region appears to provide useful information. Although a small fraction of ambulatory-care practices uses automated medical records, the effective population that would be covered by surveillance systems based on these automated records is substantial, including many of the major population centers in the country. Combining information from these sites with other information sources, such as those maintained by health plans or by hospitals, would rapidly provide at least some monitoring capability for a much larger overall population. We can suggest additional methods for supplementing a surveillance system that counts syndromes encountered in ambulatory-care visits. Other sources of data, such as school and work absenteeism, over-the-counter medication sales, and even sales of products such as facial tissues and orange juice might contain potentially useful surveillance information. However, whether such data can be cheaply and efficiently gathered and processed and whether the data will yield valid and worthwhile signals remain to be demonstrated. Many aspects of the current system will be improved with experience. The development of standardized grouping of ICD codes into syndromes is a priority to allow uniform reporting. A great deal of work remains in developing statistical methods capable of detecting different types of illness clusters, ranging from acute, localized increases (for instance, due to release of a toxic chemical agent) to more slowly emerging, widespread conditions, as might be expected from contamination of a water supply. The implementation described here demonstrates that existing electronic data developed in the course of routine medical care by a wide array of providers and health plans can yield substantial improvements in current public health capabilities for assessment of bioterrorism and other acute illness clusters. Appendix 1 We used this SAS code in fitting the generalized linear mixed model (GLMM) that generates the parameter estimates used in our reports. This SAS code relies on the GLIMMIX macro (17) , which has been distributed by SAS (18) since version 6.12.%glimmix (data=test,procopt= noclprint covtest ,stmts=%str(class tract month dayofweek;model lri/pop=month dayofweek holiday day;random int/subject=tract solution type=un;),error=binomial); The data set is structured to contain a row for each day in the historical period for each census tract. In the code, lri is the variable that contains the count for census tract tract on day day . Pop contains the number of subjects in the tract on that day . Month is the month of the day , dayofweek is the day of week of the day , and holiday indicates whether the day is a national holiday. Days are standardized to prevent numerical difficulties with computation.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5880166/

What Is the Predictive Value of Animal Models for Vaccine Efficacy in Humans?

Animal models have played a pivotal role in all stages of vaccine development. Their predictive value for vaccine effectiveness depends on the pathogen, the robustness of the animal challenge model, and the correlates of protection (if known). This article will cover key questions regarding bridging animal studies to efficacy trials in humans. Examples include human papillomavirus (HPV) vaccine in which animal protection after vaccination with heterologous prototype virus-like particles (VLPs) predicted successful efficacy trials in humans, and a recent approval of anthrax vaccine in accordance with the "Animal Rule." The establishment of animal models predictive of vaccine effectiveness in humans has been fraught with difficulties with low success rate to date. Challenges facing the use of animal models for vaccine development against Ebola and HIV will be discussed.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2643477/

The Adenylate Cyclase Toxins of Bacillus anthracis and Bordetella pertussis Promote Th2 Cell Development by Shaping T Cell Antigen Receptor Signaling

The adjuvanticity of bacterial adenylate cyclase toxins has been ascribed to their capacity, largely mediated by cAMP, to modulate APC activation, resulting in the expression of Th2–driving cytokines. On the other hand, cAMP has been demonstrated to induce a Th2 bias when present during T cell priming, suggesting that bacterial cAMP elevating toxins may directly affect the Th1/Th2 balance. Here we have investigated the effects on human CD4 + T cell differentiation of two adenylate cyclase toxins, Bacillus anthracis edema toxin (ET) and Bordetella pertussis CyaA, which differ in structure, mode of cell entry, and subcellular localization. We show that low concentrations of ET and CyaA, but not of their genetically detoxified adenylate cyclase defective counterparts, potently promote Th2 cell differentiation by inducing expression of the master Th2 transcription factors, c-maf and GATA-3. We also present evidence that the Th2–polarizing concentrations of ET and CyaA selectively inhibit TCR–dependent activation of Akt1, which is required for Th1 cell differentiation, while enhancing the activation of two TCR–signaling mediators, Vav1 and p38, implicated in Th2 cell differentiation. This is at variance from the immunosuppressive toxin concentrations, which interfere with the earliest step in TCR signaling, activation of the tyrosine kinase Lck, resulting in impaired CD3ζ phosphorylation and inhibition of TCR coupling to ZAP-70 and Erk activation. These results demonstrate that, notwithstanding their differences in their intracellular localization, which result in focalized cAMP production, both toxins directly affect the Th1/Th2 balance by interfering with the same steps in TCR signaling, and suggest that their adjuvanticity is likely to result from their combined effects on APC and CD4 + T cells. Furthermore, our results strongly support the key role of cAMP in the adjuvanticity of these toxins. Introduction Development of an effective humoral immune response is crucially dependent on T cell help. The last step of B cell differentiation, involving immunoglobulin affinity maturation and isotype switching, occurs in peripheral lymphoid organs under the guidance of a specialized CD4 + T cell subset, known as T helper 2 (Th2). These cells provide both soluble (IL-4) and membrane-bound (CD40L) factors essential for terminal differentiation of antigen specific B cells [1] . Th2 cells are characterized by expression of a unique complement of cytokines, including IL-4, IL-5, IL-10 and IL-13, which are expressed through a complex transcriptional program involving chromatin remodelling at the Th2 cytokine locus control region and de novo expression of the lineage specific transcription factors c-maf and GATA-3 [2] . Priming the Th2 differentiation program in naive CD4 + T cells requires essential cues which are provided by antigen presenting cells (APC) in the form of cytokines. Engagement of the T cell antigen receptor (TCR) on naive T cells in the presence of IL-4 promotes their differentiation to Th2 effector cells, whilst simultaneously antagonising committment to the alternative Th1 lineage, which controls cell mediated immunity [1] , [2] . Additional factors present during T cell priming may profoundly affect the developmental program of helper T cells. Among these, of paramount importance is the second messenger cAMP, which is produced by cellular adenylate cyclases in response to heterotrimeric G-protein coupled surface receptors, such as the receptors for prostaglandin E 2 , a proinflammatory prostanoid produced by activated APC [3] . cAMP has been shown to favour Th2 cell differentiation and GATA-3 dependent production of IL-4 and IL-5 through a pathway regulated by phosphoinositide-dependent kinase 1 (PDK1) and protein kinase A (PKA) [4] – [9] . Suppression of both innate and adaptive immune responses through elevation of intracellular cAMP to supraphysiological levels represents a powerful strategy of immune evasion by many bacterial pathogens. This can be achieved indirectly, as for the bacterial enterotoxins, cholera toxin (CT) and E. coli heat-labile enterotoxin (LT), which enhance intracellular cAMP production by activating the Gsα subunit of heterotrimeric G-proteins coupled to cellular adenylate cyclases [10] . Alternatively, bacteria such as B. anthracis or B. pertussis produce and deliver into target cells an adenylate cyclase toxin, the edema factor (EF) and CyaA respectively, respectively, which are themselves adenylate cyclases that catalyze the production of large amounts of cAMP [11] , [12] . Notwithstanding their immmunosuppressive activity, when administered to mice at subtoxic concentrations together with antigen these toxins potentiate antibody responses, an effect associated with enhanced generation of antigen specific Th2 cells [13] – [17] . The adjuvanticity of cAMP elevating toxins is believed to result from their capacity to modulate APC differentiation and function. This is exemplified by ET and CyaA, which have been reported to selectively inhibit the production by macrophages and dendritic cells of the master Th1 polarizing cytokine, IL-12, while upregulating IL-4 and IL-10 production, thereby enhancing the induction of Th2 cells [13] – [15] , [17] – [19] . The finding that both non-hydrolysable cAMP analogues and PGE 2 evoke similar effects on APC [18] – [21] strongly supports the notion that the cAMP elevating activity of these toxins largely accounts for their capacity to differentially affect cytokine production by APC. We and others have demonstrated that ET and CyaA potently suppress T cell activation [22] – [24] . This activity results from their capacity to uncouple TCR engagement from activation of the MAP kinase cascade, which is essential for the initiation of the transcriptional program governing T cell activation, proliferation and subsequent differentiation to armed effector cells. Here we have investigated the additional possibility, suggested by the instructive role of the cAMP/PKA axis in Th2 cell differentiation [4] – [9] , that these bacterial toxins might alter TCR signaling to promote naive T cell committment to the Th2 lineage when used at low concentrations. The results show that both ET and CyaA, but not their enzymatically deficient counterparts, directly affect the Th1/Th2 balance by selectively inhibiting TCR dependent activation of the Th1 driving kinase Akt1 while enhancing activation of two essential components of the TCR signaling cascade selectively implicated in human Th2 cell differentiation, the guanine nucleotide exchanger Vav1 and the stress kinase p38. These data support the notion that the adjuvanticity of these cAMP elevating toxins results from their combined effects on APC and CD4 + T cells. Results Low concentrations of ET and CyaA catalyze sustained production of cAMP and induce PKA activation in T cells Both B. anthracis ET and B. pertussis CyaA potently suppress T cell activation through their cAMP elevating activity [22] – [24] . To assess the potential effect of these toxins on CD4 + T cell differentiation, a permissive concentration of ET or CyaA was identified in a T cell proliferation assay. Genetically inactivated mutants of ET and CyaA, EL1 [25] and CyaA-E5 [26] respectively, were included as controls. Both ET and CyaA inhibited T cell proliferation in a concentration-dependent manner ( Figure 1A ). Conversely, neither EL1 or CyaA-E5 affected T cell proliferation ( Figure 1B ). The toxin concentration selected for the polarization experiments, 0.11 nM and 0.28 nM for ET and CyaA respectively, resulted in ∼40% inhibition in the proliferation assays (see arrow in Figure 1A ). 10.1371/journal.ppat.1000325.g001 Figure 1 Concentration dependent suppression of T cell proliferation by CyaA and ET. (A) [ 3 H]-thymidine uptake by PBL stimulated for 48 h by CD3 cross-linking in the presence or absence of the indicated concentrations of CyaA ( top ) or ET ( bottom ). The results, obtained on triplicate samples of PBL from 5 independent donors, are expressed as % [ 3 H]-thymidine uptake (cpm) by CyaA or ET treated cells compared to control cells stimulated in the absence of either toxin (taken as 100%). The arrow shows the toxin concentration selected for the polarization experiments (CyaA, 0.28 nM; ET, 0.11 nM). (B) [ 3 H]-thymidine uptake by PBL stimulated for 48 h by CD3 cross-linking in the presence or absence of either CyaA and ET or the respective adenylase deficient mutants (45 nM CyaA/CyaA-E5, 110 nM ET/EL1). The results are expressed as in A. *** P ≤0.001; ** P ≤0.01; * P ≤0.05. Error bars, SD. A time course analysis of intracellular cAMP production in purified peripheral blood T cells showed that these concentrations of ET and CyaA induced a detectable accumulation of intracellular cAMP, albeit at much lower levels as compared to that obtained with high toxin concentrations ( Figure 2A, top ). The kinetics of cAMP production by ET and CyaA were significantly different. A fully immunosuppressive ET concentration resulted in a slow increase in intracellular cAMP beginning from 2 h, with a further progressive rise up to 8 h ( Figure 2A, top left ). On the other hand, a fully immunosuppressive concentration of CyaA evoked a rapid rise of cAMP to plateau levels beginning from the earliest time point analyzed, and the levels of cAMP remained high up to 8 h ( Figure 2A, top left ). High cAMP concentrations were still measurable after 24 h (data not shown). The kinetics of cAMP production by immunosuppressive concentrations of ET and CyaA were largely reproduced by the low toxin concentrations selected for the studies on T cell polarization ( Figure 2A, top right , and data not shown for 24 h). No increase in cAMP was elicited by EL1 and CyaA-E5, even at the highest concentration used ( Figure 2A, bottom left ). 10.1371/journal.ppat.1000325.g002 Figure 2 cAMP production and PKA activation in T cells treated with high and low concentrations of CyaA and ET. (A) Time course analysis of cAMP production in purified peripheral blood T lymphocytes treated with high (CyaA hi, 45 nM ; ET hi, 110 nM) ( top left ) or low (CyaA lo, 0.28 nM; ET lo, 0.11 nM) ( top right ) concentrations of CyaA or ET, or activated by TCR/CD3 cross-linking ( bottom right ). The histogram on the bottom left panel also includes the quantification of cAMP in lysates of T cells treated with the adenylase cyclase deficient CyaA and ET mutants (45 nM CyaA-E5, 110 nM EL1) for 2 h or 6 h, respectively. The results, which show the levels of cAMP measured in T cell lysates, are expressed as fmoles/10 6 cells. Representative experiments, each carried out on duplicate samples from individual healthy donors, are shown ( n ≥4). (B) Top , Immunoblot analysis of the phosphorylation state of PKA substrates in post-nuclear supernatants of T cells treated with 45 nM CyaA (CyaA hi) or 0.28 nM CyaA (CyaA lo), or 110 nM ET (ET hi) or 0.11 nM ET (ET lo), for 2 h (CyaA) or 6 h (ET), and then lysed as such or after stimulation for 1 min with anti-CD3 mAb (CD3). A sample stimulated with anti-CD3 mAb alone was also included. The immunoblot was carried out using an antibody which recognizes a phosphorylated PKA consensus sequence (see Materials and Methods ). The stripped filter was reprobed with a phosphospecific antibody which recognizes the active form of CREB (middle) . The fold activation of CREB in CyaA/ET treated samples vs untreated control in the experiment shown was the following: CyaA low, 8.3; CyaA high, 19.0; ET low, 8.7; ET high, 79.9. The levels of phospho-CREB in the samples treated with CyaA or ET in combination with anti-CD3 mAb vs samples treated with anti-CD3 mAb alone (taken as 100%) were the following: CyaA low+CD3, 98.1±4.8%; CyaA high+CD3, 103.4±8.7%; ET low+CD3, 102.1±3.2%; ET high+CD3, 112.1±8.1% ( n = 3). A control anti-actin blot is shown below. None of the treatments modified the expression levels of CREB (data not shown). Representative experiments are presented ( n ≥3). The migration of molecular mass markers is indicated. (C) Quantification of CREB phosphorylation in post-nuclear supernatants of T cells treated with 45 nM CyaA (CyaA hi)/CyaA-E5 or 0.28 nM CyaA (CyaA lo), or 110 nM ET (ET hi) /EL1 or 0.11 nM ET (ET lo), for 2 h (CyaA) or 6 h (ET). Where indicated, cells were pretreated for 1 h with 20 µM H89. A sample stimulated for 30 min with 100 µM 8-CPT was included as positive control. The data were obtained by laser densitometry of anti-phospho-CREB immunoblots. The results are expressed as relative CREB phosphorylation (fold activation vs untreated controls) ( n = 2). (D) Quantification of cAMP in lysates of T cells treated as in B. The results, which show the levels of cAMP measured in T cell lysates, are expressed as fmoles/10 6 cells. A representative experiment, carried out on duplicate samples from an individual healthy donor, is shown ( n = 3). *** P ≤0.001; ** P ≤0.01; * P ≤0.05. Error bars, SD. To understand whether the modest increase in the levels of cAMP catalyzed by low concentrations of ET or CyaA was sufficient to elicit a biological response, we measured the activity of PKA, one of the major cellular targets of cAMP. As a readout of PKA activation we used an antibody specific for the phosphorylated PKA consensus, R-X-X-pT-X-X/R-R-X-pS-X-X, which recognizes phosphorylated PKA substrates. The increase in intracellular cAMP following T cell treatment with high concentrations of ET or CyaA resulted in a strong potentiation of PKA activity, as shown by the qualitative and quantitative changes in the phosphoprotein pattern in lysates from toxin-treated cells compared to untreated cells ( Figure 2B, top panel ). A similar enhancement in PKA activity was also observed in cells treated with low concentrations of ET or CyaA, despite the smaller increase in intracellular cAMP measured under these conditions ( Figure 2B, top panel ). Interestingly, notwithstanding the different interacellular localization of the two adenylate cyclase toxins, there was a general overlap in the phosphoprotein pattern observed in cells treated with ET or CyaA. Consistent with the agonistic activity of ET and CyaA on PKA, analysis of the phosphorylation state of the transcriptional activator CREB, a specific PKA substrate, showed that low toxin concentrations induced CREB phosphorylation, albeit to a lesser extent compared to high toxin concentrations ( Figure 2B, middle panel ). The agonistic effect of the toxins was abrogated to a significant extent when cells were pretreated with pharmacological PKA inhibitors (H89 or KT5720) ( Figure 2C and data not shown). Moroever, no CREB phosphorylation was observed in T cells treated with the adenylate cyclase defective ET or CyaA mutant ( Figure 2C ), supporting the notion that the effects of the toxins are mediated by the cAMP/PKA axis. Of note, maximal activation of both PKA and CREB was observed in cells stimulated by TCR/CD3 cross-linking ( Figure 2B ), consistent with the potent agonistic activity of the receptor on cAMP production ( Figure 2A, bottom right ). However, as opposed to the long-lasting increase in cAMP elicited by the toxins, TCR engagement resulted in a transient increase in intracellular cAMP ( Figure 2A, bottom right ). No further significant enhancement in TCR-dependent PKA or CREB activation was observed in T cells treated with ET or CyaA ( Figure 2B and respective legend), despite the increase in cAMP production observed under these conditions ( Figure 2D ), indicating that PKA and CREB activation reaches plateau levels in response to the cAMP burst elicited by the TCR. ET and CyaA promote Th2 cell differentiation To assess the impact of low concentrations of ET and CyaA on human helper T cell polarization, enriched human CD4 + T cells from healthy donors were exposed to ET or CyaA and subsequently primed by TCR/CD3 cross-linking using immobilized anti-CD3 mAb. After 10 days, cells were washed, and restimulated for 24 h or 48 h using the same anti-CD3 mAb. The identity of the Th subset into which cells had differentiated was determined by ELISPOT analysis of cytokine production. As shown in Figure 3A , priming of cells that had been exposed to low concentrations of ET or CyaA resulted in a dramatic increase in production of the Th2 cytokines, IL-4 and IL-13, to levels close to those measured in cells primed to differentiate to the Th2 subset (TCR/CD3 cross-linking in the presence of IL-4). Conversely, consistent with the mutual antagonism of the Th1/Th2 differentiation programs, ET or CyaA had a modest inhibitory effect on production of the Th1 cytokines IFNγ and TNF-α ( Figure 3B ). No significant enhancement in Th2 cytokines above the levels produced by T cells primed in neutral conditions (TCR/CD3 cross-linking alone) was observed when cells were pretreated with the adenylate cyclase defective EL1 or CyaA-E5 mutants ( Figure 3A ), indicating that the Th2 driving activity of ET and CyaA is dependent on their capacity to produce cAMP. 10.1371/journal.ppat.1000325.g003 Figure 3 CyaA and ET promote Th2 cell differentiation through their cAMP elevating activity. Enriched CD4 + T cells from 6 healthy donors were primed with anti-CD3 mAb, as such or following pretreatment for 2 h with 0.28 nM CyaA/CyaA-E5, or with 0.11 nM ET/EL1. Cells primed in Th2- (IL-4) or Th1- (IL-12) inducing conditions were included as controls. After 10 days cells were washed and restimulated with anti-CD3 mAb for 48 and 24 h respectively, and the levels of IL-4 and IL-13 (A) and IFN-γ and TNF-α (B) were quantified by ELISPOT. The results, obtained on duplicate samples, are expressed as % spot-forming units by CyaA or ET treated cells compared to control cells primed in the absence of either toxin (taken as 100%). *** P ≤0.0001; ** P ≤0.001; * P ≤0.01. Error bars, SD. Differentiation of helper T cells to the Th2 subset is crucially dependent on expression of the lineage specific transcription factors c-maf and GATA-3, which are essential for transcriptional regulation of the Th2 cytokine control locus [2] . To understand whether ET and CyaA promote production of IL-4 and IL-13 by shaping the transcriptional program triggered by the TCR in naive T cells, resulting in expression of lineage specific transcription factors, the levels of c-maf and GATA-3 mRNA in T cells primed in the presence of either toxin were measured by real-time RT-PCR. Both ET and CyaA potently upregulated expression of c-maf and GATA-3 to levels comparable or higher than those detectable in T cells primed in the presence of IL-4 ( Figure 4A ). Conversely, expression of the Th1 lineage specific transcription factor T-bet was not significantly affected in T cells primed in the presence of either CyaA or ET ( Figure 4B ). 10.1371/journal.ppat.1000325.g004 Figure 4 CyaA and ET promote c-maf and GATA-3 expression in primed T cells. Enriched CD4 + T cells from 3 healthy donors were primed with anti-CD3 mAb, as such or following pretreatment for 2 h with 0.28 nM CyaA or 0.11 nM ET. Cells primed in Th2- (IL-4) or Th1- (IL-12) inducing conditions were included as controls. After 10 days cells were restimulated with anti-CD3 mAb for 24 h. The levels of mRNA encoding c-maf and GATA-3 (A) and T-bet (B) were quantified by real-time RT–PCR. Transcript levels were normalized to the expression level of GAPDH. Syber green runs were performed with cDNAs from the same reverse transcription reaction from 400 ng of total RNA. The ΔΔC T method was applied as a comparative method of quantification, using cells primed in neutral conditions (anti-CD3 mAb) as reference. The data are representative of 3 independent experiments, each in duplicate. *** P ≤0.00001; ** P ≤0.0001; * P ≤0.001. Error bars, SD. Low concentrations of ET and CyaA selectively impair TCR–dependent activation of Akt1, while potentiating activation of Vav1 and p38 TCR signaling is initiated by Lck, a T cell specific Src family protein tyrosine kinase which is responsible for phosphorylation of the ITAMs within the ζ chain of the CD3 complex. As all Src kinases, Lck is negatively regulated by a C-terminal tyrosine residue, Y505, which, when phosphorylated, establishes an intramolecular interaction with the SH2 domain, resulting in a close, inactive conformation. The inhibitory tyrosine residue is phosphorylated by Csk, which in resting cells is maintained close to Lck in lipid rafts through interaction with the PAG adaptor and whose activity is potentiated by PKA dependent phosphorylation of a serine residue at position 364 [27] . By elevating intracellular cAMP, ET and CyaA have therefore the potential to antagonize TCR signaling beginning from the earliest step. Analysis of the phosphorylation state of Y505 on Lck using a phosphospecific antibody revealed that high concentrations of ET or CyaA effectively block TCR dependent dephosphorylation of Lck ( Figure 5A and 5B ). This activity was not reproduced by the respective adenylate cyclase defective mutants ( Figure 5B ), supporting the notion that the suppressive effect of ET and CyaA on TCR dependent Lck activation is mediated by cAMP. No enhancement in Lck kinase activity in response to TCR engagement was moreover observed when cells were pretreated with high concentrations of ET or CyaA, as assessed by measuring Lck autophosphorylation in in vitro kinase assays (data not shown). Consistent with the failure of the TCR to trigger activation of Lck in the presence of either toxin, both TCR dependent CD3ζ phosphorylation and activation of the effector kinase ZAP-70, which occurs following recruitment to the phosphorylated ITAMs of CD3ζ, were found to be inhibited by ET or CyaA ( Figure 5A and 5B ). In agreement with previous reports [22] – [24] , activation of the MAP kinase cascade, which couples these early signaling events to gene transcription, was found to be impaired by immunosuppressive concentrations of ET or CyaA, as assessed using as a readout phosphorylation of Erk1/2 ( Figure 5A and 5B ). Conversely, neither CD3ζ and ZAP-70 phosphorylation, nor Erk1/2 phosphorylation, were affected when the TCR was stimulated in cells exposed to low, Th2 polarizing concentrations of ET or CyaA ( Figure 6A and 6B ). 10.1371/journal.ppat.1000325.g005 Figure 5 Immunosuppressive concentrations of CyaA or ET prevent initiation of TCR signaling. (A) Top left , Immunoblot analysis, using a phosphospecific antibody, of Lck phosphorylation on the inhibitory C-terminal tyrosine residue (Y505) in postnuclear supernatants from PBL activated for 1 min by CD3 cross-linking in the presence or absence of either 45 nM CyaA or 110 nM ET (CyaA hi, ET hi). Top right , Immunoblot analysis, using an anti-phosphotyrosine antibody, of CD3ζ specific immunoprecipitates from PBL treated as above. Bottom , Immunoblot analysis, using phosphospecific antibodies, of ZAP-70 (left) and Erk1/2 (right) phosphorylation in postnuclear supernatants from PBL activated for 1 min (ZAP-70) or 5 min (Erk1/2) by CD3 cross-linking in the presence or absence of 45 nM CyaA or 110 nM ET (CyaA hi, ET hi). (B) Quantification by laser densitometry of the relative levels of Lck (phosphorylation in unstimulated cells taken as 100%), or CD3ζ, ZAP-70 and Erk1/2 phosphorylation (phosphorylation in anti-CD3 stimulated cells taken as 100%, indicated as a dotted line) in PBL activated by CD3 cross-linking in the presence of 45 nM CyaA or 110 nM ET (CyaA hi, ET hi) ( n ≥3). Where indicated, cells were activated in the presence of the adenylase cyclase deficient CyaA and ET mutants (45 nM CyaA-E5, 110 nM EL1) for 2 h or 6 h, respectively ( n = 2). *** P ≤0.001; ** P ≤0.01; * P ≤0.05. Error bars, SD. 10.1371/journal.ppat.1000325.g006 Figure 6 Low CyaA or ET concentrations do not affect initiation of TCR signaling. (A) Immunoblot analysis, using phosphospecific antibodies, of CD3ζ, ZAP-70, or Erk1/2 phosphorylation in postnuclear supernatants of cells activated as above in the presence or absence of either 0.28 nM CyaA or 0.11 nM ET (CyaA lo, ET lo). Filters were stripped and re-probed with control antibodies. Representative experiments are shown ( n ≥3). The migration of molecular mass markers is indicated. (B) Quantification by laser densitometry of the relative levels of CD3ζ, ZAP-70, and Erk1/2 phosphorylation (phosphorylation in anti-CD3 stimulated cells taken as 100%, indicated as a dotted line) in PBL activated by CD3 cross-linking in the presence or absence of either 0.28 nM CyaA or 0.11 nM ET (CyaA lo, ET lo) ( n ≥3). Error bars, SD. To understand whether Th2 polarizing concentrations of ET and CyaA could selectively affect downstream components of TCR signaling specifically implicated in Th lineage committment, we focused on two molecules in the TCR signaling cascade, the Rac/Cdc42 specific guanine nucleotide exchanger Vav1 and the stress-activated kinase p38, which have been implicated in human Th2 cell differentiation [28] – [32] . Furthermore, we assessed the effect of the toxins on the serine/threonine kinase Akt1, which has been associated to Th1 cell differentiation [9] . Strikingly, analysis of Akt1 activation using phosphospecific antibodies which recognize two critical residues, T308 and S473, showed that low, Th2 polarizing concentrations of ET or CyaA were sufficient to potently impair TCR dependent Akt phosphorylation ( Figure 7A and data not shown). Conversely, both basal and TCR dependent Vav1 phosphorylation on Y174, which positively regulates Vav1 activity, was potentiated by low concentrations of ET or CyaA ( Figure 7B ). A similar enhancement was observed for p38 ( Figure 7B ), consistent with the capacity of PKA to act as an agonist of this kinase [31] , [33] . The phosphodiesterase inhibitor, IBMX, further potentiated the agonistic activity of the toxins on p38 activation (data not shown), further supporting the notion that the effects of the toxins are mediated by cAMP. Hence ET and CyaA alter the Th1/Th2 balance at least in part by antagonizing the Akt1 dependent pathway leading to Th1 cell differentiation and by potentiating the Vav1 and p38 dependent pathway(s) leading to Th2 cell differentiation. Of note, TCR dependent Vav1 and p38 activation was not impaired, but actually enhanced, when cells were pretreated with high concentrations of ET or CyaA ( Figure 7B ), despite their potent inhibitory activity on initiation of TCR signaling, suggesting that an Lck independent pathway triggered by the TCR, which can be potentiated by cAMP, may contribute to a significant extent to their activation. 10.1371/journal.ppat.1000325.g007 Figure 7 Low CyaA or ET concentrations impair TCR–dependent Akt1 phosphorylation while enhancing Vav1 and p38 phosphorylation. (A) Immunoblot analysis, using a phosphospecific antibody, of Akt1 activation in postnuclear supernatants from PBL activated for 5 min by CD3 cross-linking in the presence or absence of 45 nM CyaA or 110 nM ET (CyaA hi, ET hi) or alternatively in the presence or absence of 0.28 nM CyaA or 0.11 nM ET (CyaA lo, ET lo). A representative experiment is shown ( n = 3). (B) Immunoblot analysis, using phosphospecific antibodies, of Vav1 ( top ) or p38 ( bottom ) activation in postnuclear supernatants from PBL activated for 1 min (Vav) or 5 min (p38) by CD3 cross-linking in the presence or absence of 45 nM CyaA or 110 nM ET (CyaA hi, ET hi) or alternatively in the presence or absence of 0.28 nM CyaA or 0.11 nM ET (CyaA lo, ET lo). Filters were stripped and re-probed with control antibodies. Representative experiments are shown ( n ≥4). The migration of molecular mass markers is indicated. The graphs on the right of the immunoblots show the quantification by laser densitometry of the relative levels of Akt1, Vav1, and p38 phosphorylation (phosphorylation in anti-CD3 stimulated cells taken as 100%) in PBL activated by CD3 cross-linking in the presence or absence of 45 nM CyaA or 110 nM ET (CyaA hi, ET hi) or alternatively in the presence or absence of 0.28 nM CyaA or 0.11 nM ET (CyaA lo, ET lo). *** P ≤0.001** P ≤0.01; * P ≤0.05. Error bars, SD. Low concentrations of ET and CyaA catalyze sustained production of cAMP and induce PKA activation in T cells Both B. anthracis ET and B. pertussis CyaA potently suppress T cell activation through their cAMP elevating activity [22] – [24] . To assess the potential effect of these toxins on CD4 + T cell differentiation, a permissive concentration of ET or CyaA was identified in a T cell proliferation assay. Genetically inactivated mutants of ET and CyaA, EL1 [25] and CyaA-E5 [26] respectively, were included as controls. Both ET and CyaA inhibited T cell proliferation in a concentration-dependent manner ( Figure 1A ). Conversely, neither EL1 or CyaA-E5 affected T cell proliferation ( Figure 1B ). The toxin concentration selected for the polarization experiments, 0.11 nM and 0.28 nM for ET and CyaA respectively, resulted in ∼40% inhibition in the proliferation assays (see arrow in Figure 1A ). 10.1371/journal.ppat.1000325.g001 Figure 1 Concentration dependent suppression of T cell proliferation by CyaA and ET. (A) [ 3 H]-thymidine uptake by PBL stimulated for 48 h by CD3 cross-linking in the presence or absence of the indicated concentrations of CyaA ( top ) or ET ( bottom ). The results, obtained on triplicate samples of PBL from 5 independent donors, are expressed as % [ 3 H]-thymidine uptake (cpm) by CyaA or ET treated cells compared to control cells stimulated in the absence of either toxin (taken as 100%). The arrow shows the toxin concentration selected for the polarization experiments (CyaA, 0.28 nM; ET, 0.11 nM). (B) [ 3 H]-thymidine uptake by PBL stimulated for 48 h by CD3 cross-linking in the presence or absence of either CyaA and ET or the respective adenylase deficient mutants (45 nM CyaA/CyaA-E5, 110 nM ET/EL1). The results are expressed as in A. *** P ≤0.001; ** P ≤0.01; * P ≤0.05. Error bars, SD. A time course analysis of intracellular cAMP production in purified peripheral blood T cells showed that these concentrations of ET and CyaA induced a detectable accumulation of intracellular cAMP, albeit at much lower levels as compared to that obtained with high toxin concentrations ( Figure 2A, top ). The kinetics of cAMP production by ET and CyaA were significantly different. A fully immunosuppressive ET concentration resulted in a slow increase in intracellular cAMP beginning from 2 h, with a further progressive rise up to 8 h ( Figure 2A, top left ). On the other hand, a fully immunosuppressive concentration of CyaA evoked a rapid rise of cAMP to plateau levels beginning from the earliest time point analyzed, and the levels of cAMP remained high up to 8 h ( Figure 2A, top left ). High cAMP concentrations were still measurable after 24 h (data not shown). The kinetics of cAMP production by immunosuppressive concentrations of ET and CyaA were largely reproduced by the low toxin concentrations selected for the studies on T cell polarization ( Figure 2A, top right , and data not shown for 24 h). No increase in cAMP was elicited by EL1 and CyaA-E5, even at the highest concentration used ( Figure 2A, bottom left ). 10.1371/journal.ppat.1000325.g002 Figure 2 cAMP production and PKA activation in T cells treated with high and low concentrations of CyaA and ET. (A) Time course analysis of cAMP production in purified peripheral blood T lymphocytes treated with high (CyaA hi, 45 nM ; ET hi, 110 nM) ( top left ) or low (CyaA lo, 0.28 nM; ET lo, 0.11 nM) ( top right ) concentrations of CyaA or ET, or activated by TCR/CD3 cross-linking ( bottom right ). The histogram on the bottom left panel also includes the quantification of cAMP in lysates of T cells treated with the adenylase cyclase deficient CyaA and ET mutants (45 nM CyaA-E5, 110 nM EL1) for 2 h or 6 h, respectively. The results, which show the levels of cAMP measured in T cell lysates, are expressed as fmoles/10 6 cells. Representative experiments, each carried out on duplicate samples from individual healthy donors, are shown ( n ≥4). (B) Top , Immunoblot analysis of the phosphorylation state of PKA substrates in post-nuclear supernatants of T cells treated with 45 nM CyaA (CyaA hi) or 0.28 nM CyaA (CyaA lo), or 110 nM ET (ET hi) or 0.11 nM ET (ET lo), for 2 h (CyaA) or 6 h (ET), and then lysed as such or after stimulation for 1 min with anti-CD3 mAb (CD3). A sample stimulated with anti-CD3 mAb alone was also included. The immunoblot was carried out using an antibody which recognizes a phosphorylated PKA consensus sequence (see Materials and Methods ). The stripped filter was reprobed with a phosphospecific antibody which recognizes the active form of CREB (middle) . The fold activation of CREB in CyaA/ET treated samples vs untreated control in the experiment shown was the following: CyaA low, 8.3; CyaA high, 19.0; ET low, 8.7; ET high, 79.9. The levels of phospho-CREB in the samples treated with CyaA or ET in combination with anti-CD3 mAb vs samples treated with anti-CD3 mAb alone (taken as 100%) were the following: CyaA low+CD3, 98.1±4.8%; CyaA high+CD3, 103.4±8.7%; ET low+CD3, 102.1±3.2%; ET high+CD3, 112.1±8.1% ( n = 3). A control anti-actin blot is shown below. None of the treatments modified the expression levels of CREB (data not shown). Representative experiments are presented ( n ≥3). The migration of molecular mass markers is indicated. (C) Quantification of CREB phosphorylation in post-nuclear supernatants of T cells treated with 45 nM CyaA (CyaA hi)/CyaA-E5 or 0.28 nM CyaA (CyaA lo), or 110 nM ET (ET hi) /EL1 or 0.11 nM ET (ET lo), for 2 h (CyaA) or 6 h (ET). Where indicated, cells were pretreated for 1 h with 20 µM H89. A sample stimulated for 30 min with 100 µM 8-CPT was included as positive control. The data were obtained by laser densitometry of anti-phospho-CREB immunoblots. The results are expressed as relative CREB phosphorylation (fold activation vs untreated controls) ( n = 2). (D) Quantification of cAMP in lysates of T cells treated as in B. The results, which show the levels of cAMP measured in T cell lysates, are expressed as fmoles/10 6 cells. A representative experiment, carried out on duplicate samples from an individual healthy donor, is shown ( n = 3). *** P ≤0.001; ** P ≤0.01; * P ≤0.05. Error bars, SD. To understand whether the modest increase in the levels of cAMP catalyzed by low concentrations of ET or CyaA was sufficient to elicit a biological response, we measured the activity of PKA, one of the major cellular targets of cAMP. As a readout of PKA activation we used an antibody specific for the phosphorylated PKA consensus, R-X-X-pT-X-X/R-R-X-pS-X-X, which recognizes phosphorylated PKA substrates. The increase in intracellular cAMP following T cell treatment with high concentrations of ET or CyaA resulted in a strong potentiation of PKA activity, as shown by the qualitative and quantitative changes in the phosphoprotein pattern in lysates from toxin-treated cells compared to untreated cells ( Figure 2B, top panel ). A similar enhancement in PKA activity was also observed in cells treated with low concentrations of ET or CyaA, despite the smaller increase in intracellular cAMP measured under these conditions ( Figure 2B, top panel ). Interestingly, notwithstanding the different interacellular localization of the two adenylate cyclase toxins, there was a general overlap in the phosphoprotein pattern observed in cells treated with ET or CyaA. Consistent with the agonistic activity of ET and CyaA on PKA, analysis of the phosphorylation state of the transcriptional activator CREB, a specific PKA substrate, showed that low toxin concentrations induced CREB phosphorylation, albeit to a lesser extent compared to high toxin concentrations ( Figure 2B, middle panel ). The agonistic effect of the toxins was abrogated to a significant extent when cells were pretreated with pharmacological PKA inhibitors (H89 or KT5720) ( Figure 2C and data not shown). Moroever, no CREB phosphorylation was observed in T cells treated with the adenylate cyclase defective ET or CyaA mutant ( Figure 2C ), supporting the notion that the effects of the toxins are mediated by the cAMP/PKA axis. Of note, maximal activation of both PKA and CREB was observed in cells stimulated by TCR/CD3 cross-linking ( Figure 2B ), consistent with the potent agonistic activity of the receptor on cAMP production ( Figure 2A, bottom right ). However, as opposed to the long-lasting increase in cAMP elicited by the toxins, TCR engagement resulted in a transient increase in intracellular cAMP ( Figure 2A, bottom right ). No further significant enhancement in TCR-dependent PKA or CREB activation was observed in T cells treated with ET or CyaA ( Figure 2B and respective legend), despite the increase in cAMP production observed under these conditions ( Figure 2D ), indicating that PKA and CREB activation reaches plateau levels in response to the cAMP burst elicited by the TCR. ET and CyaA promote Th2 cell differentiation To assess the impact of low concentrations of ET and CyaA on human helper T cell polarization, enriched human CD4 + T cells from healthy donors were exposed to ET or CyaA and subsequently primed by TCR/CD3 cross-linking using immobilized anti-CD3 mAb. After 10 days, cells were washed, and restimulated for 24 h or 48 h using the same anti-CD3 mAb. The identity of the Th subset into which cells had differentiated was determined by ELISPOT analysis of cytokine production. As shown in Figure 3A , priming of cells that had been exposed to low concentrations of ET or CyaA resulted in a dramatic increase in production of the Th2 cytokines, IL-4 and IL-13, to levels close to those measured in cells primed to differentiate to the Th2 subset (TCR/CD3 cross-linking in the presence of IL-4). Conversely, consistent with the mutual antagonism of the Th1/Th2 differentiation programs, ET or CyaA had a modest inhibitory effect on production of the Th1 cytokines IFNγ and TNF-α ( Figure 3B ). No significant enhancement in Th2 cytokines above the levels produced by T cells primed in neutral conditions (TCR/CD3 cross-linking alone) was observed when cells were pretreated with the adenylate cyclase defective EL1 or CyaA-E5 mutants ( Figure 3A ), indicating that the Th2 driving activity of ET and CyaA is dependent on their capacity to produce cAMP. 10.1371/journal.ppat.1000325.g003 Figure 3 CyaA and ET promote Th2 cell differentiation through their cAMP elevating activity. Enriched CD4 + T cells from 6 healthy donors were primed with anti-CD3 mAb, as such or following pretreatment for 2 h with 0.28 nM CyaA/CyaA-E5, or with 0.11 nM ET/EL1. Cells primed in Th2- (IL-4) or Th1- (IL-12) inducing conditions were included as controls. After 10 days cells were washed and restimulated with anti-CD3 mAb for 48 and 24 h respectively, and the levels of IL-4 and IL-13 (A) and IFN-γ and TNF-α (B) were quantified by ELISPOT. The results, obtained on duplicate samples, are expressed as % spot-forming units by CyaA or ET treated cells compared to control cells primed in the absence of either toxin (taken as 100%). *** P ≤0.0001; ** P ≤0.001; * P ≤0.01. Error bars, SD. Differentiation of helper T cells to the Th2 subset is crucially dependent on expression of the lineage specific transcription factors c-maf and GATA-3, which are essential for transcriptional regulation of the Th2 cytokine control locus [2] . To understand whether ET and CyaA promote production of IL-4 and IL-13 by shaping the transcriptional program triggered by the TCR in naive T cells, resulting in expression of lineage specific transcription factors, the levels of c-maf and GATA-3 mRNA in T cells primed in the presence of either toxin were measured by real-time RT-PCR. Both ET and CyaA potently upregulated expression of c-maf and GATA-3 to levels comparable or higher than those detectable in T cells primed in the presence of IL-4 ( Figure 4A ). Conversely, expression of the Th1 lineage specific transcription factor T-bet was not significantly affected in T cells primed in the presence of either CyaA or ET ( Figure 4B ). 10.1371/journal.ppat.1000325.g004 Figure 4 CyaA and ET promote c-maf and GATA-3 expression in primed T cells. Enriched CD4 + T cells from 3 healthy donors were primed with anti-CD3 mAb, as such or following pretreatment for 2 h with 0.28 nM CyaA or 0.11 nM ET. Cells primed in Th2- (IL-4) or Th1- (IL-12) inducing conditions were included as controls. After 10 days cells were restimulated with anti-CD3 mAb for 24 h. The levels of mRNA encoding c-maf and GATA-3 (A) and T-bet (B) were quantified by real-time RT–PCR. Transcript levels were normalized to the expression level of GAPDH. Syber green runs were performed with cDNAs from the same reverse transcription reaction from 400 ng of total RNA. The ΔΔC T method was applied as a comparative method of quantification, using cells primed in neutral conditions (anti-CD3 mAb) as reference. The data are representative of 3 independent experiments, each in duplicate. *** P ≤0.00001; ** P ≤0.0001; * P ≤0.001. Error bars, SD. Low concentrations of ET and CyaA selectively impair TCR–dependent activation of Akt1, while potentiating activation of Vav1 and p38 TCR signaling is initiated by Lck, a T cell specific Src family protein tyrosine kinase which is responsible for phosphorylation of the ITAMs within the ζ chain of the CD3 complex. As all Src kinases, Lck is negatively regulated by a C-terminal tyrosine residue, Y505, which, when phosphorylated, establishes an intramolecular interaction with the SH2 domain, resulting in a close, inactive conformation. The inhibitory tyrosine residue is phosphorylated by Csk, which in resting cells is maintained close to Lck in lipid rafts through interaction with the PAG adaptor and whose activity is potentiated by PKA dependent phosphorylation of a serine residue at position 364 [27] . By elevating intracellular cAMP, ET and CyaA have therefore the potential to antagonize TCR signaling beginning from the earliest step. Analysis of the phosphorylation state of Y505 on Lck using a phosphospecific antibody revealed that high concentrations of ET or CyaA effectively block TCR dependent dephosphorylation of Lck ( Figure 5A and 5B ). This activity was not reproduced by the respective adenylate cyclase defective mutants ( Figure 5B ), supporting the notion that the suppressive effect of ET and CyaA on TCR dependent Lck activation is mediated by cAMP. No enhancement in Lck kinase activity in response to TCR engagement was moreover observed when cells were pretreated with high concentrations of ET or CyaA, as assessed by measuring Lck autophosphorylation in in vitro kinase assays (data not shown). Consistent with the failure of the TCR to trigger activation of Lck in the presence of either toxin, both TCR dependent CD3ζ phosphorylation and activation of the effector kinase ZAP-70, which occurs following recruitment to the phosphorylated ITAMs of CD3ζ, were found to be inhibited by ET or CyaA ( Figure 5A and 5B ). In agreement with previous reports [22] – [24] , activation of the MAP kinase cascade, which couples these early signaling events to gene transcription, was found to be impaired by immunosuppressive concentrations of ET or CyaA, as assessed using as a readout phosphorylation of Erk1/2 ( Figure 5A and 5B ). Conversely, neither CD3ζ and ZAP-70 phosphorylation, nor Erk1/2 phosphorylation, were affected when the TCR was stimulated in cells exposed to low, Th2 polarizing concentrations of ET or CyaA ( Figure 6A and 6B ). 10.1371/journal.ppat.1000325.g005 Figure 5 Immunosuppressive concentrations of CyaA or ET prevent initiation of TCR signaling. (A) Top left , Immunoblot analysis, using a phosphospecific antibody, of Lck phosphorylation on the inhibitory C-terminal tyrosine residue (Y505) in postnuclear supernatants from PBL activated for 1 min by CD3 cross-linking in the presence or absence of either 45 nM CyaA or 110 nM ET (CyaA hi, ET hi). Top right , Immunoblot analysis, using an anti-phosphotyrosine antibody, of CD3ζ specific immunoprecipitates from PBL treated as above. Bottom , Immunoblot analysis, using phosphospecific antibodies, of ZAP-70 (left) and Erk1/2 (right) phosphorylation in postnuclear supernatants from PBL activated for 1 min (ZAP-70) or 5 min (Erk1/2) by CD3 cross-linking in the presence or absence of 45 nM CyaA or 110 nM ET (CyaA hi, ET hi). (B) Quantification by laser densitometry of the relative levels of Lck (phosphorylation in unstimulated cells taken as 100%), or CD3ζ, ZAP-70 and Erk1/2 phosphorylation (phosphorylation in anti-CD3 stimulated cells taken as 100%, indicated as a dotted line) in PBL activated by CD3 cross-linking in the presence of 45 nM CyaA or 110 nM ET (CyaA hi, ET hi) ( n ≥3). Where indicated, cells were activated in the presence of the adenylase cyclase deficient CyaA and ET mutants (45 nM CyaA-E5, 110 nM EL1) for 2 h or 6 h, respectively ( n = 2). *** P ≤0.001; ** P ≤0.01; * P ≤0.05. Error bars, SD. 10.1371/journal.ppat.1000325.g006 Figure 6 Low CyaA or ET concentrations do not affect initiation of TCR signaling. (A) Immunoblot analysis, using phosphospecific antibodies, of CD3ζ, ZAP-70, or Erk1/2 phosphorylation in postnuclear supernatants of cells activated as above in the presence or absence of either 0.28 nM CyaA or 0.11 nM ET (CyaA lo, ET lo). Filters were stripped and re-probed with control antibodies. Representative experiments are shown ( n ≥3). The migration of molecular mass markers is indicated. (B) Quantification by laser densitometry of the relative levels of CD3ζ, ZAP-70, and Erk1/2 phosphorylation (phosphorylation in anti-CD3 stimulated cells taken as 100%, indicated as a dotted line) in PBL activated by CD3 cross-linking in the presence or absence of either 0.28 nM CyaA or 0.11 nM ET (CyaA lo, ET lo) ( n ≥3). Error bars, SD. To understand whether Th2 polarizing concentrations of ET and CyaA could selectively affect downstream components of TCR signaling specifically implicated in Th lineage committment, we focused on two molecules in the TCR signaling cascade, the Rac/Cdc42 specific guanine nucleotide exchanger Vav1 and the stress-activated kinase p38, which have been implicated in human Th2 cell differentiation [28] – [32] . Furthermore, we assessed the effect of the toxins on the serine/threonine kinase Akt1, which has been associated to Th1 cell differentiation [9] . Strikingly, analysis of Akt1 activation using phosphospecific antibodies which recognize two critical residues, T308 and S473, showed that low, Th2 polarizing concentrations of ET or CyaA were sufficient to potently impair TCR dependent Akt phosphorylation ( Figure 7A and data not shown). Conversely, both basal and TCR dependent Vav1 phosphorylation on Y174, which positively regulates Vav1 activity, was potentiated by low concentrations of ET or CyaA ( Figure 7B ). A similar enhancement was observed for p38 ( Figure 7B ), consistent with the capacity of PKA to act as an agonist of this kinase [31] , [33] . The phosphodiesterase inhibitor, IBMX, further potentiated the agonistic activity of the toxins on p38 activation (data not shown), further supporting the notion that the effects of the toxins are mediated by cAMP. Hence ET and CyaA alter the Th1/Th2 balance at least in part by antagonizing the Akt1 dependent pathway leading to Th1 cell differentiation and by potentiating the Vav1 and p38 dependent pathway(s) leading to Th2 cell differentiation. Of note, TCR dependent Vav1 and p38 activation was not impaired, but actually enhanced, when cells were pretreated with high concentrations of ET or CyaA ( Figure 7B ), despite their potent inhibitory activity on initiation of TCR signaling, suggesting that an Lck independent pathway triggered by the TCR, which can be potentiated by cAMP, may contribute to a significant extent to their activation. 10.1371/journal.ppat.1000325.g007 Figure 7 Low CyaA or ET concentrations impair TCR–dependent Akt1 phosphorylation while enhancing Vav1 and p38 phosphorylation. (A) Immunoblot analysis, using a phosphospecific antibody, of Akt1 activation in postnuclear supernatants from PBL activated for 5 min by CD3 cross-linking in the presence or absence of 45 nM CyaA or 110 nM ET (CyaA hi, ET hi) or alternatively in the presence or absence of 0.28 nM CyaA or 0.11 nM ET (CyaA lo, ET lo). A representative experiment is shown ( n = 3). (B) Immunoblot analysis, using phosphospecific antibodies, of Vav1 ( top ) or p38 ( bottom ) activation in postnuclear supernatants from PBL activated for 1 min (Vav) or 5 min (p38) by CD3 cross-linking in the presence or absence of 45 nM CyaA or 110 nM ET (CyaA hi, ET hi) or alternatively in the presence or absence of 0.28 nM CyaA or 0.11 nM ET (CyaA lo, ET lo). Filters were stripped and re-probed with control antibodies. Representative experiments are shown ( n ≥4). The migration of molecular mass markers is indicated. The graphs on the right of the immunoblots show the quantification by laser densitometry of the relative levels of Akt1, Vav1, and p38 phosphorylation (phosphorylation in anti-CD3 stimulated cells taken as 100%) in PBL activated by CD3 cross-linking in the presence or absence of 45 nM CyaA or 110 nM ET (CyaA hi, ET hi) or alternatively in the presence or absence of 0.28 nM CyaA or 0.11 nM ET (CyaA lo, ET lo). *** P ≤0.001** P ≤0.01; * P ≤0.05. Error bars, SD. Discussion The B. anthracis ET and B. pertussis CyaA adenylate cyclase toxins act as potent suppressors of T cell activation and proliferation in the 10 −9 –10 −6 molar range of concentrations [22] – [24] . In the absence of systemic intoxication, these high concentrations are likely to be reached only locally through accumulation of the toxins at the primary site of infection. However, there are anatomical districts and localized infections ( e.g. cutaneous anthrax) where low amount of toxins may be released and might modulate the host immune response. We found that both ET and CyaA are potent promoters of naive CD4 + T cell differentiation to Th2 effectors when used at subnanomolar concentrations (0.1–0.3 nM). Interestingly, distinct effects of high vs low concentrations of CyaA have also been observed in neutrophils and other phagocytes, ranging from cytolysis to apoptosis to impairment of effector functions [34] , suggesting the biological outcome of host cell exposure to the toxin is likely to be dictated by its proximity to the bacterium. The sensitivity of T cells to such low ET concentration can be accounted for by the fact that human leukocytes express the high affinity CMG2 receptor for protective antigen (PA), the receptor binding subunit of ET [35] . Furthermore, although T cells lack CD11b/CD18, the only known CyaA receptor, CyaA can effectively insert into cell membranes or artificial lipid bilayers in the absence of CD11b/CD18, albeit with a reduced efficacy [36] . The presence on T cells of a putative alternative CyaA receptor cannot however be ruled out. The immunosuppressant activity of high concentrations of ET and CyaA is fully consistent with the known inhibitory effects of cAMP on T cell activation. In physiological conditions cAMP production by a TCR-coupled adenylate cyclase is part of a negative feed-back loop which ensures extinction of TCR signaling through PKA dependent activation of Csk, a kinase that inhibits Lck by phosphorylating its C-terminal tyrosine residue [27] . This feed-back loop does not become immediately operational because cAMP production is counterbalanced by TCR dependent recruitment of PDE-4 to lipid rafts, where also the activated TCR localizes, thereby allowing the protein tyrosine kinase cascade to start [37] . Once PDE-4 dissociates from lipid rafts, the feed-back loop can terminate the signal. Since cAMP production and PKA activation are TCR-dependent, cAMP returns to basal levels after signal extinction. Alterations in this finely regulated cAMP balance by adenylate cyclase agonists, such as PGE 2 receptors, result in impaired TCR signaling and T cell activation [27] . The inhibitory activity of high concentrations of ET and CyaA on Lck activation and CD3ζ phosphorylation indicates that these toxins preventing firing of the TCR signaling cascade by altering the cAMP balance through the massive and sustained production of cAMP. In this context, it should be underlined that TCR engagement results in a rapid elevation in the levels of cAMP, which elicit a potent enhancement in PKA activation, comparable to the one observed in cells exposed to high toxin concentrations. Nevertheless, under these conditions the TCR triggers a productive signaling cascade, as opposed to cells pre-exposed to high concentrations of ET or CyaA, supporting the importance of the spatiotemporal regulation of cAMP in TCR signaling. Low toxin concentrations on the other hand, do not inhibit initiation of TCR signaling. The intracellular concentration of cAMP measured under these conditions, which is very modest compared to the burst of cAMP evoked by the TCR, may be locally and transiently neutralized by PDE-4. At variance with their inability to impair initiation of TCR signaling, low concentrations of ET or CyaA were found to selectively affect specific downstream nodes -Akt1, Vav1 and p38 activation- crucial to Th1/Th2 lineage committment. This activity is likely to result from their PKA dependent modulation of intracellular signaling mediators implicated in Th2 cell differentiation downstream of signal initiation. Akt1 has been reported to favour Th1 cell differentiation by providing the CD28 costimulatory signal required for expression of the Th1 cytokines IL-2 and IFN-γ [9] . Although our study was carried out on T cells stimulated by TCR/CD3 cross-linking in the absence of CD28 costimulation, the results show that Akt1 is effectively phosphorylated in response to TCR engagement and that this event is potently inhibited by low concentrations of ET or CyaA. The negative regulation of Akt by PKA [38] , [39] is likely to underlie this inhibitory, TCR-distal effect of the two toxins, which would moreover favour differentiation to the Th2 lineage by potentiating the PDK1/PKA pathway coupling the TCR to IL-4 gene transcription [9] . Under the same conditions, both toxins enhance TCR dependent phosphorylation of the guanine nucleotide exchanger Vav1 and activation of the stress kinase p38, which participate in Th2 lineage commitment. A skewing of the Th1/Th2 balance to Th1, as well as defects in Th2 dependent B cell responses, have been indeed observed in Vav1 −/− mice [28] , [40] . Furthermore, p38 has been implicated in human Th2 cell differentiation, at least in part through its capacity to promote activation of GATA-3 [29] – [32] . The similar enhancement of TCR-dependent Vav and p38 activation in the presence of high toxin concentrations, which block signal initiation, supports a local effect on a specific signaling module independent of the TCR proximal, Lck-dependent signaling cascade. While there is evidence for an agonistic role of cAMP in p38 activation [31] , the potential function of cAMP in the modulation of Vav1 activity has not been directly addressed. We have previously characterized a Fyn dependent, Lck independent pathway linking the TCR to Vav1 phosphorylation and p38 activation which could be potentiated by PGE 2 [41] . Together with the evidence implicating Vav1 and p38 in Th2 cell differentiation, the Th1 bias observed in Fyn −/− mice [42] , [43] may suggest a potential involvement of this cAMP sensitive pathway in the Th2 promoting activity of ET and CyaA. It is noteworthy that the effects of ET and CyaA are almost undistinguishable, notwithstanding their differences in stucture, mode of cell entry and intracellular localization. ET is an A–B type toxin, consisting of a cell binding component, PA, which targets cells via the receptors TEM8 or CMG2, and a toxin component, EF. ET enters the cell by receptor mediated internalization and is transported to the endosomes, wherefrom it is released into the cytosol [11] . CyaA is a single polypeptide which binds to target cells both directly and through a membrane receptor, which in macrophages and other APC is the integrin CD11b/CD18 [44] . Following binding, the adenylate cyclase N-terminal domain is translocated accross the plasma membrane of target cells [12] . Hence EF and CyaA produce cAMP not only with different kinetics, which is delayed for ET probably due to the multistep mechanism of delivery into host cells, but also at different subcellular locations, in the cytosol with a prevalent perinuclear localization, and close to the plasma membrane, respectively [45] – [47] . The role of AKAPs in segregation of PKA pools at specific subcellular localizations and dynamic recruitment of phosphodiesterases underscores the importance of the spatiotemporal control of cAMP signaling [48] . By focalizing cAMP production at distinct sites within the cell, EF and CyaA could differentially affect early and late events in TCR signaling. Our findings indicate that the critical targets of the cAMP dependent Th2 polarizing activities of ET and CyaA can be activated independently of the subcellular site of cAMP production. It is likely that the sustained cAMP production overrides the negative local feedback mechanisms, resulting in loss of compartmentalization of cAMP dependent signaling. It should be however underlined that low ET concentrations are almost as effective as high concentrations in triggering PKA activation, suggesting that ET may modulate other functions, such as CREB mediated gene expression [45] , [47] , through activation of specific PKA pools. Both ET and CyaA have been reported to potentiate antibody responses and development of antigen specific Th2 cells in mice when coadministered with antigen [13] – [17] . This adjuvant activity had been related to their capacity to modulate cytokine production by dendritic cells and macrophages in vitro , resulting in reduction in IL-12 and enhancement in IL-10 and IL-4 expression [13] – [16] , [18] , [19] . Our finding that ET and CyaA directly affect the Th1/Th2 balance by favouring Th2 cell development suggests that their adjuvanticity is also due to their effects on CD4 + T cells. This activity appears mediated by cAMP, as it cannot be reproduced by the respective enzymatically deficient mutants. The Th2 polarizing effects of these toxins resulting from their modulation of cytokine expression by APC have also been related to their cAMP elevating activity [15] , [18] , [19] , [49] . Consistent with these findings, both non-hydrolysable cAMP analogues and PGE 2 have been reported to favour Th2 cell development both by directly affecting CD4 + T cell differentiation [4] , [6] , [8] and by shaping the pattern of cytokine production by APC [19] – [21] . The Th2 driving activity of ET may be very relevant to cutaneous anthrax, where there is a limited toxin production and where resolution of infection has been causally linked to the development of an antibody response against the toxin [50] , [51] . As opposed to the clear-cut role of cAMP in both the direct and the APC dependent Th2 driving activity of ET or CyaA in vitro , the role of cAMP in the adjuvanticity of the toxins in vivo is more controvertial. While the adjuvant activity of adenylate cyclase deficient ET mutants has as yet not been tested in vivo , there are discrepancies as to the adjuvanticity of catalytically inactive CyaA mutants, which have been proposed to result from a number of factors, including the genetic background of the mouse strain, the route of antigen delivery, the dose of CyaA mutant and the vaccination schedule [14] , [15] , [52] , [53] . Genetically detoxified mutants of other cAMP elevating toxins such as CT or LT-I, or their individual B (binding) subunits, have been demonstrated to retain adjuvant activity [10] , indicating that both cAMP production and toxin binding to specific receptors contribute to their adjuvanticity. This possibility has been ruled out for ET, as PA does not harbour any activity either on APC or T cells in vitro nor acts as an adjuvant in vivo [16] . On the other hand, at variance with other reports [15] , [53] , an enzymatically deficient CyaA mutant, highly purified to rule out a contamination by LPS, has been reported to display adjuvant properties comparable or even superior to wild-type CyaA [14] , [52] . A potential implication of CD11b/CD18 in the adjuvanticity of CyaA appears unlikely, as this integrin suppresses cytokine production by dendritic cells [54] , [55] , and moreover does not account either for the immunodeviating activity of CyaA on APC in vitro , which is cAMP dependent [15] , [18] , [19] , or for the conflicting results obtained by different groups in vivo [14] , [15] , [52] , [53] . An integrated and detailed analysis of the structural and functional interaction of adenylate cyclase toxins with the different cellular components which together orchestrate the immune response is expected not only to clarify their mechanism of adjuvanticity but also to lead to the development of more specific and effective adjuvants. Materials and Methods Cells, antibodies, reagents, and toxins Peripheral blood mononuclear cells were purified from buffy coats from anonymous healthy donors (collectively ∼30, available from authorised blood banks) by density gradient centrifugation on Ficoll-Paque (Amersham Biosciences, Buckinghamshire, UK), using a Beckman GS-6R tabletop centrifuge (Beckman Coulter SpA, Milan, Italy). Cells were washed 2× in phosphate buffered saline (PBS), resuspended in RPMI 1640 (Invitrogen Ltd, Paisley, UK) (buffered with sodium bicarbonate to pH 7.2) supplemented with 7.5% fetal calf serum (FCS) (Hyclone, Thermofischer Scientific Inc, SouthLogan, UT), plated in plastic flasks (Sarstedt AG, Numbrecht, Germany) and incubated overnight at 37°C in a humidified atmosphere with 5% CO 2 . Non-adherent cells, which consisted principally of peripheral blood lymphocytes (PBL) and of which >90% were T cells (CD3 + ), were centrifuged at 800× g for 5 min at room temperature in Beckman GS-6R tabletop centrifuge and resuspended in fresh RPMI 1640 supplemented with 7.5% FCS. For cAMP measurement, T cells were purified from peripheral blood mononuclear cell suspensions using the StemSep Human T cell enrichment kit (Voden Medical Instruments SpA, Milan, Italy). Phosphospecific antibodies recognizing the phosphorylated active forms of CD3ζ, ZAP-70 (Y493), Vav1 (Y160), Akt1 (T308/S473), Erk1/2 (T202/Y204), p38 (T180/Y182) and CREB (S133), as well as an antibody recognizing phosphorylated Y505 on Lck, were from Cell Signaling Technology (Beverly, MA), Santa Cruz Biotechnology (Santa Cruz, CA) and Biosource Europe SA (Nivelles, Belgium). An antibody against the phosphorylated PKA consensus phosphorylation site, R-X-X-pT-X-X/R-R-X-pS-X-X, was purchased from Cell Signaling Technology. Anti-CD3ζ, -Lck, ZAP-70, -Vav, -Erk2, -p38, -CREB and anti-actin antibodies were from Santa Cruz Biotechnology, Upstate Biotechnology (Dundee, UK) and Cell Signaling Technology. A mAb suitable for immunoprecipitation of tyrosine phosphorylated CD3ζ was kindly provided by M. Banyiash. Fluorochrome-labeled anti-CD3 mAb were obtained from Becton Dickinson Biosciences (Milan, Italy). Unlabeled secondary antibodies were purchased from Cappel (ICN Pharmaceuticals Inc, CA) and peroxidase labeled antibodies from Amersham Biosciences. IgG antibodies from OKT3 (anti-CD3; American Type Culture Collection, Manassas, VA) hybridoma supernatants were purified on Mabtrap (Amersham Biosciences, Inc) and titrated by flow cytometry. CyaA and the enzymatically inactive variant CyaA-E5 (resulting from a Leu-Gln dipeptide insertion between D188 and I189 in the catalytic core of the enzyme) were expressed in E. coli and purified to near homogeneity by previously established procedures modified as described [56] in order to eliminate most of the contaminating endotoxin. The specific activity of CyaA, measured as described in Ladant et al. [26] was higher than 500 µmol cAMP/min.mg whereas CyaA-E5 had no detectable enzymatic activity. In both preparations the endotoxin content, determined using a LAL assay (QCL-1000 kit from Lonza), was below 0.5 EU/µg protein. PA, LF and EL1 were expressed in E. coli and purified as described [25] , [45] , [57] . H89, KT5720, IBMX and 8-CPT were purchased from Sigma-Aldrich (Milan, Italy) and Calbiochem (Merck Biosciences GmbH, Schwalboch, Germany). Cell activations and lysis, immunoblots For immunoblot analyses cells were plated at 5×10 6 cells/ml in plastic flasks in RPMI 1640 supplemented with 7.5% FCS and 2 mM CaCl 2 (required for CyaA entry into the cells), added with CyaA/CyaA-E5, and incubated at 37°C in a humidified atmosphere with 5% CO 2 for 2 h before activation. Alternatively, cells were plated as above, added with ET (ratio PA∶EF 1.6), and incubated at 37°C for 6 h before activation. Activations by TCR/CD3 cross-linking were performed by incubating PBL with saturating concentrations of anti-CD3 mAb (as assessed by flow cytometry) and 50 µg ml −1 secondary antibodies (goat anti-mouse immunoglobulin Ig) in RPMI 1640 for 1–5 min at 37°C as previously described [58] . None of the above mentioned treatments affected cell viability, as assessed by Trypan blue exclusion (data not shown). When required, cells were pretreated with the PKA inhibitors, H89 (20 µM) and KT5720 (56 nM), or with the PDE inhibitor, IBMX (0.5 mM), for 1 h before addition of ET or CyaA. Alternatively, cells were treated for 30 min with 8-CPT (100 µM). Cells were recovered by centrifugation at 16,000× g for 30 sec at 4°C in an Eppendorf 5415R microcentrifuge (Eppendorf srl, Milan, Italy), washed 2× in PBS and lysed in 1% (v/v) Triton X-100 in 20 mM Tris-HCl pH 8, 150 mM NaCl (in the presence of 0.2 mg/ml Na orthovanadate, 1 µg/ml pepstatin, leupeptin, and aprotinin, and 10 mM phenyl methyl sulfonyl fluoride). To normalize for variations in protein content among samples, equal amounts of proteins from each sample (measured using a kit from Pierce, Rockford, IL) were resolved by 12% SDS-PAGE and transferred to 0.45-µm nitrocellulose filters Whatman GmbH, Dassel, Germany). Prestained molecular mass markers (Invitrogen) were included in each gel. Immunoblots were carried out using primary antibodies and peroxidase-labeled secondary antibodies according to the manufacturers' instructions and a chemiluminescence detection kit (Pierce). Blots were scanned using a laser densitometer (Duoscan T2500 Agfa, Milan, Italy) and quantified using the ImageQuant 5.0 software (Molecular Dynamics, Sunnyvale, CA). Data were normalized to loading controls. T cell proliferation assays, cytokine, and cAMP measurements For proliferation assays, cells (2×10 5 /sample) were plated in 96-well plates in RPMI 1640 supplemented with 7.5% FCS (and 2 mM CaCl 2 for CyaA treatments and respective controls), added with CyaA/CyaA-E5 (0.07–45 nM) or ET/EL1 (0.01 pM–1.1 nM), and incubated at 37°C in a humidified atmosphere with 5% CO 2 for 2 h (CyaA) or 6 h (ET) before activation. Cells were activated by CD3 cross-linking on secondary antibody-coated plates as described [58] and processed 16–48 h after activation. [ 3 H]-thymidine (1 mCi) was added to each microtiter well (96-well plates) for the last 18 h of culture. After harvesting the cells with an automatic harvester (Micromate 196, Canberra Packard, Meriden, CT), proliferation was determined by measuring the [ 3 H]thymidine (Amersham, Buckinghamshire, UK) incorporation in a β-counter (Matrix 9600, Canberra Packard, Meriden, CT). To measure Th cell differentiation, enriched human CD4 + T cells were activated by immobilized anti-CD3 mAb (purified from OKT3 hybridoma supernatants on Mabtrap, Amersham Biosciences Europe) as described [59] , in the absence or presence of recombinant hIL-12 (2 ng/ml, Sigma-Aldrich Milan, Italy) to promote Th1 differentiation, or recombinant hIL-4 (10 ng/ml, Sigma-Aldrich Milan, Italy) to promote Th2 differentiation. Recombinant hIL-2 (kindly provided by Eurocetus Milan, Italy) was added to the cultures on day 4 and 7. After 10 days, cells (1×10 6 ) were washed, stimulated for 24 h or 48 h using anti-CD3 mAb and the levels of IL-4, IL-13, IFNγ and TNF-α were measured by ELISPOT as described [59] . Intracellular cAMP was quantitated by enzyme-linked immunoassay kit (Biotrak EIA, Amersham Biosciences) according to the manufacturers' instructions. For these experiments, cells (1×10 6 plated in 96-well plates in 200 µl RPMI 1640/7.5% FCS) were treated with CyaA/CyaA-E5 or ET/EL1 as described above for 30 min to 24 h in a humidified atmosphere with 5% CO 2 . At the end of the treatment, cells were washed 2× in PBS and lysed in the lysis reagent included in the kit. RNA purification and real-time quantitative RT–PCR Total RNA was extracted from Th cells, polarized as described above, using Tri Reagent (Ambion, Austin, TX). Reverse transcription-polymerase chain reaction (RT-PCR) was carried out on 400 ng total RNA using ImProm-II™ reverse transcriptase and and oligo-dT (Promega Italia srl, Milan, Italy) as first strand primer. Real-time quantitative PCR was performed using SYBR Green I SensiMix™ dT Kit (Quantace, Watford, UK) according to the manufacturer's instructions, in an Opticon 2 Continuous Fluorescence Detection System (MJ Research, Bio-Rad Laboratories, Waltham, MA). All samples were run in duplicate on 96-well optical PCR plates (Roche Diagnostics, Milan, Italy). The specific primers used to amplify cDNA fragments corresponding to c-maf, GATA-3, T-bet and GAPDH were: 5′-TGGAGTCGGAGAAGAACCAG-3′ ( sense ), 5′-GCTTCCAAAATGTGGCGTAT-3′ ( antisense ) for c-Maf; 5′-GAAGGAAGGCATCCAGACCAG-3′ ( sense ), 5′-ACCCATGGCGGTGACCATGC-3′ ( antisense ) for GATA-3; 5′-TAATAACCCCTTTGCCAAAGG-3′ ( sense ); and 5′-TCCCCCAAGGAATTGACAGT-3′ ( anti-sense ) for T-bet and 5′-TGCACCACCAACTGCTTAGC-3′ ( sense ) and 5′-GGCATGGACTGTGGTCATGAG-3′ ( anti-sense ) for GAPDH. After an initial denaturation for 10 min at 95°C, denaturation at the subsequent 40 cycles was performed for 15 s at 95°C, followed by 15 s primer annealing at 60°C and a final extension at 72°C for 30 s. The ΔΔC T method [60] was applied as a comparative quantification method. The specificity of the amplified fragment was demonstrated by the melting curve, where a single peak was observed for each sample amplified with c-maf, GATA-3, T-bet and GAPDH primers. c-maf, GATA-3 and T-bet mRNA levels were normalized to GAPDH, used as a housekeeping gene. Statistical analyses Mean values, standard deviation values and Student's t test (unpaired) were calculated using the Microsoft Excel application. A level of P 90% were T cells (CD3 + ), were centrifuged at 800× g for 5 min at room temperature in Beckman GS-6R tabletop centrifuge and resuspended in fresh RPMI 1640 supplemented with 7.5% FCS. For cAMP measurement, T cells were purified from peripheral blood mononuclear cell suspensions using the StemSep Human T cell enrichment kit (Voden Medical Instruments SpA, Milan, Italy). Phosphospecific antibodies recognizing the phosphorylated active forms of CD3ζ, ZAP-70 (Y493), Vav1 (Y160), Akt1 (T308/S473), Erk1/2 (T202/Y204), p38 (T180/Y182) and CREB (S133), as well as an antibody recognizing phosphorylated Y505 on Lck, were from Cell Signaling Technology (Beverly, MA), Santa Cruz Biotechnology (Santa Cruz, CA) and Biosource Europe SA (Nivelles, Belgium). An antibody against the phosphorylated PKA consensus phosphorylation site, R-X-X-pT-X-X/R-R-X-pS-X-X, was purchased from Cell Signaling Technology. Anti-CD3ζ, -Lck, ZAP-70, -Vav, -Erk2, -p38, -CREB and anti-actin antibodies were from Santa Cruz Biotechnology, Upstate Biotechnology (Dundee, UK) and Cell Signaling Technology. A mAb suitable for immunoprecipitation of tyrosine phosphorylated CD3ζ was kindly provided by M. Banyiash. Fluorochrome-labeled anti-CD3 mAb were obtained from Becton Dickinson Biosciences (Milan, Italy). Unlabeled secondary antibodies were purchased from Cappel (ICN Pharmaceuticals Inc, CA) and peroxidase labeled antibodies from Amersham Biosciences. IgG antibodies from OKT3 (anti-CD3; American Type Culture Collection, Manassas, VA) hybridoma supernatants were purified on Mabtrap (Amersham Biosciences, Inc) and titrated by flow cytometry. CyaA and the enzymatically inactive variant CyaA-E5 (resulting from a Leu-Gln dipeptide insertion between D188 and I189 in the catalytic core of the enzyme) were expressed in E. coli and purified to near homogeneity by previously established procedures modified as described [56] in order to eliminate most of the contaminating endotoxin. The specific activity of CyaA, measured as described in Ladant et al. [26] was higher than 500 µmol cAMP/min.mg whereas CyaA-E5 had no detectable enzymatic activity. In both preparations the endotoxin content, determined using a LAL assay (QCL-1000 kit from Lonza), was below 0.5 EU/µg protein. PA, LF and EL1 were expressed in E. coli and purified as described [25] , [45] , [57] . H89, KT5720, IBMX and 8-CPT were purchased from Sigma-Aldrich (Milan, Italy) and Calbiochem (Merck Biosciences GmbH, Schwalboch, Germany). Cell activations and lysis, immunoblots For immunoblot analyses cells were plated at 5×10 6 cells/ml in plastic flasks in RPMI 1640 supplemented with 7.5% FCS and 2 mM CaCl 2 (required for CyaA entry into the cells), added with CyaA/CyaA-E5, and incubated at 37°C in a humidified atmosphere with 5% CO 2 for 2 h before activation. Alternatively, cells were plated as above, added with ET (ratio PA∶EF 1.6), and incubated at 37°C for 6 h before activation. Activations by TCR/CD3 cross-linking were performed by incubating PBL with saturating concentrations of anti-CD3 mAb (as assessed by flow cytometry) and 50 µg ml −1 secondary antibodies (goat anti-mouse immunoglobulin Ig) in RPMI 1640 for 1–5 min at 37°C as previously described [58] . None of the above mentioned treatments affected cell viability, as assessed by Trypan blue exclusion (data not shown). When required, cells were pretreated with the PKA inhibitors, H89 (20 µM) and KT5720 (56 nM), or with the PDE inhibitor, IBMX (0.5 mM), for 1 h before addition of ET or CyaA. Alternatively, cells were treated for 30 min with 8-CPT (100 µM). Cells were recovered by centrifugation at 16,000× g for 30 sec at 4°C in an Eppendorf 5415R microcentrifuge (Eppendorf srl, Milan, Italy), washed 2× in PBS and lysed in 1% (v/v) Triton X-100 in 20 mM Tris-HCl pH 8, 150 mM NaCl (in the presence of 0.2 mg/ml Na orthovanadate, 1 µg/ml pepstatin, leupeptin, and aprotinin, and 10 mM phenyl methyl sulfonyl fluoride). To normalize for variations in protein content among samples, equal amounts of proteins from each sample (measured using a kit from Pierce, Rockford, IL) were resolved by 12% SDS-PAGE and transferred to 0.45-µm nitrocellulose filters Whatman GmbH, Dassel, Germany). Prestained molecular mass markers (Invitrogen) were included in each gel. Immunoblots were carried out using primary antibodies and peroxidase-labeled secondary antibodies according to the manufacturers' instructions and a chemiluminescence detection kit (Pierce). Blots were scanned using a laser densitometer (Duoscan T2500 Agfa, Milan, Italy) and quantified using the ImageQuant 5.0 software (Molecular Dynamics, Sunnyvale, CA). Data were normalized to loading controls. T cell proliferation assays, cytokine, and cAMP measurements For proliferation assays, cells (2×10 5 /sample) were plated in 96-well plates in RPMI 1640 supplemented with 7.5% FCS (and 2 mM CaCl 2 for CyaA treatments and respective controls), added with CyaA/CyaA-E5 (0.07–45 nM) or ET/EL1 (0.01 pM–1.1 nM), and incubated at 37°C in a humidified atmosphere with 5% CO 2 for 2 h (CyaA) or 6 h (ET) before activation. Cells were activated by CD3 cross-linking on secondary antibody-coated plates as described [58] and processed 16–48 h after activation. [ 3 H]-thymidine (1 mCi) was added to each microtiter well (96-well plates) for the last 18 h of culture. After harvesting the cells with an automatic harvester (Micromate 196, Canberra Packard, Meriden, CT), proliferation was determined by measuring the [ 3 H]thymidine (Amersham, Buckinghamshire, UK) incorporation in a β-counter (Matrix 9600, Canberra Packard, Meriden, CT). To measure Th cell differentiation, enriched human CD4 + T cells were activated by immobilized anti-CD3 mAb (purified from OKT3 hybridoma supernatants on Mabtrap, Amersham Biosciences Europe) as described [59] , in the absence or presence of recombinant hIL-12 (2 ng/ml, Sigma-Aldrich Milan, Italy) to promote Th1 differentiation, or recombinant hIL-4 (10 ng/ml, Sigma-Aldrich Milan, Italy) to promote Th2 differentiation. Recombinant hIL-2 (kindly provided by Eurocetus Milan, Italy) was added to the cultures on day 4 and 7. After 10 days, cells (1×10 6 ) were washed, stimulated for 24 h or 48 h using anti-CD3 mAb and the levels of IL-4, IL-13, IFNγ and TNF-α were measured by ELISPOT as described [59] . Intracellular cAMP was quantitated by enzyme-linked immunoassay kit (Biotrak EIA, Amersham Biosciences) according to the manufacturers' instructions. For these experiments, cells (1×10 6 plated in 96-well plates in 200 µl RPMI 1640/7.5% FCS) were treated with CyaA/CyaA-E5 or ET/EL1 as described above for 30 min to 24 h in a humidified atmosphere with 5% CO 2 . At the end of the treatment, cells were washed 2× in PBS and lysed in the lysis reagent included in the kit. RNA purification and real-time quantitative RT–PCR Total RNA was extracted from Th cells, polarized as described above, using Tri Reagent (Ambion, Austin, TX). Reverse transcription-polymerase chain reaction (RT-PCR) was carried out on 400 ng total RNA using ImProm-II™ reverse transcriptase and and oligo-dT (Promega Italia srl, Milan, Italy) as first strand primer. Real-time quantitative PCR was performed using SYBR Green I SensiMix™ dT Kit (Quantace, Watford, UK) according to the manufacturer's instructions, in an Opticon 2 Continuous Fluorescence Detection System (MJ Research, Bio-Rad Laboratories, Waltham, MA). All samples were run in duplicate on 96-well optical PCR plates (Roche Diagnostics, Milan, Italy). The specific primers used to amplify cDNA fragments corresponding to c-maf, GATA-3, T-bet and GAPDH were: 5′-TGGAGTCGGAGAAGAACCAG-3′ ( sense ), 5′-GCTTCCAAAATGTGGCGTAT-3′ ( antisense ) for c-Maf; 5′-GAAGGAAGGCATCCAGACCAG-3′ ( sense ), 5′-ACCCATGGCGGTGACCATGC-3′ ( antisense ) for GATA-3; 5′-TAATAACCCCTTTGCCAAAGG-3′ ( sense ); and 5′-TCCCCCAAGGAATTGACAGT-3′ ( anti-sense ) for T-bet and 5′-TGCACCACCAACTGCTTAGC-3′ ( sense ) and 5′-GGCATGGACTGTGGTCATGAG-3′ ( anti-sense ) for GAPDH. After an initial denaturation for 10 min at 95°C, denaturation at the subsequent 40 cycles was performed for 15 s at 95°C, followed by 15 s primer annealing at 60°C and a final extension at 72°C for 30 s. The ΔΔC T method [60] was applied as a comparative quantification method. The specificity of the amplified fragment was demonstrated by the melting curve, where a single peak was observed for each sample amplified with c-maf, GATA-3, T-bet and GAPDH primers. c-maf, GATA-3 and T-bet mRNA levels were normalized to GAPDH, used as a housekeeping gene. Statistical analyses Mean values, standard deviation values and Student's t test (unpaired) were calculated using the Microsoft Excel application. A level of P <0.05 was considered statistically significant.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8621548/

Cross-Strain Neutralizing and Protective Monoclonal Antibodies against EEEV or WEEV

The three encephalitic alphaviruses, namely, the Venezuelan, eastern, and western equine encephalitis viruses (VEEV, EEEV, and WEEV), are classified by the Centers for Disease Control and Prevention (CDC) as biothreat agents. Currently, no licensed medical countermeasures (MCMs) against these viruses are available for humans. Neutralizing antibodies (NAbs) are fast-acting and highly effective MCMs for use in both pre- and post-exposure settings against biothreat agents. While significant work has been done to identify anti-VEEV NAbs, less has been done to identify NAbs against EEEV and WEEV. In order to develop anti-EEEV or -WEEV NAbs, mice were immunized using complementary strategies with a variety of different EEEV or WEEV immunogens to maximize the generation of NAbs to each of these viruses. Of the hybridomas generated, three anti-EEEV and seven anti-WEEV monoclonal antibodies were identified with in vitro neutralization activity. The most potent neutralizers (two anti-EEEV NAbs and three anti-WEEV NAbs) were further evaluated for neutralization activity against additional strains of EEEV, a single strain of Madariaga virus (formerly South American EEEV), or WEEV. Of these, G1-2-H4 and G1-4-C3 neutralized all three EEEV strains and the Madariaga virus strain, whereas G8-2-H9 and 12 WA neutralized six out of eight WEEV strains. To determine the protective efficacy of these NAbs, the five most potent neutralizers were evaluated in respective mouse aerosol challenge models. All five NAbs demonstrated various levels of protection when administered at doses of 2.5 mg/kg or 10 mg/kg 24 h before the respective virus exposure via the aerosol route. Of these, anti-EEEV NAb G1-4-C3 and anti-WEEV NAb 8C2 provided 100% protection at both doses and all surviving mice were free of clinical signs throughout the study. Additionally, no virus was detected in the brain 14 days post virus exposure. Taken together, efficacious NAbs were developed that demonstrate the potential for the development of cross-strain antibody-based MCMs against EEEV and WEEV infections. 1. Introduction The three encephalitic alphaviruses, namely, the Venezuelan, eastern, and western equine encephalitis viruses (VEEV, EEEV, and WEEV), belong to the Alphavirus genus of the family Togaviridae [ 1 ]. All three encephalitic alphaviruses share a number of sequence, structural, and functional similarities, including a positive-sense and single-stranded RNA genome with two polyprotein gene clusters, namely, nonstructural and structural [ 2 ]. The nonstructural proteins are translated directly from the 5′ two-thirds of the genomic RNA. A subgenomic positive-stranded RNA (the 26S RNA) is identical to the 3′ one-third of the genome and serves as the translational template for the structural proteins, capsid (C), E3, E2, 6K, and E1 [ 3 ]. Three of these proteins, C, E1, and E2 are found on all mature encephalitic alphavirus virions, while the E3 protein has only been positively identified in VEEV capsids to date [ 4 ]. The C encapsidates the viral genome and lies beneath the viral lipid bilayer. E1 and E2 project from the virus envelope as trimeric spikes of E1/E2 heterodimers. The E1 protein is responsible for membrane fusion, while E2, the receptor-binding protein, is believed to be the major protective antigen [ 5 , 6 ]. EEEV formerly encompassed North American and South American strains, with the North American strains being much more virulent than their South American counterparts. The South American strains are now classified as a separate species, namely, the Madariaga virus, which has an approximately 23% difference in nucleotide sequence from EEEV [ 7 ]. In nature, these viruses primarily circulate through animal populations and infect humans via bites from mosquito carriers that have fed on infected animals. Human infection typically results in an acute and highly incapacitating disease that is characterized by severe symptoms that are similar to influenza. However, severe or fatal encephalitis can result from these viruses crossing the cerebral vascular endothelium or the olfactory epithelium [ 8 ]. In 2019, the northern USA provinces experienced the worst outbreak of EEEV since monitoring of the disease began 15 years ago and a total of 38 cases, including 12 deaths, were reported [ 9 ]. Accidental laboratory infections [ 1 ] and experimental studies in animals [ 10 ] with these three alphaviruses have demonstrated that they are highly infectious via the aerosol route. Furthermore, alphavirus infections via the aerosol route develop much faster, displaying higher morbidity compared with the natural (mosquito bite) route, likely because the aerosol route allows for more of the virus to contact olfactory neurons, thus expediting viral invasion of the brain [ 11 ]. In addition, high titers of these alphaviruses are easily obtained in cell culture and are relatively stable (either liquid or dry) in the environment. As such, VEEV, EEEV, and WEEV are classified by the Centers for Disease Control and Prevention (CDC) as biothreat agents [ 12 ]. Indeed, VEEV was weaponized and aerosolized as an incapacitating agent by the pre-1992 Soviet Union and pre-1969 United States biological warfare programs [ 13 ]. Currently, no licensed vaccines or therapeutics are available for any of these three encephalitic alphaviruses for the protection or treatment of humans [ 14 ]. Antibodies, which are naturally produced in the body as part of the immune response to infectious agents, can also be used in the form of polyclonal serum/plasma preparations or recombinantly manufactured neutralizing antibodies (NAbs) to passively immunize patients to prevent or treat infectious diseases [ 15 , 16 , 17 , 18 ]. NAbs can confer immediate and consistent protection against infectious agents when administered, regardless of the recipient's immune status. The use of NAbs in an infectious disease emergency has been demonstrated in the 2010 and 2012 outbreaks of Hendra virus in Australia [ 19 ], and in the unprecedented 2013–2016 outbreak of the Ebola virus in West Africa [ 20 ], highlighting that NAbs can play a critical role in the management of a deadly infectious disease crisis. NAbs that are licensed by regulatory agencies are available for prophylaxis of respiratory syncytial virus and therapeutically to treat anthrax [ 21 , 22 ], demonstrating their utility as effective medical countermeasures (MCMs) against infectious agents [ 23 , 24 ]. While considerable progress has been made in the development of potential therapeutic anti-VEEV NAbs [ 25 , 26 , 27 ], the development of anti-EEEV and anti-WEEV NAbs has not been significantly explored. To date, there have been only two reports of anti-WEEV NAbs [ 28 , 29 ], and two reports of anti-EEEV NAbs [ 30 , 31 ]. To address this, we set out to generate and identify NAbs against these viruses. Two different approaches were used to generate anti-EEEV and anti-WEEV neutralizing hybridoma clones, yielding a total of three anti-EEEV and seven anti-WEEV NAbs. Among them, two anti-EEEV NAbs, namely, G1-2-H4 and G1-4-C3, and two anti-WEEV NAbs, namely, G8-2-H9 and 12 WA, were shown to possess cross-strain neutralizing activity against multiple strains of EEEV and WEEV, respectively. In addition, in an in vivo pre-exposure prophylaxis setting, all five NAbs that were tested demonstrated a level of protection against EEEV or WEEV aerosol infection. 2. Materials and Methods 2.1. Reagents, Cells, and Viruses High-glucose Dulbecco's minimal essential medium (DMEM), Leibovitz (L-15) medium, phosphate-buffered saline (PBS), non-fat dry milk, fetal bovine serum (FBS), Synth-a-Freeze defined protein-free cryopreservation medium, Tween-20, SDS-polyacrylaminde gels, StartingBlock (PBS) blocking buffer, polyvinylidene difluoride (PVDF) membrane, agarose, penicillin, streptomycin, Melon Gel purification kits, Nunc cryovials, and Invitrogen cloning kits were purchased from ThermoFisher Scientific (Loughborough, UK; Waltham, MA, USA; Ottawa, ON, Canada) or Sigma Aldrich (Poole, UK). TiterMax Gold adjuvant was purchased from Cedarlane (Burlington, ON, Canada). Clonacell-HY Kit was from Stem Cell Technologies (Vancouver, BC, Canada). The AdEasy system was purchased from Qbiogene (Carlsbad, CA, USA). Restriction enzymes were from New England Biolabs (Mississauga, ON, Canada). Cell culture flasks and plates were from VWR (Mississauga, ON, Canada). Goat anti-mouse IgG horseradish peroxidase (HRP)-conjugated antibody and 2,2′-Azino-di(3-ethylbenzthiazoline-6-sulfonate) (ABTS) substrate were from KPL (Gaithersburg, MD, USA). Isoflurane was obtained from Baxter (Mississauga, ON, Canada). Alkaline phosphate substrate and goat anti-mouse alkaline phosphate conjugated antibody were from Bio-Rad Laboratories (Hercules, CA, USA). A non-specific murine IgG monoclonal antibody (MAb) and dimethyl sulfoxide (DMSO) were obtained from Sigma Aldrich (Poole, UK). Vero (ATCC CCL-81) cells, Sp 2/0 mouse myeloma cells, and HEK 293 cells were from the American Type Culture Collection (ATCC, Manassas, VA, USA), and the European Collection of Animal Cell Cultures (Salisbury, UK). EEEV strains PE6, FL93-969, and Williams; Madariaga virus (435731); and WEEV strains B11 and CBA-87 were kindly provided by Dr. George Ludwig, United States Army Medical Research Institute for Infectious Diseases (USAMRIID), Frederick, MD, USA. WEEV strain 71V-1658 was kindly provided by Dr. Nick Karabatsos, CDC, Fort Collins, CO, USA; WEEV strains Fleming and California were purchased from ATCC; WEEV strains McMillan, Mn548, and Mn520 were kindly provided by Drs. Mike Drebot and Harvey Artsorb, National Microbiology Laboratory, Winnipeg, MN, Canada. 2.2. Cell and Virus Maintenance Hybridoma cell lines were maintained in DMEM supplemented with 10% FBS. HEK 293 cells were grown in DMEM supplemented with 5% FBS. Vero cells were grown in DMEM supplemented with 5% FBS, or DMEM supplemented with 10% FBS, 2 mM l -glutamine, 50 IU/mL penicillin, and 50 µg/mL streptomycin. All cell cultures were at 37 °C in a 5% CO 2 humidified atmosphere, except for the virus culture in L15 medium, which had no added CO 2 . For cryopreservation, cell lines were resuspended in Synth-a-Freeze defined protein-free cryopreservation medium, or FBS supplemented with 20% ( v/v ) DMSO, and 1 mL aliquots were transferred to cryovials and stored in liquid nitrogen. Experiments with live viruses were carried out at two establishments: the Defence Science and Technology Laboratory (Dstl), UK, under UK Advisory Committee on Dangerous Pathogens, Containment Level 3 (CL3) conditions, and the Defence Research and Development Canada, Suffield Research Centre (DRDC SRC) CL3 facilities in compliance with Public Health Agency of Canada and Canadian Food Inspection Agency Guidelines. Virulent virus stocks were prepared at Dstl by inoculating suckling mouse pups intra-cranially with the virus and allowing them to become moribund (24 h after inoculation) before culling with an overdose of sodium pentobarbital. All animal studies were carried out in accordance with the UK Animals (Scientific Procedures) Act of 1986 and the Codes of Practice for the Housing and Care of Animals used in Scientific Procedures 1989. The work was performed under Project licence 30/3166 and was approved by the UK Home Office on 16 April 2014. The virus was harvested by extracting tissue through the dorsal cranium with a large-bore syringe needle and mixed with L-15 medium supplemented with 2% ( v/v ) FBS, 2 mM l -glutamine, 50 IU/mL penicillin, and 50 µg/mL streptomycin (virus culture medium). This was then passed through a 70 µm nylon cell strainer, clarified at 10,000 rpm for 10 min in an SW28 rotor (Beckman Coulter, UK), and stored at −80 °C. Standard plaque assays were performed in a 24-well plate format with 100 µL/well of virus inoculum applied in duplicate to Vero cell monolayers under a carboxymethyl cellulose overlay. Plates were incubated in humidified conditions at 37 °C for 3–4 days prior to fixation with 10% ( v/v ) formal saline solution and visualization of plaques was achieved using 1% crystal violet solution. The limit of detection in this assay is 10 plaque-forming units (pfu)/mL of the original sample. The use of this production method minimizes the potential for loss of virulence factors as a result of cell culture adaptation, and also provides a more representative wild-type virus population (quasispecies). Virus seed stocks were made at DRDC SRC via the inoculation of Vero cells with virus suspensions at a multiplicity of infection of less than 0.1. The supernates were clarified using centrifugation and stored in aliquots at −70 °C [ 32 ]. 2.3. Immunogens An encephalomyelitis vaccine for horses, consisting of formalin-inactivated EEEV and WEEV, and tetanus toxoid (Zoetis vaccine) was purchased from Zoetis Canada INC (Kirkland, QC, Canada). An adenovirus DNA vaccine, namely, pAd-EEEV PE6, was constructed by cloning the structural proteins of EEEV strain PE6 into an adenovirus vector using the AdEasy system according to the manufacturer's protocol. The recombinant adenoviral construct pAd-EEEV PE6 was linearized with Pac I and transfected into HEK 293 cells cultured in DMEM with 5% FBS for amplification and then the amplified adenovirus was purified via chromatography. A plasmid DNA vaccine, namely, pVHX-6-WEEV 71V-1658, was described previously [ 33 ] and expresses the structural proteins (capsid, E3, E2, 6K, and E1) of WEEV 71V-1658. Recombinant E2 (rE2) of EEEV V105-00210 was prepared by cloning the E2 gene into a mammalian expression plasmid for expression in HEK293 cells and bacmid expression plasmid for Baculovirus expression in SF9 cells. Each recombinant protein contained a C-terminal 8× His tag. Expressed proteins were purified using metal affinity chromatography [ 34 ]. The recombinant E1 (rE1) or rE2 antigen of WEEV Fleming was prepared by cloning the E1 or E2 gene into a pCRT7 bacterial expression vector and the C-terminal 6× His-tagged rE1 or rE2 was expressed in bacteria and purified using metal affinity chromatography [ 35 ]. 2.4. The First Approach to Develop Anti-WEEV or Anti-EEEV NAbs The rE2 antigens of EEEV V105-00210 and WEEV Fleming were shipped to a custom antibody service provider (Precision Antibody; PA; Columbia, MD, USA) for immunization of mice and generation of hybridoma clones that were reactive to the rE2 antigen of EEEV V105-00210 or WEEV Fleming. A total of seven hybridoma clones that were reactive to the rE2 antigen of EEEV V105-00210 and 66 clones reactive to the rE2 antigen of WEEV Fleming were generated. 2.5. The Second Approach to Develop Anti-WEEV and Anti-EEEV NAbs Female BALB/c mice (4–6 weeks old) were obtained from Charles River Canada (St Constant, QC, Canada). All mouse experiments were performed in strict accordance with the guidelines set out by the Canadian Council on Animal Care. The animal care protocol was reviewed and approved in 2013 by the Committee on the Ethics of Animal Experiments of DRDC SRC (protocol number: W1H-13-1-2-0). All efforts were made to minimize the suffering of the mice. The mice were intramuscularly (i.m.) immunized with immunogens in different formats from different strains to develop anti-EEEV or anti-WEEV Nabs, as shown in Figure 1 and Figure 2 . Three days after the last booster, five mice for each of EEEV and WEEV were sacrificed. Splenocytes were prepared and fused with myeloma cells in a standard hybridoma fusion protocol. Briefly, spleens were aseptically dissected from the immunized mice three days after the last booster, ground gently with autoclaved frosted-glass slides in DMEM, and filtered through a wire mesh screen to prepare splenocytes. Hybridomas were produced by fusing the splenocytes with Sp 2/0 myeloma cells using a Clonacell-HY Kit. After 2 weeks in a semisolid medium, five-thousand individual hybridoma clones for each virus were picked from semisolid medium and transferred to 96-well plates and cultured for 7 days in Clonacell Medium E, as previously described [ 36 ]. Hybridoma culture supernates were screened using an in vitro neutralization assay. Briefly, 20 pfu/well of EEEV PE6 or 125 pfu/well of WEEV 71V-1658 were incubated with a hybridoma culture supernate at 37 °C for 1 h. The mixtures were then added to Vero cells in 96-well plates (1 × 10 4 cells/well). Two or three days later, cells were observed under the microscope for cytopathic effects (CPE). Hybridoma clones that suppressed CPE were expanded. NAbs were purified from the cell culture supernate using a Melon Gel purification kit according to the manufacturer's protocol. The supernate was dialyzed for two exchanges (1 h each) in Melon Gel IgG Purification Buffer pH 7.0 and then was added to a column containing the Melon Gel resin. After a 5 min incubation with end-over-end mixing, the purified IgG was collected in the flowthrough. All purified IgG samples were stored in aliquots at −70 °C. 2.6. Enzyme-Linked Immunosorbent Assay (ELISA) ELISA plates were coated with sucrose-purified WEEV CBA-87 in PBS (3 µg/well) overnight at 4 °C. Following incubation, the excess was decanted and plates were blocked with PBS containing 0.02% Tween-20 and 5% non-fat dry milk (PBSTM) for 2 h at 37 °C. Following incubation, plates were either used immediately or stored at −20 °C for up to one month. Plates were washed three times with PBS containing 0.02% Tween-20 (PBST). Samples were added to the plate in duplicate and diluted down the plate in two-fold dilutions in PBSTM with 1% heat-inactivated FBS, starting at a dilution of 1:20 to 1:80. Plates were incubated for 1–2 h at 37 °C and then washed three times with PBST. After washing, goat anti-mouse IgG HRP conjugated antibody diluted 1:50,000 in PBSTM with 1% FBS was added (100 µL/well) and then incubated for 1 h at 37 °C. After another three washes with PBST, ABTS substrate (100 µL/well) was added and incubated for approximately 10 min at 37 °C. Plates were read at 410 nm with a SpectraMax M5 plate reader (Molecular Devices, Sunnyvale, CA, USA). 2.7. Western Blot Analysis rE2 antigens of EEEV V105-00210 and WEEV Fleming, or inactivated EEEV V105-00210, were denatured under reducing conditions and run on a gradient SDS-polyacrylamide gel. The proteins were transferred to a PVDF membrane, followed by a blocking step with StartingBlock to prevent antibodies from non-specifically sticking to the membrane. The membrane was incubated with supernate from hybridoma cells secreting antibodies to the rE2 antigens, followed by a secondary alkaline phosphatase-conjugated antibody. Alkaline phosphate substrate was added to visualize antibody reactivity to the rE2 antigens or EEEV V105-00210. 2.8. Plaque Reduction Neutralization Test (PRNT) Virus culture media or MAb (20 µg/mL) was mixed with an equal volume of virus suspension (400 pfu/mL) and incubated overnight at 4 °C. A total of 500 µL of each mixture was inoculated in duplicate onto Vero cells (6-well plate) and incubated for 4 days at 37 °C under a carboxymethyl cellulose overlay. After 4 days of incubation, monolayers were fixed with 10% formal saline overnight and stained with 1% crystal violet. A reduction in plaque numbers compared with positive control wells (cell culture media mixed with virus only) of ≥50% was considered to qualify as neutralization. 2.9. Titration of NAbs against Various Strains of EEEV or WEEV As shown in Figure 3 , the alphavirus neutralization test (ANT) was carried out in 96-well plates. Three-fold serial dilutions of each NAb were screened, starting with a maximum concentration of 100 µg/mL. The volume of each well was 50 µL. Subsequently, 100× the 50% tissue culture infective dose (TCID 50 ) of the virus in 50 µL was added to each well and then pre-incubated at 37 °C for 1 h to allow for neutralization of the virus. Thereafter, 1 × 10 4 cells/well (Vero) were added in a volume of 50 µL. Plates were then incubated at 37 °C under 5% CO 2 . After 3 days, the plates were examined microscopically. The NAb titer was identified as the highest dilution that resulted in 50% inhibition of CPE. 2.10. In Vivo Protective Efficacy Evaluation In vivo studies were performed in accordance with the UK Scientific Procedures (Animals) Act 1986 and the UK Codes of Practice for the Housing and Care of Animals Used in Scientific Procedures 1989 (as well the Animal Care and Use Review Office, Fort Detrick, MD, USA). The work performed in the UK was performed under Project licence 30/3166, approved by the UK Home Office 16 April 2014. Micro-chipped female BALB/c mice, aged 7–9 weeks old (18–20 g; Charles River Laboratories, Margate, UK) were suitably housed with access to food and water ad libitum in a rigid-walled CL3 isolator. Acclimatization within the CL3 isolator was for a minimum of 5 days and all mice were weighed prior to performing any procedures. Mice were inoculated with either 2.5 or 10 mg/kg of NAb via the intraperitoneal (IP) route (100 µL/mouse) 24 h prior to aerosol exposure to a cognate virus. A non-specific murine IgG MAb was administered to a group of 5 mice at a dose of 10 mg/kg, serving as the negative control. A clinical scoring system was used to closely monitor the clinical course of infection [ 37 ] and all mice were individually weighed daily. Any mouse that was observed to have pronounced mobility issues (unable to reach food and water) was immediately culled on welfare grounds in accordance with UK Home Office Project License requirements. Mice were observed a minimum of twice daily for clinical signs of infection post-exposure by an independent technician. All culls were performed according to the UK Schedule 1 method (cervical dislocation followed by confirmation of cessation of the heartbeat). When mice succumbed to disease, the brain and lungs were excised, weighed, and stored for subsequent determination of the viral load. Additionally, all mice surviving out to 14 days post-exposure were culled and brain and lungs were similarly excised and stored. Stored organs were examined for the presence of the virus. Briefly, organs were allowed to thaw and were homogenized through a fine cell sieve (40 µm sieve, Corning Falcon cell strainer, Fisher Scientific, Loughborough, UK) into 1 mL virus culture medium. Serial dilutions were prepared from homogenates in virus culture media and titrated using a standard plaque assay. Aerosol exposure was achieved using a 3-jet Collison nebulizer containing a minimum of 10 mL of the virus in virus culture medium without FBS and antibiotics, controlled and conditioned to 50% (±5%) relative humidity, using an AeroMP platform system (Biaera Technologies, Hagerstown, MD, USA). Mice were physically restrained in holding tubes and nose-only exposed for 10 min. A single 1 min sample of each aerosol exposure was taken using an all-glass impinger (AGI-30; Ace Glass, Vineland, NJ, USA) containing 10 mL PBS (at a flow rate of 12 L/min). The mean calculated presented challenge dose was determined using the viral titers obtained from the AGI-30 and Guyton's formula for the respiratory volumes of laboratory animals [ 38 ]. 2.11. Statistical Analysis Graphs were prepared using Microsoft Excel 2010 Ink and Graphpad PRISM v8. Statistical analysis was performed using IBM SPSS (V21.0) and survival characteristics were compared using stratified and unstratified pairwise comparisons of the data using the log-rank (Mantel–Cox) test. 2.1. Reagents, Cells, and Viruses High-glucose Dulbecco's minimal essential medium (DMEM), Leibovitz (L-15) medium, phosphate-buffered saline (PBS), non-fat dry milk, fetal bovine serum (FBS), Synth-a-Freeze defined protein-free cryopreservation medium, Tween-20, SDS-polyacrylaminde gels, StartingBlock (PBS) blocking buffer, polyvinylidene difluoride (PVDF) membrane, agarose, penicillin, streptomycin, Melon Gel purification kits, Nunc cryovials, and Invitrogen cloning kits were purchased from ThermoFisher Scientific (Loughborough, UK; Waltham, MA, USA; Ottawa, ON, Canada) or Sigma Aldrich (Poole, UK). TiterMax Gold adjuvant was purchased from Cedarlane (Burlington, ON, Canada). Clonacell-HY Kit was from Stem Cell Technologies (Vancouver, BC, Canada). The AdEasy system was purchased from Qbiogene (Carlsbad, CA, USA). Restriction enzymes were from New England Biolabs (Mississauga, ON, Canada). Cell culture flasks and plates were from VWR (Mississauga, ON, Canada). Goat anti-mouse IgG horseradish peroxidase (HRP)-conjugated antibody and 2,2′-Azino-di(3-ethylbenzthiazoline-6-sulfonate) (ABTS) substrate were from KPL (Gaithersburg, MD, USA). Isoflurane was obtained from Baxter (Mississauga, ON, Canada). Alkaline phosphate substrate and goat anti-mouse alkaline phosphate conjugated antibody were from Bio-Rad Laboratories (Hercules, CA, USA). A non-specific murine IgG monoclonal antibody (MAb) and dimethyl sulfoxide (DMSO) were obtained from Sigma Aldrich (Poole, UK). Vero (ATCC CCL-81) cells, Sp 2/0 mouse myeloma cells, and HEK 293 cells were from the American Type Culture Collection (ATCC, Manassas, VA, USA), and the European Collection of Animal Cell Cultures (Salisbury, UK). EEEV strains PE6, FL93-969, and Williams; Madariaga virus (435731); and WEEV strains B11 and CBA-87 were kindly provided by Dr. George Ludwig, United States Army Medical Research Institute for Infectious Diseases (USAMRIID), Frederick, MD, USA. WEEV strain 71V-1658 was kindly provided by Dr. Nick Karabatsos, CDC, Fort Collins, CO, USA; WEEV strains Fleming and California were purchased from ATCC; WEEV strains McMillan, Mn548, and Mn520 were kindly provided by Drs. Mike Drebot and Harvey Artsorb, National Microbiology Laboratory, Winnipeg, MN, Canada. 2.2. Cell and Virus Maintenance Hybridoma cell lines were maintained in DMEM supplemented with 10% FBS. HEK 293 cells were grown in DMEM supplemented with 5% FBS. Vero cells were grown in DMEM supplemented with 5% FBS, or DMEM supplemented with 10% FBS, 2 mM l -glutamine, 50 IU/mL penicillin, and 50 µg/mL streptomycin. All cell cultures were at 37 °C in a 5% CO 2 humidified atmosphere, except for the virus culture in L15 medium, which had no added CO 2 . For cryopreservation, cell lines were resuspended in Synth-a-Freeze defined protein-free cryopreservation medium, or FBS supplemented with 20% ( v/v ) DMSO, and 1 mL aliquots were transferred to cryovials and stored in liquid nitrogen. Experiments with live viruses were carried out at two establishments: the Defence Science and Technology Laboratory (Dstl), UK, under UK Advisory Committee on Dangerous Pathogens, Containment Level 3 (CL3) conditions, and the Defence Research and Development Canada, Suffield Research Centre (DRDC SRC) CL3 facilities in compliance with Public Health Agency of Canada and Canadian Food Inspection Agency Guidelines. Virulent virus stocks were prepared at Dstl by inoculating suckling mouse pups intra-cranially with the virus and allowing them to become moribund (24 h after inoculation) before culling with an overdose of sodium pentobarbital. All animal studies were carried out in accordance with the UK Animals (Scientific Procedures) Act of 1986 and the Codes of Practice for the Housing and Care of Animals used in Scientific Procedures 1989. The work was performed under Project licence 30/3166 and was approved by the UK Home Office on 16 April 2014. The virus was harvested by extracting tissue through the dorsal cranium with a large-bore syringe needle and mixed with L-15 medium supplemented with 2% ( v/v ) FBS, 2 mM l -glutamine, 50 IU/mL penicillin, and 50 µg/mL streptomycin (virus culture medium). This was then passed through a 70 µm nylon cell strainer, clarified at 10,000 rpm for 10 min in an SW28 rotor (Beckman Coulter, UK), and stored at −80 °C. Standard plaque assays were performed in a 24-well plate format with 100 µL/well of virus inoculum applied in duplicate to Vero cell monolayers under a carboxymethyl cellulose overlay. Plates were incubated in humidified conditions at 37 °C for 3–4 days prior to fixation with 10% ( v/v ) formal saline solution and visualization of plaques was achieved using 1% crystal violet solution. The limit of detection in this assay is 10 plaque-forming units (pfu)/mL of the original sample. The use of this production method minimizes the potential for loss of virulence factors as a result of cell culture adaptation, and also provides a more representative wild-type virus population (quasispecies). Virus seed stocks were made at DRDC SRC via the inoculation of Vero cells with virus suspensions at a multiplicity of infection of less than 0.1. The supernates were clarified using centrifugation and stored in aliquots at −70 °C [ 32 ]. 2.3. Immunogens An encephalomyelitis vaccine for horses, consisting of formalin-inactivated EEEV and WEEV, and tetanus toxoid (Zoetis vaccine) was purchased from Zoetis Canada INC (Kirkland, QC, Canada). An adenovirus DNA vaccine, namely, pAd-EEEV PE6, was constructed by cloning the structural proteins of EEEV strain PE6 into an adenovirus vector using the AdEasy system according to the manufacturer's protocol. The recombinant adenoviral construct pAd-EEEV PE6 was linearized with Pac I and transfected into HEK 293 cells cultured in DMEM with 5% FBS for amplification and then the amplified adenovirus was purified via chromatography. A plasmid DNA vaccine, namely, pVHX-6-WEEV 71V-1658, was described previously [ 33 ] and expresses the structural proteins (capsid, E3, E2, 6K, and E1) of WEEV 71V-1658. Recombinant E2 (rE2) of EEEV V105-00210 was prepared by cloning the E2 gene into a mammalian expression plasmid for expression in HEK293 cells and bacmid expression plasmid for Baculovirus expression in SF9 cells. Each recombinant protein contained a C-terminal 8× His tag. Expressed proteins were purified using metal affinity chromatography [ 34 ]. The recombinant E1 (rE1) or rE2 antigen of WEEV Fleming was prepared by cloning the E1 or E2 gene into a pCRT7 bacterial expression vector and the C-terminal 6× His-tagged rE1 or rE2 was expressed in bacteria and purified using metal affinity chromatography [ 35 ]. 2.4. The First Approach to Develop Anti-WEEV or Anti-EEEV NAbs The rE2 antigens of EEEV V105-00210 and WEEV Fleming were shipped to a custom antibody service provider (Precision Antibody; PA; Columbia, MD, USA) for immunization of mice and generation of hybridoma clones that were reactive to the rE2 antigen of EEEV V105-00210 or WEEV Fleming. A total of seven hybridoma clones that were reactive to the rE2 antigen of EEEV V105-00210 and 66 clones reactive to the rE2 antigen of WEEV Fleming were generated. 2.5. The Second Approach to Develop Anti-WEEV and Anti-EEEV NAbs Female BALB/c mice (4–6 weeks old) were obtained from Charles River Canada (St Constant, QC, Canada). All mouse experiments were performed in strict accordance with the guidelines set out by the Canadian Council on Animal Care. The animal care protocol was reviewed and approved in 2013 by the Committee on the Ethics of Animal Experiments of DRDC SRC (protocol number: W1H-13-1-2-0). All efforts were made to minimize the suffering of the mice. The mice were intramuscularly (i.m.) immunized with immunogens in different formats from different strains to develop anti-EEEV or anti-WEEV Nabs, as shown in Figure 1 and Figure 2 . Three days after the last booster, five mice for each of EEEV and WEEV were sacrificed. Splenocytes were prepared and fused with myeloma cells in a standard hybridoma fusion protocol. Briefly, spleens were aseptically dissected from the immunized mice three days after the last booster, ground gently with autoclaved frosted-glass slides in DMEM, and filtered through a wire mesh screen to prepare splenocytes. Hybridomas were produced by fusing the splenocytes with Sp 2/0 myeloma cells using a Clonacell-HY Kit. After 2 weeks in a semisolid medium, five-thousand individual hybridoma clones for each virus were picked from semisolid medium and transferred to 96-well plates and cultured for 7 days in Clonacell Medium E, as previously described [ 36 ]. Hybridoma culture supernates were screened using an in vitro neutralization assay. Briefly, 20 pfu/well of EEEV PE6 or 125 pfu/well of WEEV 71V-1658 were incubated with a hybridoma culture supernate at 37 °C for 1 h. The mixtures were then added to Vero cells in 96-well plates (1 × 10 4 cells/well). Two or three days later, cells were observed under the microscope for cytopathic effects (CPE). Hybridoma clones that suppressed CPE were expanded. NAbs were purified from the cell culture supernate using a Melon Gel purification kit according to the manufacturer's protocol. The supernate was dialyzed for two exchanges (1 h each) in Melon Gel IgG Purification Buffer pH 7.0 and then was added to a column containing the Melon Gel resin. After a 5 min incubation with end-over-end mixing, the purified IgG was collected in the flowthrough. All purified IgG samples were stored in aliquots at −70 °C. 2.6. Enzyme-Linked Immunosorbent Assay (ELISA) ELISA plates were coated with sucrose-purified WEEV CBA-87 in PBS (3 µg/well) overnight at 4 °C. Following incubation, the excess was decanted and plates were blocked with PBS containing 0.02% Tween-20 and 5% non-fat dry milk (PBSTM) for 2 h at 37 °C. Following incubation, plates were either used immediately or stored at −20 °C for up to one month. Plates were washed three times with PBS containing 0.02% Tween-20 (PBST). Samples were added to the plate in duplicate and diluted down the plate in two-fold dilutions in PBSTM with 1% heat-inactivated FBS, starting at a dilution of 1:20 to 1:80. Plates were incubated for 1–2 h at 37 °C and then washed three times with PBST. After washing, goat anti-mouse IgG HRP conjugated antibody diluted 1:50,000 in PBSTM with 1% FBS was added (100 µL/well) and then incubated for 1 h at 37 °C. After another three washes with PBST, ABTS substrate (100 µL/well) was added and incubated for approximately 10 min at 37 °C. Plates were read at 410 nm with a SpectraMax M5 plate reader (Molecular Devices, Sunnyvale, CA, USA). 2.7. Western Blot Analysis rE2 antigens of EEEV V105-00210 and WEEV Fleming, or inactivated EEEV V105-00210, were denatured under reducing conditions and run on a gradient SDS-polyacrylamide gel. The proteins were transferred to a PVDF membrane, followed by a blocking step with StartingBlock to prevent antibodies from non-specifically sticking to the membrane. The membrane was incubated with supernate from hybridoma cells secreting antibodies to the rE2 antigens, followed by a secondary alkaline phosphatase-conjugated antibody. Alkaline phosphate substrate was added to visualize antibody reactivity to the rE2 antigens or EEEV V105-00210. 2.8. Plaque Reduction Neutralization Test (PRNT) Virus culture media or MAb (20 µg/mL) was mixed with an equal volume of virus suspension (400 pfu/mL) and incubated overnight at 4 °C. A total of 500 µL of each mixture was inoculated in duplicate onto Vero cells (6-well plate) and incubated for 4 days at 37 °C under a carboxymethyl cellulose overlay. After 4 days of incubation, monolayers were fixed with 10% formal saline overnight and stained with 1% crystal violet. A reduction in plaque numbers compared with positive control wells (cell culture media mixed with virus only) of ≥50% was considered to qualify as neutralization. 2.9. Titration of NAbs against Various Strains of EEEV or WEEV As shown in Figure 3 , the alphavirus neutralization test (ANT) was carried out in 96-well plates. Three-fold serial dilutions of each NAb were screened, starting with a maximum concentration of 100 µg/mL. The volume of each well was 50 µL. Subsequently, 100× the 50% tissue culture infective dose (TCID 50 ) of the virus in 50 µL was added to each well and then pre-incubated at 37 °C for 1 h to allow for neutralization of the virus. Thereafter, 1 × 10 4 cells/well (Vero) were added in a volume of 50 µL. Plates were then incubated at 37 °C under 5% CO 2 . After 3 days, the plates were examined microscopically. The NAb titer was identified as the highest dilution that resulted in 50% inhibition of CPE. 2.10. In Vivo Protective Efficacy Evaluation In vivo studies were performed in accordance with the UK Scientific Procedures (Animals) Act 1986 and the UK Codes of Practice for the Housing and Care of Animals Used in Scientific Procedures 1989 (as well the Animal Care and Use Review Office, Fort Detrick, MD, USA). The work performed in the UK was performed under Project licence 30/3166, approved by the UK Home Office 16 April 2014. Micro-chipped female BALB/c mice, aged 7–9 weeks old (18–20 g; Charles River Laboratories, Margate, UK) were suitably housed with access to food and water ad libitum in a rigid-walled CL3 isolator. Acclimatization within the CL3 isolator was for a minimum of 5 days and all mice were weighed prior to performing any procedures. Mice were inoculated with either 2.5 or 10 mg/kg of NAb via the intraperitoneal (IP) route (100 µL/mouse) 24 h prior to aerosol exposure to a cognate virus. A non-specific murine IgG MAb was administered to a group of 5 mice at a dose of 10 mg/kg, serving as the negative control. A clinical scoring system was used to closely monitor the clinical course of infection [ 37 ] and all mice were individually weighed daily. Any mouse that was observed to have pronounced mobility issues (unable to reach food and water) was immediately culled on welfare grounds in accordance with UK Home Office Project License requirements. Mice were observed a minimum of twice daily for clinical signs of infection post-exposure by an independent technician. All culls were performed according to the UK Schedule 1 method (cervical dislocation followed by confirmation of cessation of the heartbeat). When mice succumbed to disease, the brain and lungs were excised, weighed, and stored for subsequent determination of the viral load. Additionally, all mice surviving out to 14 days post-exposure were culled and brain and lungs were similarly excised and stored. Stored organs were examined for the presence of the virus. Briefly, organs were allowed to thaw and were homogenized through a fine cell sieve (40 µm sieve, Corning Falcon cell strainer, Fisher Scientific, Loughborough, UK) into 1 mL virus culture medium. Serial dilutions were prepared from homogenates in virus culture media and titrated using a standard plaque assay. Aerosol exposure was achieved using a 3-jet Collison nebulizer containing a minimum of 10 mL of the virus in virus culture medium without FBS and antibiotics, controlled and conditioned to 50% (±5%) relative humidity, using an AeroMP platform system (Biaera Technologies, Hagerstown, MD, USA). Mice were physically restrained in holding tubes and nose-only exposed for 10 min. A single 1 min sample of each aerosol exposure was taken using an all-glass impinger (AGI-30; Ace Glass, Vineland, NJ, USA) containing 10 mL PBS (at a flow rate of 12 L/min). The mean calculated presented challenge dose was determined using the viral titers obtained from the AGI-30 and Guyton's formula for the respiratory volumes of laboratory animals [ 38 ]. 2.11. Statistical Analysis Graphs were prepared using Microsoft Excel 2010 Ink and Graphpad PRISM v8. Statistical analysis was performed using IBM SPSS (V21.0) and survival characteristics were compared using stratified and unstratified pairwise comparisons of the data using the log-rank (Mantel–Cox) test. 3. Results 3.1. Generation and Selection of Neutralizing Hybridoma Clones Two independent mouse immunization regimens were designed with varying immunogens to maximize the identification of novel cross-strain NAbs for EEEV and WEEV. First, the rE2 antigen of EEEV V105-00210 or WEEV Fleming was used to immunize the mice. As a result, 7 hybridoma clones were generated, secreting antibodies that bound to EEEV V105-00210 rE2, and 66 clones that were reactive with WEEV Fleming rE2. The seven clones that were reactive with EEEV V105-00210 rE2 were further tested via Western blotting using homologous rE2 or the whole virus as the target antigen. None of the seven MAbs bound to the rE2 or authentic E2 of EEEV in the Western blot of the denatured antigen. The 66 WEEV Fleming rE2 reactive MAbs were also tested via Western blotting, using rE2 of WEEV Fleming as the target antigen and 55 MAbs were confirmed to be capable of binding to the Western-blotted rE2 of WEEV Fleming. The 66 MAbs were also assessed using ELISA with a heterologous strain (WEEV CBA-87 as the coating antigen), where 47 MAbs were found to be able to bind to WEEV CBA-87. Seven EEEV rE2-reactive MAbs and the top 14 WEEV rE2-reactive MAbs based on the Western blot results were screened for their ability to neutralize the cognate viable virus. The remainder of the WEEV rE2-reactive MAbs were not further screened since their binding signals to the rE2 in the Western blot were quite low. None of the anti-EEEV MAbs effectively neutralized the EEEV PE6 infectivity for Vero cells. Meanwhile, two anti-WEEV Fleming E2 MAbs, namely, 8C2 and 8H10, neutralized the WEEV Fleming infectivity for Vero cells. As an alternative approach, mice were immunized with EEEV or WEEV immunogens in different formats with different strains (see Figure 1 and Figure 2 ). These included pAd-EEEV PE6, formalin-inactivated WEEV/EEEV viruses (Zoetis vaccine), pVHX-6-WEEV 71V-1658, and rE1/rE2 of WEEV Fleming. The hybridomas were initially screened with an in vitro neutralization assay in Vero cells to identify anti-EEEV PE6 or anti-WEEV 71V-1658 NAbs. Ultimately, three anti-EEEV and five additional anti-WEEV MAbs were identified that demonstrated neutralizing activity against EEEV PE6 or WEEV 71V-1658. 3.2. NAb Titration and Evaluation against Various Strains of EEEV or WEEV The three anti-EEEV and seven anti-WEEV NAbs were produced and purified from the hybridoma culture supernate. Anti-EEEV and anti-WEEV neutralizing activity was evaluated using ANT ( Figure 3 ). The ANT results for anti-EEEV NAbs are summarized in Table 1 . All three anti-EEEV NAbs yielded positive results, neutralizing 50% EEEV PE6 infectivity in Vero cells at concentrations ranging from 2.4 to 21.6 µg/mL. G1-2-H4 and G1-4-C3 possessed the highest efficacy and were further evaluated against two additional EEEV strains (FL93-969 and Williams) and a single strain of Madariaga virus (435731), demonstrating the neutralization of all the strains tested. As shown in Table 2 , similar experiments with all seven anti-WEEV NAbs showed positive results, neutralizing 50% of WEEV Fleming infectivity in Vero cells at concentrations ranging from 0.03 to 18 µg/mL. Three anti-WEEV NAbs with the highest efficacy (0.03 µg/mL), namely, G8-2-H9, 12 WA, and 8C2, were evaluated against another seven WEEV strains. Two of these NAbs, namely, G8-2-H9 and 12WA, neutralized 71V-1658, CBA-87, B11, Mn548, and Mn520, but not California or McMillan. Additionally, 8C2 and 8H10 were assessed using standard PRNT; 10 µg/mL of antibodies resulted in 45% (8H10) and 100% (8C2) inhibition of plaques (mean plaque number of 87 in the untreated control wells). 3.3. In Vivo Protective Efficacy Two anti-EEEV and three anti-WEEV NAbs were down-selected for assessment in vivo to determine the protective efficacy against EEEV or WEEV aerosol challenges in a pre-exposure prophylaxis setting [ 37 ]. As shown in Figure 4 , anti-EEEV NAb G1-2-H4 provided 40% (at a dose of 2.5 mg/kg) and 60% (at a dose of 10 mg/kg) protection, while G1-4-C3 provided 100% protection at both doses in an EEEV PE6 lethal aerosol challenge model. Surviving mice in the Nab-inoculated groups were free of clinical signs and maintained a typical weight profile throughout the study. Statistical analysis (log-rank Mantel–Cox) using pairwise comparisons identified a significant benefit in the prophylactic use of anti-EEEV NAb G1-4-C3 ( p ≤ 0.001) at either dose when compared with the non-specific IgG control group. As mice succumbed to disease (the humane endpoint had been reached), the viral load in the brain and lungs was determined. Mice that were inoculated with either dose of G1-2-H4 that succumbed to disease (days 3–5 post aerosol exposure) had a mean of 2.17 × 10 10 pfu/g in the brain and 2.82 × 10 4 pfu/g in the lungs. This was comparable with viral loads that were observed in mice inoculated with non-specific IgG that succumbed to disease on days 3–5 post aerosol exposure, with a mean of 2.98 × 10 10 pfu/g in the brain and 2.58× 10 4 pfu/g in the lung. In a WEEV Fleming sublethal aerosol challenge model, all three anti-WEEV Nabs, either at a dose of 10 or 2.5 mg/kg, provided up to 100% protection ( Figure 5 ). NAb 8C2 was able to provide 100% protection when mice were inoculated with either dose. The mortality rate of control mice did not reach 100% in this study due to the lower than anticipated exposure dose of WEEV (mean of 7.4 median lethal dose (MLD)). In addition to the mortality rate, it is important to note the differences in clinical outcomes, as all NAb-treated mice were free of clinical signs (ruffled fur, hunched posture, lack of mobility, behavioral changes, and weight loss) throughout the study. Statistical analysis (log-rank Mantel–Cox) using pairwise comparisons identified a significant benefit ( p ≤ 0.05) in the prophylactic use of these anti-WEEV NAbs when compared with the non-specific IgG control group for 8C2 at either dose, 12WA at 10 mg/kg, and G8-2-H9 at 2.5 mg/kg ( p = 0.034). Including both doses as stratification pairwise comparisons did not identify any significant differences between the anti-WEEV NAbs in this study ( p ≥ 0.05), highlighting the ability of all three NAbs to provide protection against aerosol exposure. During the acute phase of infection, several mice succumbed to disease (the humane endpoint had been reached) and the viral load in the brain and lungs of these animals was determined. Mice inoculated with 10 mg/kg non-specific IgG that succumbed to disease on days 3–4 post aerosol exposure achieved mean titers of 6.05 × 10 9 pfu/g of WEEV Fleming in the brain and 1.05 × 10 2 pfu/g in the lung. The individual mice inoculated with 2.5 mg/kg 12WA and 10 mg/kg G8-2-H9 that succumbed to disease on day 4 post aerosol exposure were also found to have high viral loads in the brain (1.3 × 10 10 and 4.8 × 10 9 pfu/g, respectively) and lung (3.6 × 10 7 and 6.9 × 10 5 pfu/g, respectively). 3.1. Generation and Selection of Neutralizing Hybridoma Clones Two independent mouse immunization regimens were designed with varying immunogens to maximize the identification of novel cross-strain NAbs for EEEV and WEEV. First, the rE2 antigen of EEEV V105-00210 or WEEV Fleming was used to immunize the mice. As a result, 7 hybridoma clones were generated, secreting antibodies that bound to EEEV V105-00210 rE2, and 66 clones that were reactive with WEEV Fleming rE2. The seven clones that were reactive with EEEV V105-00210 rE2 were further tested via Western blotting using homologous rE2 or the whole virus as the target antigen. None of the seven MAbs bound to the rE2 or authentic E2 of EEEV in the Western blot of the denatured antigen. The 66 WEEV Fleming rE2 reactive MAbs were also tested via Western blotting, using rE2 of WEEV Fleming as the target antigen and 55 MAbs were confirmed to be capable of binding to the Western-blotted rE2 of WEEV Fleming. The 66 MAbs were also assessed using ELISA with a heterologous strain (WEEV CBA-87 as the coating antigen), where 47 MAbs were found to be able to bind to WEEV CBA-87. Seven EEEV rE2-reactive MAbs and the top 14 WEEV rE2-reactive MAbs based on the Western blot results were screened for their ability to neutralize the cognate viable virus. The remainder of the WEEV rE2-reactive MAbs were not further screened since their binding signals to the rE2 in the Western blot were quite low. None of the anti-EEEV MAbs effectively neutralized the EEEV PE6 infectivity for Vero cells. Meanwhile, two anti-WEEV Fleming E2 MAbs, namely, 8C2 and 8H10, neutralized the WEEV Fleming infectivity for Vero cells. As an alternative approach, mice were immunized with EEEV or WEEV immunogens in different formats with different strains (see Figure 1 and Figure 2 ). These included pAd-EEEV PE6, formalin-inactivated WEEV/EEEV viruses (Zoetis vaccine), pVHX-6-WEEV 71V-1658, and rE1/rE2 of WEEV Fleming. The hybridomas were initially screened with an in vitro neutralization assay in Vero cells to identify anti-EEEV PE6 or anti-WEEV 71V-1658 NAbs. Ultimately, three anti-EEEV and five additional anti-WEEV MAbs were identified that demonstrated neutralizing activity against EEEV PE6 or WEEV 71V-1658. 3.2. NAb Titration and Evaluation against Various Strains of EEEV or WEEV The three anti-EEEV and seven anti-WEEV NAbs were produced and purified from the hybridoma culture supernate. Anti-EEEV and anti-WEEV neutralizing activity was evaluated using ANT ( Figure 3 ). The ANT results for anti-EEEV NAbs are summarized in Table 1 . All three anti-EEEV NAbs yielded positive results, neutralizing 50% EEEV PE6 infectivity in Vero cells at concentrations ranging from 2.4 to 21.6 µg/mL. G1-2-H4 and G1-4-C3 possessed the highest efficacy and were further evaluated against two additional EEEV strains (FL93-969 and Williams) and a single strain of Madariaga virus (435731), demonstrating the neutralization of all the strains tested. As shown in Table 2 , similar experiments with all seven anti-WEEV NAbs showed positive results, neutralizing 50% of WEEV Fleming infectivity in Vero cells at concentrations ranging from 0.03 to 18 µg/mL. Three anti-WEEV NAbs with the highest efficacy (0.03 µg/mL), namely, G8-2-H9, 12 WA, and 8C2, were evaluated against another seven WEEV strains. Two of these NAbs, namely, G8-2-H9 and 12WA, neutralized 71V-1658, CBA-87, B11, Mn548, and Mn520, but not California or McMillan. Additionally, 8C2 and 8H10 were assessed using standard PRNT; 10 µg/mL of antibodies resulted in 45% (8H10) and 100% (8C2) inhibition of plaques (mean plaque number of 87 in the untreated control wells). 3.3. In Vivo Protective Efficacy Two anti-EEEV and three anti-WEEV NAbs were down-selected for assessment in vivo to determine the protective efficacy against EEEV or WEEV aerosol challenges in a pre-exposure prophylaxis setting [ 37 ]. As shown in Figure 4 , anti-EEEV NAb G1-2-H4 provided 40% (at a dose of 2.5 mg/kg) and 60% (at a dose of 10 mg/kg) protection, while G1-4-C3 provided 100% protection at both doses in an EEEV PE6 lethal aerosol challenge model. Surviving mice in the Nab-inoculated groups were free of clinical signs and maintained a typical weight profile throughout the study. Statistical analysis (log-rank Mantel–Cox) using pairwise comparisons identified a significant benefit in the prophylactic use of anti-EEEV NAb G1-4-C3 ( p ≤ 0.001) at either dose when compared with the non-specific IgG control group. As mice succumbed to disease (the humane endpoint had been reached), the viral load in the brain and lungs was determined. Mice that were inoculated with either dose of G1-2-H4 that succumbed to disease (days 3–5 post aerosol exposure) had a mean of 2.17 × 10 10 pfu/g in the brain and 2.82 × 10 4 pfu/g in the lungs. This was comparable with viral loads that were observed in mice inoculated with non-specific IgG that succumbed to disease on days 3–5 post aerosol exposure, with a mean of 2.98 × 10 10 pfu/g in the brain and 2.58× 10 4 pfu/g in the lung. In a WEEV Fleming sublethal aerosol challenge model, all three anti-WEEV Nabs, either at a dose of 10 or 2.5 mg/kg, provided up to 100% protection ( Figure 5 ). NAb 8C2 was able to provide 100% protection when mice were inoculated with either dose. The mortality rate of control mice did not reach 100% in this study due to the lower than anticipated exposure dose of WEEV (mean of 7.4 median lethal dose (MLD)). In addition to the mortality rate, it is important to note the differences in clinical outcomes, as all NAb-treated mice were free of clinical signs (ruffled fur, hunched posture, lack of mobility, behavioral changes, and weight loss) throughout the study. Statistical analysis (log-rank Mantel–Cox) using pairwise comparisons identified a significant benefit ( p ≤ 0.05) in the prophylactic use of these anti-WEEV NAbs when compared with the non-specific IgG control group for 8C2 at either dose, 12WA at 10 mg/kg, and G8-2-H9 at 2.5 mg/kg ( p = 0.034). Including both doses as stratification pairwise comparisons did not identify any significant differences between the anti-WEEV NAbs in this study ( p ≥ 0.05), highlighting the ability of all three NAbs to provide protection against aerosol exposure. During the acute phase of infection, several mice succumbed to disease (the humane endpoint had been reached) and the viral load in the brain and lungs of these animals was determined. Mice inoculated with 10 mg/kg non-specific IgG that succumbed to disease on days 3–4 post aerosol exposure achieved mean titers of 6.05 × 10 9 pfu/g of WEEV Fleming in the brain and 1.05 × 10 2 pfu/g in the lung. The individual mice inoculated with 2.5 mg/kg 12WA and 10 mg/kg G8-2-H9 that succumbed to disease on day 4 post aerosol exposure were also found to have high viral loads in the brain (1.3 × 10 10 and 4.8 × 10 9 pfu/g, respectively) and lung (3.6 × 10 7 and 6.9 × 10 5 pfu/g, respectively). 4. Discussion The receptor-binding envelope protein E2 of alphaviruses is considered to be the major viral antigen that elicits protective antibodies [ 5 , 6 ]. Therefore, we chose to use a recombinant form of this antigen, namely, rE2 of EEEV and WEEV, to immunize mice and generate NAbs. A total of 73 MAbs, 7 reactive to EEEV V105-00210 rE2 and 66 reactive to WEEV Fleming rE2 in ELISA, were generated. None of EEEV V105-00210 reactive MAbs bound recombinant or authentic E2 in denaturing Western blotting, whereas 55 out of the 66 WEEV Fleming reactive MAbs bound to rE2 of WEEV Fleming in denaturing Western blotting. Two of these 55 anti-WEEV Fleming rE2 clones, namely, 8C2 and 8H10, were found to have neutralizing activity against WEEV Fleming infectivity for Vero cells. However, none of the seven anti-EEEV MAbs were able to neutralize EEEV PE6 infectivity for Vero cells, although it is possible that these clones might neutralize the homologous strain, namely, EEEV V105-00210, which was not tested in this study. The host antibody response plays a pivotal role in the prevention of, recovery from, and treatment of viral infections. Antibodies against the numerous epitopes on virus proteins can be divided into neutralizing and non-neutralizing categories. Only a small fraction of antibodies are NAbs, which are capable of blocking the virus infection of host cells by interfering with virion binding to receptors on cells, blocking uptake, preventing uncoating of the genomes in cells, or inducing aggregation of virus particles [ 39 ]. The vast majority of antibodies are non-NAbs, which bind specifically to virus particles, but do not neutralize viral infectivity [ 40 ]. This may explain why only 2 out of 14 anti-WEEV MAbs that were tested had neutralizing activity. However, other factors, such as improperly folded rE2 and inadequately presented rE2 as compared with the natural E1–E2 dimers, are not excluded as explanations. The E2 protein of alphaviruses binds to host cell surface receptors to trigger virus internalization and transport the virus into acidic intracellular vesicles. Within the low pH environment of the endosome, conformational changes in the envelope proteins allow E1 to insert its hydrophobic fusion loop into the membrane of the endosome. Additional conformational changes bring the endosomal and viral membranes together, thus causing membrane fusion and virus infection [ 41 ]. E1 plays a pivotal role in alphavirus internalization into host cells. Das' study showed that both E1 and E2 were directly involved in contact with the host immune system and elicitation of host immune responses [ 42 ]. Indeed, E1 and E2 are translated and processed through the endoplasmic reticulum and Golgi apparatus together and E2 would not be expressed on the surface of infected cells without association with E1. Without E1, E2 might not be processed and folded into the same conformation that it has on the surface of the virus [ 43 ]. Therefore, our second approach was to use both E1 and E2 as immunogens in either heterodimer or mixture formats from various strains to immunize the mice and increase the probability of eliciting not only neutralizing but also cross-strain neutralizing anti-EEEV or anti-WEEV hybridoma clones. As such, the inactivated EEEV and WEEV viruses that were provided by the Zoetis vaccine were used to prime the mice. The EEEV and WEEV strains in this vaccine are unknown due to commercial confidentiality. Afterward, the mice were given two boosters immunizations, providing all the structural proteins of EEEV PE6 (adenovirus vector) or WEEV 71V-1658 (DNA plasmid vector, supplemented with rE1 and rE2 proteins of WEEV Fleming). In this way, the opportunities for B-cell clones to recognize epitopes that were shared by different EEEV or WEEV strains were improved. In general, the conventional approach for isolating NAbs from immunized mice is to identify antigen-binding clones first using an immunoassay and then identify neutralizing clones via an in vitro neutralization assay using these antigen-binding clones. Since infectious alphaviruses are CL3 agents, which cause potentially lethal human diseases, their use in an immunoassay is prohibited in CL2 laboratories. Instead, either inactivated viruses or recombinant antigens are used for the immunoassay. However, the conformation of inactivated virus or recombinant antigens might not be identical to an authentic virus. Any change might affect the neutralizing epitopes that are recognized by NAbs and these NAb clones would be missed by the immunoassay. In order to overcome this hurdle, an in vitro neutralization assay using live viruses was applied to screen for anti-EEEV or anti-WEEV neutralizing clones from approximately 5000 hybridoma clones each. This led to the identification of three anti-EEEV and five additional anti-WEEV NAbs. Eight WEEV strains were previously compared based on virulence and genetic diversity, where the eight strains could be divided into two groups: a high-virulence group consisting of strains California, Fleming, and McMillan; and a low-virulence group, including strains CBA-87, Mn548, B11, Mn520, and 71V-1658 [ 32 ]. To investigate whether these NAbs had cross-strain activity against EEEV or WEEV infectivity, the most promising candidates with the highest neutralizing titers (two anti-EEEV NAbs and two anti-WEEV NAbs) were evaluated against multiple EEEV or WEEV strains, respectively. Anti-EEEV NAbs G1-2-H4 and G1-4-C3, which were developed from the second immunization approach, were able to neutralize all EEEV strains tested and one Madariaga strain tested. Meanwhile, anti-WEEV NAbs neutralized six out of eight tested strains. Only the California and McMillan strains were resistant to neutralization by these two NAbs. It is notable that NAbs G8-2-H9 and 12WA had activity against members of both groups. Interestingly, the California and McMillan strains share four amino acids in E2 that are not present in the other strains of WEEV tested [ 32 ], suggesting the epitope for these two NAbs may involve this region. WEEV 71V-1658 was used to initially screen the hybridoma clones, and it is possible to have missed antibodies that neutralize the California and McMillan strains via an epitope involving this four-amino-acid stretch. Both G8-2-H9 and 12WA were developed using the second immunization approach in which the mice were immunized with the antigens from multiple WEEV strains. Unlike G8-2-H9 and 12WA, 8C2 was developed from the first immunization approach in which the mice were immunized with the rE2 from only the Fleming strain. This difference in the immunization approaches may explain why 8C2 was the only identified anti-WEEV NAb that did not neutralize any of the heterologous strains tested. These results indicate that these cross-strain NAbs recognize a common or conserved neutralizing epitope that is shared by several strains of EEEV or WEEV, respectively. Furthermore, the results indicated that immunization strategies for antiviral MAbs should include antigens from multiple strains to improve the potential for generating MAb-based MCMs with activity across the target species. In this study, the five most promising NAbs were evaluated in an in vivo prophylactic efficacy setting, inoculating mice one day prior to aerosol exposure to the virus. All five NAbs demonstrated high levels of protection against EEEV or WEEV infection in mice. Regardless of the NAb dose, complete protection was afforded by G1-4-C3 against a lethal aerosol exposure of EEEV PE6, where surviving animals were free of clinical signs throughout the study and maintained typical weight profiles. Complete protection was also afforded by 8C2 at both NAb dose levels against a sub-lethal aerosol exposure of WEEV Fleming, where surviving animals were free of clinical signs throughout the study and maintained typical weight profiles and clinical scores. The lower clinical score in the WEEV control animals relative to EEEV control animals was reflected by the less than 100% mortality in the WEEV control mice. This was due to the nature of aerosol challenge experiments, where the anticipated, or intended, dose and the actual delivered dose might vary. Importantly, we were able to calculate the actual delivered dose after exposure and it was thus clear that the delivered dose of WEEV was less than intended and sub-lethal for this established model [ 37 ], whereas the EEEV challenge was in the intended lethal range. Nevertheless, the extensive clinical score data, and its correlation with the level of mortality, provided a clear test of the ability of the NAbs to protect against a WEEV challenge. To date, this is the first description of cross-strain neutralizing anti-EEEV and anti-WEEV NAbs. These data provide an evidence base on which to further explore the potential of NAbs as cross-strain antibody-based MCMs against these encephalitic alphaviruses, for which there are no licensed MCMs for use in humans.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9534887/

Targeted inhibition of tumor-derived exosomes as a novel therapeutic option for cancer

Mounting evidence indicates that tumor-derived exosomes (TDEs) play critical roles in tumor development and progression by regulating components in the tumor microenvironment (TME) in an autocrine or paracrine manner. Moreover, due to their delivery of critical molecules that react to chemotherapy and immunotherapy, TDEs also contribute to tumor drug resistance and impede the effective response of antitumor immunotherapy, thereby leading to poor clinical outcomes. There is a pressing need for the inhibition or removal of TDEs to facilitate the treatment and prognosis of cancer patients. Here, in the present review, we systematically overviewed the current strategies for TDE inhibition and clearance, providing novel insights for future tumor interventions in translational medicine. Moreover, existing challenges and potential prospects for TDE-targeted cancer therapy are also discussed to bridge the gaps between progress and promising applications. Introduction Since the discovery of exosomes in 1983, a new method of cell-to-cell communication was introduced to extend our perspectives of numerous physiological and pathological processes 1 , 2 . Exosomes are specific extracellular vesicles generated from the endosomal system instead of outward budding, typically ranging in size from 30 to 150 nm. Exosomes contain a large number of active constituents (e.g., proteins, lipids, and nuclear acids) and are considered key mediators of intercellular transportation 3 , 4 . Exosomes can be secreted by multiple kinds of donor cells, among which tumor cell-derived exosomes (TDEs) attract the most interest since they are involved in a series of critical functions, such as tumor growth and metastasis 5 , 6 . Accelerating evidence suggests that ubiquitous TDEs in the tumor microenvironment (TME) play critical roles in tumor progression (Fig. 1 ). Transferring biological information locally and distantly, TDEs regulate the fate of their target cells through autocrine and paracrine pathways 7 , 8 . TDEs communicate with tumor cells, immune cells, cancer-associated fibroblasts (CAFs), and host vasculature in the TME and from a distance 9 , 10 . It has been indicated that TDEs can be taken up by tumor cells and inhibit the further release of TDEs, forming a negative feedback loop regulation. Moreover, TDEs also modulate the function of recipient tumor cells in proliferation and metastasis 11 , 12 . Immune cells are critical antitumor effectors in the TME 13 , 14 . Suppressing T-cell proliferation and inhibiting CD8 T-cell activation, TDEs contribute to immune escape 15 , 16 . In addition, TDEs may also induce T-cell differentiation into a suppressive regulatory T-cell (Treg) phenotype, favoring immunosuppression 17 – 19 . Producing immunoglobulins and presenting antigens, B cells play critical roles in cancer immunity. Regulatory B cells (Bregs) can be induced by TDEs, facilitating immune tolerance 20 , 21 . In addition, TDEs also present suppressive effects on natural killer cells 22 , 23 , dendritic cells 24 – 26 , and macrophages 27 – 29 in the TME, promoting tumor progression and benefiting immune escape 30 . Fig. 1 Tumor-derived exosomes (TDEs) play critical roles in tumor progression and impede tumor therapy. TDEs promote the expansion and activation of regulatory T cells and B cells and inhibit the function of effector T cells and NK cells to create immunosuppressive tumor microenvironments. By not only reprogramming normal fibroblasts into CAFs but also inducing CAFs to transform epithelial cells to a mesenchymal phenotype, TDEs also induce tumor progression and metastasis. TDEs mediate the activation of endothelial cells, leading to tumor angiogenesis. For tumor therapy, TDEs mediate chemotherapeutic drug resistance phenotypes in tumor cells by delivering multidrug transporters. In addition, TDE impedes the effect of tumor immunotherapy through the contained immune checkpoints, especially PD-L1. CAFs constitute the major cellular component in solid tumors. Distinct from normal fibroblasts, CAFs are able to secrete multiple proinflammatory factors that contribute to tumor growth and metastasis 31 , 32 . Tumor cells communicate with CAFs through TDEs. TDEs may reprogram normal fibroblasts into CAFs, mainly through the delivery of critical proteins and microRNAs (miRNAs) to activate the transforming growth factor-β (TGF-β) signaling pathway 33 – 35 . In addition, TDEs could also induce CAFs to transform epithelial cells to a mesenchymal phenotype, promoting tumor metastasis 36 . TDEs accelerate angiogenesis in the TME to construct a new blood vessel network for tumor progression 37 , 38 . Containing vascular endothelial growth factor (VEGF) and other critical modulators in angiogenesis and targeting endothelial cells (ECs), TDEs reprogram ECs and activate angiogenic signaling pathways, inducing neovascularization in the TME 39 – 41 . In addition, ligand/receptor-mediated interactions are also involved in TDE-induced angiogenesis of ECs 42 . In addition to the critical roles of TDEs in tumor progression, TDEs also contribute to drug resistance and impede the effective response to antitumor immunotherapy 43 , 44 . Accelerating evidence indicates that TDEs contain a large number of nucleic acids that may transition drug-sensitive cancer cells to a resistant phenotype 45 , 46 . In addition, TDEs deliver P-glycoprotein (P-gp), an ATP-dependent multidrug transporter, in an autocrine way to induce the extrusion of cytotoxic drugs 47 . Acquired drug resistance could also be achieved by the transmission of proto-oncogenes such as PTPRZ1-MET 48 . Checkpoint blockade immunotherapies targeting the programmed cell death-1 receptor (PD-1)/programmed cell death 1 ligand 1 (PD-L1) axis have emerged as promising treatments for cancer patients 49 . Nevertheless, the response rate is not satisfactory and partially due to the adaptive resistance mechanism mediated by TDEs 50 , 51 . Carrying PD-L1 inherited from their donor cells, TDEs bind to immune cells through the PD-1/PD-L1 axis, leading to the dysfunction of antitumor effectors 52 – 54 . Studies also suggested that PD-L1 on TDEs could directly bind to anti-PD-L1 antibody and may lead to immunotherapy resistance 55 . Due to their involvement in tumor progression and therapy resistance, TDEs need to be inhibited to improve the prognosis of cancer patients. In the present review, we systematically overviewed the current strategies for TDE inhibition and clearance and suggested opportunities for tumor interventions in future translational medicine. Moreover, existing challenges, as well as potential prospects for TDE-targeting cancer therapy, are also discussed to bridge the gaps between TDE inhibition and the promising future of cancer therapy. TDE biogenesis and secretion As an important subset of exosomes, TDEs share common mechanisms of biogenesis with non-TDEs (Fig. 2 ) that begin with the inward membrane budding of early endosome (EE) to form multivesicular bodies (MVBs) 56 . It is well known that the endosomal sorting complex required for transport (ESCRT) machinery pathways are critically involved in the mechanism of TDEs. The ESCRT system consists of four complexes (named ESCRT-0, ESCRT-I, ESCRT-II, ESCRT-III) and accessory components 57 , 58 . The early ESCRTs (ESCRT-0 and ESCRT-I) are more responsible for cargo sorting. Initially, the ESCRT-0 complex binds phosphatidinositol-3-phosphate (PI3P) on the EE. Then, ESCRT-I and ESCRT-II are recruited and aggregated on the endosomal membrane. With high affinities, the rigid polyvalent membrane binding structures of ESCRT-I and ESCRT-II promote early endosomal membrane deformation and inward budding to form a narrow membrane neck. ESCRT-III is subsequently recruited to cleave the neck, forming the intraluminal vesicles (ILVs) of MVB 58 . TDEs are consequently released by the fusion of MVBs with the tumor cell plasma membrane 59 . Studies have reported that most of the key regulators, for example, ESCRT-0 proteins (HRS, STAM1, STAM2) 60 – 62 , ESCRT-I protein TSG101 63 , ESCRT-III proteins (CHMP4A, CHMP4B, CHMP4C) 64 , and accessory protein Alix 65 , are overexpressed and/or hyperactivated in various tumor cells, contributing to the aberrant secretion of TDEs (Fig. 3 ). In this context, although the mechanism is shared, the biogenesis and secretion activity of TDEs are different from those of non-TDEs, owing to aberrantly expressed regulators. Fig. 2 Critical modulators involved in TDE biogenesis and secretion. The biogenesis of TDEs begins with early endosomes budding inward to form the MVB, which depends on ESCRT complexes or through the ESCRT-independent pathway based on nSMase2. Subsequently, MVBs fuse with lysosomes for degradation or are docked to the cell periphery with the favor of Rab GTPases for secretion. Finally, SNARE complexes drive membrane fusion for TDE release. Fig. 3 Exclusive machinery for TDE biogenesis and secretion. Key regulators of exosome biogenesis are overexpressed and/or hyperactivated in various tumor cells, contributing to the aberrant secretion of TDEs. The tumor microenvironment (TME) benefits the biogenesis and secretion of TDEs in multiple manners. Under hypoxic and acidic conditions, the biogenesis and secretion of TDEs are crucially regulated, making them different from non-TDEs. Due to specific genomic mutations (e.g., RAS and EGFR) of tumor cells, there are distinguishing mechanisms underlying the biogenesis and secretion of TDEs compared with non-TDEs. Colombo et al. revealed that ESCRT-0/I (HRS, STAM1, and TSG101) knockdown critically inhibited TDE secretion by HeLa-CIITA cells 66 . In addition to abundant numbers, the function of TDEs strongly depends on the encapsulated critical protein cargos, such as PD-L1 55 . In this context, fully revealing the mechanism of cargo sorting would provide new insights for TDE inhibition. ESCRT-0 is generally considered the driver responsible for sorting ubiquitinated protein cargo into ILVs. Bache et al. revealed that HRS, STAM proteins and EPS15 formed a multivalent complex, which had multiple ubiquitin-binding domains to cluster ubiquitin cargos on the endosomal membrane 67 . Deletion of certain ESCRT subunits leads to changes in the protein composition of TDEs. In our previous study, when HRS was downregulated in malignant melanoma cells, a decrease in the level of exosomal PD-L1, as well as an increase in cellular PD-L1, was achieved, suggesting HRS as a potential target for functional cargo sorting of TDEs 52 . Further elucidation of critical TDE cargo sorting regulated by ESCRT-0 would be beneficial for functional TDE inhibition. Baietti et al. indicated that the Syndecan-Syntenin-Alix axis was crucial for TDE biogenesis and cargo sorting manipulation in breast cancer cells 68 . Syndecan is a ubiquitinated transmembrane protein that is supplied with heparan sulfate on the membrane. Alix binds to syntenin, the cytoplasmic adaptor of syndecan, regulating the biogenesis of ILVs. Recent evidence demonstrated that Alix also recruited ESCRT-III proteins to regulate TDE biogenesis and specific cargo sorting, such as tetraspanins, independent of other ESCRT subunits 69 , 70 . Being able to bind both ubiquitinated and nonubiquitinated proteins, Alix may be considered a potential target for TDE inhibition. However, Monypenny et al. indicated that Alix depletion resulted in defective PD-L1 trafficking through MVBs. Loss of Alix promoted PD-L1 redistribution to the cell surface and conferred an EGFR-dependent immunosuppressive phenotype 71 . Taken together, future studies should take combinations between TDE inhibition and the related effects into consideration to achieve effective TDE-targeted cancer therapy. It has also been found that the lipid raft microdomains distinctly segregated protein cargos on the endosome membrane 72 , 73 . Neutral sphingomyelinase 2 (nSMase2) is able to hydrolyze the sphingomyelin of the endosome membrane into conical ceramide, resulting in negative membrane curvature by its cone-shaped structure. Subsequently, the endosome membrane bends toward the inner cavity and sorts the lipids and protein cargos into the ILVs 73 . When nSMase2 was knocked out in the PC3 prostate cancer cell line, the secretion of TDEs was critically inhibited 53 . Before the release of TDEs, secretory MVB is transported to the cell periphery and docked to the plasma membrane. The Rab GTPase family contributes to the underlying mechanism of this trandport 74 , 75 . It has been noted that the 2 isoforms of Rab27, i.e., Rab27a and Rab27b, are involved in membrane transport with distinguished roles. Specifically, Rab27a regulates the docking and fusion of MVBs with the plasma membrane, while Rab27b participates in membrane transfer to MVEs from the Golgi network (TGN) 76 . Rab27a or Rab27b knockdown reduces TDE secretion of various types of cancer cells, such as HeLa cervical cancer cells 76 and T24 bladder cancer cells 77 . Rab7 has also been found to play a critical role in TDE secretion by MCF-7 human breast cancer cells 68 . Other Rab GTPases, such as Rab11 and Rab35, participate in exosome formation of human retinal pigmented epithelial 1 (RPE1) cells 78 , also indicating a potential role in TDE biogenesis. After MVB docking, the soluble N-ethylmaleimide-sensitive component attachment protein receptor (SNARE) complexes start to drive membrane fusion and subsequent TDE secretion. The SNARE complexes are formed by v-SNAREs on the secretory MVB and t-SNAREs on the plasma membrane 79 – 81 . It has been indicated that syntaxin 6 81 , YKT6 82 , and VAMP7 83 , all SNARE proteins, regulate the TDE secretion of C4-2B and CWR-R1 prostate cancer cells, A549 lung cancer cells and K562 leukemic cells, respectively. In addition to universal regulation, the exclusive TME may also benefit the biogenesis and secretion of TDEs in multiple manners (Fig. 3 ). Hypoxia and low pH are the key features of the TME 84 . Under hypoxic and acidic conditions, the biogenesis and secretion of TDEs are crucially regulated, making them different from non-TDEs 85 , 86 . Mechanistically, hypoxia-inducible factor-1alpha (HIF-1α) promoted the expression of pyruvate kinase 2 (PKM2) 87 . PKM2 was responsible for the phosphorylation of Ser95 of synaptosome-associated protein 23 (SNAP-23), a critical component of the SNARE complex, thereby promoting the secretion of TDEs 88 . Low pH is also considered a key microenvironmental factor that regulates the biogenesis and secretion of TDEs. It has been elucidated that acidic conditions could promote the secretion of TDEs by enhancing the function of caveolin-1 86 , which is an important initiator of exosome biogenesis through the regulation of cholesterol contents. Due to the specific genomic mutations of tumor cells, there are distinguishing mechanisms underlying the biogenesis and secretion of TDEs compared with non-TDEs (Fig. 3 ). RAS (H-RAS, N-RAS, and K-RAS) is the most frequently mutated oncogene in cancers 89 . Several key differences have been revealed between mutant RAS and wild-type RAS cell line-derived exosomes. Increased secretion levels and more oncogenic proteins were found in mutant K-RAS cell-derived exosomes 90 . Mechanistically, activation of the RAS/ERK pathway, as a result of RAS mutation, was associated with the ESCRT-dependent biogenesis and secretion of exosomes. In addition, mutation of epidermal growth factor receptor (EGFR), which is considered one of the characteristics of lung cancer, also leads to the increased biogenesis and secretion of TDEs 91 . Collectively, the aberrant levels of shared regulators, the exclusive TME and the specific genomic mutations may combine to contribute to the promoted secretion and altered molecular contents of TDEs. Inhibition strategies for TDEs Based on the mechanisms of TDE biogenesis and release, emerging strategies are employed for targeted inhibition of TDEs (Fig. 4 and Table 1 ). Genetic manipulation and pharmacological inhibitors are the most studied approaches. Fig. 4 Emerging strategies employed for TDE targeted inhibition. Genetic manipulation has been proven to inhibit TDEs. With RNAi and the CRISPR‒Cas9 system to disrupt the genes that regulate TDE biogenesis and secretion, TDE inhibition was achieved. Pharmacological inhibitors have also been adopted to block TDEs by targeting critical regulators in TDE secretion. Table 1 Inhibition strategies for TDEs. Strategies Targets Cancer cell types/Recipients References Genetic manipulation RNAi HRS HeLa-CIITA cells, SCC61 cells, WM9 cells 52 , 66 , 93 STAM1, TSG101 HeLa-CIITA cells 66 Syndecan, Syntenin, Alix, MCF-7 cells 68 ARF6, PLD2 MCF-7 cells 94 Rab27a/Rab27b HeLa cells, T24 cells, SCC61 cells 76 , 77 , 93 , 95 YKT6 A549 cells 78 Gene knockout Rab27a and nSMase2 PC3 cells 53 Pharmacological inhibition GW4869 nSMase MDA-MB-231 cells, T24 cells, A431 cells, SCC61 77 , 93 , 98 , 99 Manumycin-A Tipifarnib FTases C4-2B cells 105 106 Ketoconazole ERK signaling RCC-24, 786-O, Caki-2 107 Erlotinib EGFR HSC‐4 cells 108 HS analogs Syndecan-Syntenin-Alix B16F10 cells 109 Sulfisoxazole ETA MDA-MB-231 and CT26 cells 110 , 111 Genetic manipulation Effective biotechnological tools, such as RNA interference (RNAi) and the CRISPR‒Cas9 system, have been widely used to reduce or completely disrupt the expression of critical genes involved in TDE biogenesis and secretion 92 . Colombo et al. took advantage of RNAi to silence 23 components of ESCRT machinery in HeLa-CIITA cells 66 . It was further indicated that the silencing of HRS, STAM1, or TSG101 led to a reduction in TDE secretion and exosomal MHC class II (MHC II). HRS has been shown to be involved in TDE biogenesis and cargo sorting in several studies. Hoshino et al. used RNAi to knock down HRS in SCC61 head and neck squamous cell carcinoma (HNSCC) cells and found that the secretion of exosomes was significantly reduced, as well as the protein contents, such as TSG101 and MT1-MMP, the critical matrix-degrading proteinase 93 . In our previous study, exosomal PD-L1 from malignant melanoma cells was found to contribute to immunosuppression. Encouragingly, when the expression of HRS was disrupted by RNAi, the level of exosomal PD-L1 decreased significantly 52 . Taken together, the above evidence suggests that HRS may be targeted for both TDE and effective cargo inhibition. In MCF-7 human breast cancer cells, a portion of exosomes are formed through the Syndecan-Syntenin-Alix pathway. The disruption of Syndecan, Syntenin, or Alix by RNAi led to the reduced secretion of exosomes 68 . In another study, the biogenesis of syntenin-Alix exosomes was found to be regulated by the small GTPase ARF6 and its effector PLD2. Downregulation of ARF6 or PLD2 affected ILV formation and caused defects in MVB formation and subsequent exosome secretion 94 . In that case, ARF6 depletion did not implicate the specific cargo sorting of exosomes, which led to an aimless inhibition of both TDEs and non-TDEs. Rab GTPases, which are responsible for intracellular vesicle transport, have also been used as targets for TDE inhibition. Evidence indicated that knockdown of Rab27a or Rab27b by RNAi inhibited the secretion of exosomes from a variety of tumor cells, such as cervical cancer cells 76 , bladder cancer cells 77 , HNSCCs 93 , 95 , etc. The downregulation of Rab27a or Rab27b blocked the transport of MVB to the cell periphery as well as the docking to the cell membrane, resulting in TDE inhibition 76 . Notably, Poggio et al. achieved the inhibition of exosomal secretion of PC3 prostate cancer cells by knockout of nSMase2 or Rab27a based on CRISPR‒Cas9, indicating a potential method of TDE inhibition 53 . However, since the deletion of nSMase2 resulted in only a partial loss of TDEs, more effective strategies or targets should be revealed in future studies. In addition, RNAi targeting Rab7 in MCF-7 breast cancer cells also inhibited the transport of MVB and led to reduced TDE secretion 68 . A recent study presented that Rab31 regulated exosome biogenesis in HeLa cells through an ESCRT-independent pathway and drove the sorting of protein cargos such as EGFR, providing a novel target for TDEs as well as cargo inhibition in future therapeutics 96 . For SNARE proteins that mediate the membrane fusion of MVB and cell membrane, Peak et al. found that downregulation of t-SNARE protein syntaxin 6 resulted in significantly reduced TDE secretion in C4-2B and CWR-R1 prostate cancer cells 81 . Ruiz-Martinez et al. also revealed inhibitory effects on TDE secretion in A549 cells by targeting YKT6 82 . Pharmacological inhibition Over recent decades, pharmacological inhibitors have been extensively studied for suppressing TDEs, providing broad prospects in therapeutic applications. GW4869 is considered the most commonly used inhibitor of TDEs 97 . In both in vitro and in vivo cases, GW4869 inhibited the secretion of exosomes from a variety of tumor cells, including breast cancer cells 98 , bladder cancer cells 77 , epidermal cancer cells 99 , head and neck squamous cell carcinoma cells 93 , and malignant melanoma cells 100 , facilitating antitumor immunity. Yang et al. indicated that GW4869 inhibited TDEs from MDA-MB-231 breast cancer cells, resulting in a decrease in total protein content 98 . In a mouse model of breast cancer, GW4869 arrested tumor growth by inhibiting TDEs to promote antitumor effects, significantly enhancing the therapeutic effect of an anti-PD-L1 antibody. Given the hydrophobic nature of the GW4869 molecule, Wang et al. constructed a hyaluronic acid (HA)-based nanoplatform (HGF NPs) to deliver GW4869 in vivo 101 . The secretion of TDEs was dramatically decreased by HGFs assembled from GW4869 with a ferroptosis inducer (Fe 3+ ), leading to the antitumor effect of cytotoxic T lymphocytes as well as long-lasting immunological memory. Although inhibiting roles on TDEs have been reported in numerous cases, the practical application of GW4869 is limited due to certain shortcomings. GW4869 is a blocker of nSMase2, which mediates the biogenesis and secretion of both TDEs and non-TDEs via an ESCRT-independent way 102 . Thus, direct application of GW4896 without tumor-targeted delivery may lead to nonspecific inhibition of both TDEs and non-TDEs. Additionally, considering that both ESCRT-dependent and ESCRT-independent machineries are involved in the regulation of TDE biogenesis and secretion 103 , application of GW4869 to inhibit the nSMase-mediated ESCRT-independent pathway may achieve only limited effects. In addition, GW4869 abrogates nSMase in a noncompetitive way 102 , which may further lead to limited efficiency of TDE inhibition. Finally, biosafety assessment of GW4869 is required before its clinical application because nSMase2 also contributes to multiple central biological processes 104 , thereby resulting in unpleasant side effects in vivo. Based on the mechanisms that account for the biogenesis and secretion of TDEs, targeted inhibition strategies of TDEs were revealed. The genomic mutations that contribute to the aberrant biogenesis and secretion of TDEs could provide specific therapeutic targets. In this regard, Datta et al. conducted high-throughput drug screening and found a natural bacterial metabolite, manumycin-A (MA), which selectively affected RAS/RAF/ERK1/2 by targeting farnesyltransferases (FTases), resulting in the inhibited secretion of TDEs by castration-resistant prostate cancer (CRPC) cells 105 . In addition, tipifarnib, another FTase inhibitor, was also found to inhibit TDEs of prostate cancer cells, suggesting that FTase inhibitors can function as a class of potential effectors to block TDEs 106 . In a recent study, ketoconazole was adopted as an adjunctive therapy to enhance the efficacy of sunitinib in renal cell carcinoma treatment by inhibiting TDEs through downstream ERK signaling, providing updated evidence for the use TDE inhibitors as a novel option for tumor therapy 107 . Sasabe et al. employed erlotinib, an EGFR inhibitor, to suppress the negative effects of TDEs in oral squamous cell carcinoma (OSCC) 108 . This finding suggested that anti-EGFR agents may be effective for the treatment of cancer patients with EGFR mutations by not only blocking the EGFR signaling pathway but also attenuating the unpleasant roles of TDEs. In addition to the specific inhibition of genomic mutations or exclusive TME, a shared regulatory mechanism could also be employed. Since most of the key regulators were found to be overexpressed and/or hyperactivated in tumor cells, they may serve as potential candidates for TDE inhibition. Wu et al. indicated that heparan sulfate (HS) analogs (heparin, low molecular weight heparin, and 6-O-desulfated heparin) specifically and efficiently inhibited TDE secretion by targeting Syndecan-Syntenin-Alix, leading to weakened tumor proliferation and invasion 109 . When B16F10 melanoma cells were treated with different HS analogs, both TDE secretion and protein cargo were inhibited. By screening 1163 drugs from FDA-approved libraries, Im et al. revealed that sulfisoxazole (SFX) selectively inhibited the secretion of TDEs from breast cancer cells 110 . SFX is generally employed as an oral antibiotic that is noncytotoxic at effective doses. By suppressing the transcription of Rab GTPases (Rab5, Rab7, and Rab27a) and ESCRT components (Alix, VPS4B), SFX inhibited the formation and secretion of MVB and induced their degradation within lysosomes, ultimately leading to TDE inhibition. In breast cancer xenograft mouse models, SFX presented significant antitumor and antimetastatic effects by inhibiting TDEs. Endothelin receptor A (ETA) was identified as the downstream effector of SFX, providing a potential target for TDE inhibition in breast cancer cells. In a recent study, SFX was found to be effective in reducing the level of circulating exosomes carrying PD-L1 in CT26 tumor-bearing mice, which reinvigorated the function of CD8 cytotoxic T cells and enhanced the efficacy of anti-PD-1 immunotherapy 111 . However, the specificity is limited since non-TDEs would also be inhibited, leading us to focus on tumor-targeted delivery of the drugs for improvement. Based on the key features of the TME (e.g., hypoxia and low pH), hypoxia- and/or pH-responsive drug delivery systems may be developed for targeted drug delivery. Taking lessons from well-designed TME-responsive systems, pharmacological inhibitors may be encapsulated and specifically delivered to tumor tissues, thereby achieving selective inhibition of TDEs. Genetic manipulation Effective biotechnological tools, such as RNA interference (RNAi) and the CRISPR‒Cas9 system, have been widely used to reduce or completely disrupt the expression of critical genes involved in TDE biogenesis and secretion 92 . Colombo et al. took advantage of RNAi to silence 23 components of ESCRT machinery in HeLa-CIITA cells 66 . It was further indicated that the silencing of HRS, STAM1, or TSG101 led to a reduction in TDE secretion and exosomal MHC class II (MHC II). HRS has been shown to be involved in TDE biogenesis and cargo sorting in several studies. Hoshino et al. used RNAi to knock down HRS in SCC61 head and neck squamous cell carcinoma (HNSCC) cells and found that the secretion of exosomes was significantly reduced, as well as the protein contents, such as TSG101 and MT1-MMP, the critical matrix-degrading proteinase 93 . In our previous study, exosomal PD-L1 from malignant melanoma cells was found to contribute to immunosuppression. Encouragingly, when the expression of HRS was disrupted by RNAi, the level of exosomal PD-L1 decreased significantly 52 . Taken together, the above evidence suggests that HRS may be targeted for both TDE and effective cargo inhibition. In MCF-7 human breast cancer cells, a portion of exosomes are formed through the Syndecan-Syntenin-Alix pathway. The disruption of Syndecan, Syntenin, or Alix by RNAi led to the reduced secretion of exosomes 68 . In another study, the biogenesis of syntenin-Alix exosomes was found to be regulated by the small GTPase ARF6 and its effector PLD2. Downregulation of ARF6 or PLD2 affected ILV formation and caused defects in MVB formation and subsequent exosome secretion 94 . In that case, ARF6 depletion did not implicate the specific cargo sorting of exosomes, which led to an aimless inhibition of both TDEs and non-TDEs. Rab GTPases, which are responsible for intracellular vesicle transport, have also been used as targets for TDE inhibition. Evidence indicated that knockdown of Rab27a or Rab27b by RNAi inhibited the secretion of exosomes from a variety of tumor cells, such as cervical cancer cells 76 , bladder cancer cells 77 , HNSCCs 93 , 95 , etc. The downregulation of Rab27a or Rab27b blocked the transport of MVB to the cell periphery as well as the docking to the cell membrane, resulting in TDE inhibition 76 . Notably, Poggio et al. achieved the inhibition of exosomal secretion of PC3 prostate cancer cells by knockout of nSMase2 or Rab27a based on CRISPR‒Cas9, indicating a potential method of TDE inhibition 53 . However, since the deletion of nSMase2 resulted in only a partial loss of TDEs, more effective strategies or targets should be revealed in future studies. In addition, RNAi targeting Rab7 in MCF-7 breast cancer cells also inhibited the transport of MVB and led to reduced TDE secretion 68 . A recent study presented that Rab31 regulated exosome biogenesis in HeLa cells through an ESCRT-independent pathway and drove the sorting of protein cargos such as EGFR, providing a novel target for TDEs as well as cargo inhibition in future therapeutics 96 . For SNARE proteins that mediate the membrane fusion of MVB and cell membrane, Peak et al. found that downregulation of t-SNARE protein syntaxin 6 resulted in significantly reduced TDE secretion in C4-2B and CWR-R1 prostate cancer cells 81 . Ruiz-Martinez et al. also revealed inhibitory effects on TDE secretion in A549 cells by targeting YKT6 82 . Pharmacological inhibition Over recent decades, pharmacological inhibitors have been extensively studied for suppressing TDEs, providing broad prospects in therapeutic applications. GW4869 is considered the most commonly used inhibitor of TDEs 97 . In both in vitro and in vivo cases, GW4869 inhibited the secretion of exosomes from a variety of tumor cells, including breast cancer cells 98 , bladder cancer cells 77 , epidermal cancer cells 99 , head and neck squamous cell carcinoma cells 93 , and malignant melanoma cells 100 , facilitating antitumor immunity. Yang et al. indicated that GW4869 inhibited TDEs from MDA-MB-231 breast cancer cells, resulting in a decrease in total protein content 98 . In a mouse model of breast cancer, GW4869 arrested tumor growth by inhibiting TDEs to promote antitumor effects, significantly enhancing the therapeutic effect of an anti-PD-L1 antibody. Given the hydrophobic nature of the GW4869 molecule, Wang et al. constructed a hyaluronic acid (HA)-based nanoplatform (HGF NPs) to deliver GW4869 in vivo 101 . The secretion of TDEs was dramatically decreased by HGFs assembled from GW4869 with a ferroptosis inducer (Fe 3+ ), leading to the antitumor effect of cytotoxic T lymphocytes as well as long-lasting immunological memory. Although inhibiting roles on TDEs have been reported in numerous cases, the practical application of GW4869 is limited due to certain shortcomings. GW4869 is a blocker of nSMase2, which mediates the biogenesis and secretion of both TDEs and non-TDEs via an ESCRT-independent way 102 . Thus, direct application of GW4896 without tumor-targeted delivery may lead to nonspecific inhibition of both TDEs and non-TDEs. Additionally, considering that both ESCRT-dependent and ESCRT-independent machineries are involved in the regulation of TDE biogenesis and secretion 103 , application of GW4869 to inhibit the nSMase-mediated ESCRT-independent pathway may achieve only limited effects. In addition, GW4869 abrogates nSMase in a noncompetitive way 102 , which may further lead to limited efficiency of TDE inhibition. Finally, biosafety assessment of GW4869 is required before its clinical application because nSMase2 also contributes to multiple central biological processes 104 , thereby resulting in unpleasant side effects in vivo. Based on the mechanisms that account for the biogenesis and secretion of TDEs, targeted inhibition strategies of TDEs were revealed. The genomic mutations that contribute to the aberrant biogenesis and secretion of TDEs could provide specific therapeutic targets. In this regard, Datta et al. conducted high-throughput drug screening and found a natural bacterial metabolite, manumycin-A (MA), which selectively affected RAS/RAF/ERK1/2 by targeting farnesyltransferases (FTases), resulting in the inhibited secretion of TDEs by castration-resistant prostate cancer (CRPC) cells 105 . In addition, tipifarnib, another FTase inhibitor, was also found to inhibit TDEs of prostate cancer cells, suggesting that FTase inhibitors can function as a class of potential effectors to block TDEs 106 . In a recent study, ketoconazole was adopted as an adjunctive therapy to enhance the efficacy of sunitinib in renal cell carcinoma treatment by inhibiting TDEs through downstream ERK signaling, providing updated evidence for the use TDE inhibitors as a novel option for tumor therapy 107 . Sasabe et al. employed erlotinib, an EGFR inhibitor, to suppress the negative effects of TDEs in oral squamous cell carcinoma (OSCC) 108 . This finding suggested that anti-EGFR agents may be effective for the treatment of cancer patients with EGFR mutations by not only blocking the EGFR signaling pathway but also attenuating the unpleasant roles of TDEs. In addition to the specific inhibition of genomic mutations or exclusive TME, a shared regulatory mechanism could also be employed. Since most of the key regulators were found to be overexpressed and/or hyperactivated in tumor cells, they may serve as potential candidates for TDE inhibition. Wu et al. indicated that heparan sulfate (HS) analogs (heparin, low molecular weight heparin, and 6-O-desulfated heparin) specifically and efficiently inhibited TDE secretion by targeting Syndecan-Syntenin-Alix, leading to weakened tumor proliferation and invasion 109 . When B16F10 melanoma cells were treated with different HS analogs, both TDE secretion and protein cargo were inhibited. By screening 1163 drugs from FDA-approved libraries, Im et al. revealed that sulfisoxazole (SFX) selectively inhibited the secretion of TDEs from breast cancer cells 110 . SFX is generally employed as an oral antibiotic that is noncytotoxic at effective doses. By suppressing the transcription of Rab GTPases (Rab5, Rab7, and Rab27a) and ESCRT components (Alix, VPS4B), SFX inhibited the formation and secretion of MVB and induced their degradation within lysosomes, ultimately leading to TDE inhibition. In breast cancer xenograft mouse models, SFX presented significant antitumor and antimetastatic effects by inhibiting TDEs. Endothelin receptor A (ETA) was identified as the downstream effector of SFX, providing a potential target for TDE inhibition in breast cancer cells. In a recent study, SFX was found to be effective in reducing the level of circulating exosomes carrying PD-L1 in CT26 tumor-bearing mice, which reinvigorated the function of CD8 cytotoxic T cells and enhanced the efficacy of anti-PD-1 immunotherapy 111 . However, the specificity is limited since non-TDEs would also be inhibited, leading us to focus on tumor-targeted delivery of the drugs for improvement. Based on the key features of the TME (e.g., hypoxia and low pH), hypoxia- and/or pH-responsive drug delivery systems may be developed for targeted drug delivery. Taking lessons from well-designed TME-responsive systems, pharmacological inhibitors may be encapsulated and specifically delivered to tumor tissues, thereby achieving selective inhibition of TDEs. Clearance strategies for TDEs In addition to the strategy based on genetic manipulation or pharmaceutical inhibition of TDEs in vitro and in vivo, Orme et al. pioneered the removal of TDEs from the circulation through therapeutic plasma exchange (TPE) in patients with malignant melanoma 112 . To discard the circulating exosomes, the plasma from patients was extracted by apheresis equipment and replaced by colloid solutions. With the diminished level of circulating TDEs, especially the critical cargos that modulate immunosuppression, the efficacy of immunotherapy may be improved. Dialysis is also a commonly employed treatment to remove harmful substances from the circulatory system. As widely adopted therapies for kidney diseases, hemofiltration (HF) and hemoperfusion (HP) have also been developed to treat cancer. Taking advantage of a semipermeable membrane (diameter  200 nm) microporous membrane would be promising for the clearance of TDEs. Although not yet applied in the clinic, affinity adsorbents have been adopted to selectively remove immunosuppressive cytokines, which may be a promising strategy for cancer treatment. It has been revealed that the cytokine network of the TME is involved in tumor progression and metastasis, leading Wang et al. further to utilize polyvinyl alcohol (PVA) microspheres coupled with heparin to remove tumor-induced cytokines 113 . The developed approach efficiently adsorbed immunosuppressive cytokines, such as VEGF and TGF-β, in the blood of tumor patients, facilitating cancer therapy. It is worth noting that PVA also presented promising biosafety, making it suitable for hemoperfusion in future translational medicine. Wu et al. also applied silica microspheres with a hemofiltration device to achieve the selective capture and removal of abundant circulating tumor cells as well as TDEs, providing potential choices for tumor therapy 114 . Recently, the rapid development of microfluid chips has paved the way for TDE filtration. Benefitting from the powerful compatibility and tiny size, microfluid chips assemble a large number of units with antibody coating for TDE elimination, which enables rapid and straightforward TDE clearance 115 . Current challenges and future perspectives As outlined, approaches have been developed for the effective inhibition and clearance of TDEs. However, compared with the growing demand for scientific research and clinical application, more steps are needed (Fig. 5 ). Fig. 5 Current challenges and future perspectives for future eligible TDE inhibition and clearance to facilitate tumor therapeutics. Off-target effects and biosafety risks are the leading factors that limit the application of gene manipulation for TDE inhibition. Efforts should be devoted to developing an efficient and safe delivery system as well as targeting TDE-specific effectors. For promising pharmaceutical TDE inhibition, considerations on specific TDE blocking as well as controllable side effects should be taken ahead of translational medicine. The combination of high-throughput drug screening strategies and targeted drug delivery systems would promote the development of clinically available pharmacological inhibitor-based tumor therapies. For direct TDE clearance, the risk of trauma and infection as well as nonspecific elimination hamper the clinical translation to a large extent. In this regard, more friendly and specific strategies should be developed to pave the way for future clinics. Despite its promising application in cancer research, TDE inhibition based on genetic manipulation still faces challenges in certain aspects 116 , 117 . First, off-target effects would result in unintended gene deletions, insertions, or mutations, leading to reduced block effect and inducing safety concerns 118 . Second, the specificity of the current inhibition strategy is limited. By blocking both TDEs and non-TDEs, potential adverse effects in tumor treatment were induced. Third, genetic modification may cause biosafety problems with virus-based systems, such as adeno-associated viral vectors or lentiviral vectors, resulting in uncertain virus‒host interactions such as severe immune reactions and cancer progression 119 – 121 . To obtain eligible TDE inhibition with genetic manipulation, efforts should be devoted to developing an efficient and safe delivery system as well as targeting TDE-specific effectors 122 – 124 . To this end, the mechanism of TDE biogenesis and secretion should be well recognized, providing more potential targets for future translation. With the understanding of the detailed mechanism of TDE biogenesis and secretion, multitargeting strategies may be developed to inhibit every critical step in TDE generation, leading to a whole pathway of TDE inhibition 53 , 125 . In addition, although the safety concerns of viral vectors have been dispelled in most in vitro and in vivo studies, considerations should be addressed on more alternatives in tumor therapy 126 . For TDE inhibitors, serious considerations should also be taken ahead of translational medicine. First, it is difficult to target a single molecule with an inhibitor to effectively block TDEs since complicated pathways are involved in TDE biogenesis and secretion. On the other hand, with a heterogeneous population of exosomes in circulation, including but not limited to TDEs, obstacles also exist in specific inhibition of TDEs with the current strategies 2 , 127 . In these cases, multitarget pharmacological inhibitors should be developed to block as many pathways as possible that are critical for TDE biogenesis and secretion. Moreover, to avoid potential side effects to non-TDEs, precise release of TDE inhibitors should also be achieved. Considering that the exclusive TME (e.g., hypoxia and low pH) benefits TDE biogenesis and secretion, reshaping the TME or targeting the related downstream signaling pathways would provide new opportunities for precise TDE inhibition. It should also be noted that the development of clinically available pharmacological inhibitors is a time-consuming work with substantial costs 128 , 129 . In this regard, it would be more effective to screen TDE inhibitors with a high-throughput system. The quantitative analytical methodology should also be constructed for simultaneous screening with a wide range of candidate inhibitors 105 , 106 , 130 . For direct TDE clearance, current strategies based on extracorporeal devices present limitations, such as invasion-induced trauma, bleeding risks and potential infections, hampering clinical translation. Moreover, the removal of total exosomes by blood purification would not only eliminate TDEs but also clear up the potential positive exosomes, leading to unknown effects that may further burden the tumor patients. In this case, more friendly and specific clearance strategies should be developed in future translation 112 , 131 . To this end, activating the in vivo phagocytosis system mediated by macrophages instead of developing a clearance system in vitro would be a beneficial approach 132 . Evidence has indicated that phagocytotic clearance by macrophages might be altered in the TME. The elevated PD-1 expression on macrophages in the TME was negatively correlated with phagocytic potency 133 . In this case, inhibiting or blocking PD-1 in macrophages in the TME might be an effective approach for activating phagocytosis-mediated clearance of TDEs. Furthermore, Lu et al. recently demonstrated that head and neck squamous cell carcinoma (HNSCC)-derived TDEs inhibited phagocytosis of macrophages through CD73, thereby triggering immune suppression and aggressive tumor growth 134 . Therefore, future studies may also attempt to block CD73 on TDEs to enhance the clearance of TDEs by phagocytes, thus facilitating tumor therapy. Conclusions To conclude, TDEs play critical roles in tumor progression and mediate therapy resistance, leading to poor clinical outcomes. Taking lessons from the mechanism of TDE biogenesis and secretion, we discussed the emerging strategies for TDE inhibition and clearance, providing opportunities for future cancer therapy. To address the issues occurring in current inhibition approaches, considerations should be taken into account to achieve more specific and effective methods for genetic manipulation as well as pharmacological inhibition. Alternatively, friendly TDE clearance strategies should also be introduced. Taken together, targeted inhibition or clearance of TDEs may provide novel therapeutic options for future cancer treatment.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3408201/

Autolytic Proteolysis within the Function to Find Domain (FIIND) Is Required for NLRP1 Inflammasome Activity *

Background: NLRP1 mediates the release of the inflammatory cytokine IL-1β and is linked to several human inflammatory diseases. Results: Autolytic proteolysis occurs within the C terminus of NLRP1 and is modulated by polymorphisms and alternative mRNA splicing. Conclusion: Autolytic cleavage is a key regulator of the NLRP1 inflammasome and downstream IL-1β production. Significance: Understanding the mechanisms underlying NLRP1 activation is required to develop effective therapeutics.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5997389/

Macrophage P2X4 receptors augment bacterial killing and protect against sepsis

The macrophage is a major phagocytic cell type, and its impaired function is a primary cause of immune paralysis, organ injury, and death in sepsis. An incomplete understanding of the endogenous molecules that regulate macrophage bactericidal activity is a major barrier for developing effective therapies for sepsis. Using an in vitro killing assay, we report here that the endogenous purine ATP augments the killing of sepsis-causing bacteria by macrophages through P2X4 receptors (P2X4Rs). Using newly developed transgenic mice expressing a bioluminescent ATP probe on the cell surface, we found that extracellular ATP levels increase during sepsis, indicating that ATP may contribute to bacterial killing in vivo. Studies with P2X4R-deficient mice subjected to sepsis confirm the role of extracellular ATP acting on P2X4Rs in killing bacteria and protecting against organ injury and death. Results with adoptive transfer of macrophages, myeloid-specific P2X4R-deficient mice, and P2rx4 tdTomato reporter mice indicate that macrophages are essential for the antibacterial, antiinflammatory, and organ protective effects of P2X4Rs in sepsis. Pharmacological targeting of P2X4Rs with the allosteric activator ivermectin protects against bacterial dissemination and mortality in sepsis. We propose that P2X4Rs represent a promising target for drug development to control bacterial growth in sepsis and other infections.

PMC

Anthrax

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9597862/

Structural mechanisms of inflammasome regulation revealed by cryo-EM studies

Inflammasomes are cytosolic protein complexes that form in response to pathogen or damage signals and initiate inflammation. Signal transduction in the inflammasome pathway occurs via protein–protein interaction, protein conformational change, and oligomerization. Recent advances in structural biology have provided multiple insights in inflammasome regulation that are both biologically intriguing and therapeutically valuable. In this review, we summarize the current understanding of three most studied inflammasome complexes: the NAIP/NLRC4, NLRP1, and NLRP3 inflammasomes. We discuss the general mechanisms and unique features of their regulation and how investigating these systems may contribute to therapeutic applications.

PMC