document_id
int64 185
2.68k
| context
stringlengths 2.88k
70.8k
| question
stringlengths 11
194
| id
int64 225
5.32k
| answer
stringlengths 1
933
| answer_start
int64 157
69k
|
|---|---|---|---|---|---|
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What is the word Chikungunya mean? | 2,479 | 'that which contorts or bends up' | 1,855 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What does Chikungunya mean in Swahili? | 2,480 | the illness of the bended walker | 1,921 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | How is CHIKV maintained in Africa? | 2,481 | in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents | 1,998 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What is it vectored by, in Asia? | 2,482 | Ae. aegypti and Ae. albopictus | 2,164 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | How does the transmission in Asia occur? | 2,483 | in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever | 2,231 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What spurred the discovery of the new vector Ae. albopictus? | 2,484 | The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, | 2,406 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | In the epidemic peak how many cases per week were there on the island? | 2,485 | 46,000 | 2,683 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What does this review detail? | 2,486 | the epidemiology and global expansion of CHIKV | 2,992 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What dose this review describe? | 2,487 | its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection. | 3,049 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | How many genotypes of CHIKV have been isilated? | 2,488 | three genotypes based on phylogenetic studies. | 3,229 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What are the genotypes based on? | 2,489 | the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African | 3,303 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | hen did Asian genotype emerge? | 2,490 | between 50 and 310 y ago, | 3,514 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | When didthe Asian genotype diverge from African genotype? | 2,491 | between 100 and 840 y ago | 3,588 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What is the status of Asian CHIKV since its emergence? | 2,492 | has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions | 3,640 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What are the recent activities of CHIKV? | 2,493 | the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. | 3,789 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | How was the Italian isolation found to have evolved from? | 2,494 | from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [ | 4,312 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | How many days is the incubation period? | 2,495 | 2-6 d | 5,443 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | In how many days do the symptoms arise? | 2,496 | 4-7 | 5,481 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What are exhibited in the two phases? | 2,497 | The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [ | 5,673 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What are consequences of infection? | 2,498 | Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. | 6,066 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What percentage of the patients still have the CHIKV IgM after eighteen months? | 2,499 | The chronic stage of CHIKF is characterized by | 6,379 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What is te chronic stage characterized by? | 2,500 | by polyarthralgia that can last from weeks to years beyond the acute stage | 6,422 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What is affected by CHIKV? | 2,501 | fibroblasts | 6,535 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What explains the pain associated with CHIKV? | 2,502 | The high number of nociceptive nerve endings found within the joints and muscle connective tissues | 6,623 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What percentage of people suffering from the CHIKF are over 65 years old? | 2,503 | 50% | 6,784 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What percentage die? | 2,504 | 33% | 6,862 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What other group is disproportionately affected? | 2,505 | children | 7,040 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What complications are associated with CHIKV? | 2,506 | from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure | 7,093 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What happens after host infection? | 2,507 | CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations | 7,347 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What did the Ae.Aegypti which is responsible for epidemics in Kenya, Comoros and Seychelles carry? | 2,508 | CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) | 7,580 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | what was the result of the decline in population of Ae. Aegyptus when the virus struck the Reunion Islands, due to massive use dichlorodiphenyltrichloroethane usage? | 2,509 | in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) | 7,876 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What did this mutation allow? | 2,510 | CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector | 8,003 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What vectored the large epidemic in La Reunion Islands? | 2,511 | Ae. albopictus | 8,182 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What percentage of population was affected? | 2,512 | 34% | 8,247 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | Where were the CHIKV strain found? | 2,513 | All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [ | 8,291 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What is the finding on E1-A226V in Indian Ocean? | 2,514 | mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season | 8,442 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What has the E1-A226V enabled? | 2,515 | an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti | 8,761 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What has become the preferred and lethal vector? | 2,516 | Ae. albopictus | 8,913 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What was the finding on the Green Fluorescent tagged E1-A226V? | 2,517 | E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti | 9,075 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What became the main vector in the Indian Ocean within 1-2 y after CHIKV was introduced? | 2,518 | Ae. albopictus | 9,204 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | how long Ae. aegypti been established in North America? | 2,519 | for over 300 | 9,392 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What is the presence of Ae.albopictus in North America? | 2,520 | has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. | 9,429 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What percentage of newborns were infected from their mother? | 2,521 | 50% | 11,218 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What has been some instances of mother to fetus transmission? | 2,522 | congenital illness and fetal death | 11,449 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What did the studies reveal regarding transmission from mothers during perinatal period? | 2,523 | During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. | 11,492 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What is theorized regarding transmission? | 2,524 | motherto-child transmission most likely happens transplacentally shortly before delivery | 12,096 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What did the study report? | 2,525 | neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers | 12,254 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What is a conclusion of this report? | 2,526 | DNA vaccines could play a major role in combating CHIKV | 24,357 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What is a conclusion of this report? | 2,527 | Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. | 24,421 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What is the NIAID designation of CHIKV? | 2,528 | as a Category C pathogen alongside the influenza and SARS-CoV viruses | 22,921 |
1,689 | Chikungunya: A Potentially Emerging Epidemic?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860491/
SHA: f7c3160bef4169d29e2a8bdd79dd6e9056d4774c
Authors: Thiboutot, Michelle M.; Kannan, Senthil; Kawalekar, Omkar U.; Shedlock, Devon J.; Khan, Amir S.; Sarangan, Gopalsamy; Srikanth, Padma; Weiner, David B.; Muthumani, Karuppiah
Date: 2010-04-27
DOI: 10.1371/journal.pntd.0000623
License: cc-by
Abstract: Chikungunya virus is a mosquito-borne emerging pathogen that has a major health impact in humans and causes fever disease, headache, rash, nausea, vomiting, myalgia, and arthralgia. Indigenous to tropical Africa, recent large outbreaks have been reported in parts of South East Asia and several of its neighboring islands in 2005–07 and in Europe in 2007. Furthermore, positive cases have been confirmed in the United States in travelers returning from known outbreak areas. Currently, there is no vaccine or antiviral treatment. With the threat of an emerging global pandemic, the peculiar problems associated with the more immediate and seasonal epidemics warrant the development of an effective vaccine. In this review, we summarize the evidence supporting these concepts.
Text: Chikungunya virus (CHIKV), a mosquito-borne pathogen listed by National Institute of Allergy and Infectious Diseases (NIAID) as a Category C Priority Pathogen that causes Chikungunya fever (CHIKF), has been spreading throughout Asia, Africa, and parts of Europe in recent times [1, 2, 3] . CHIKV is an arthropod-borne virus (arbovirus) and is transmitted to humans primarily by Aedes aegypti, the infamous yellow fever propagator [4, 5] . CHIKV infection is marked by severe joint pain, contorting its victims into unusual postures [6] . The disease gets its name from the Kimakonde vernacular language of Tanzania and Mozambique, and the word chikungunya means ''that which contorts or bends up'' and translates in Swahili to ''the illness of the bended walker'' [7, 8, 9] . In Africa, CHIKV is maintained in a sylvatic cycle among forest-dwelling Aedes spp. mosquitoes, wild primates, squirrels, birds, and rodents ( Figure 1 ) [10] . In Asia, the disease is vectored by Ae. aegypti and Ae. albopictus [11] . Transmission in Asia occurs in an urban cycle whereby the mosquito spreads the disease from an infected human to an uninfected human, following an epidemiological pattern similar to dengue fever [12] .
The 2005-2006 epidemic of CHIKV in La Reunion islands in the Indian Ocean, spurred the discovery of a new vector species, Ae. albopictus [5] . Wrecking over one-third of the island's population, this epidemic peaked its devastation between January and February 2006, when over 46,000 cases came into light every week, including 284 deaths [5, 13] . Ae. albopictus is common in urban areas of the United States and is already flourishing in 36 states, raising grave concerns to the immunologically naive populace of the United States [14] .
Accordingly, this review elaborately details the epidemiology and global expansion of CHIKV, describes its clinical features and pathogenesis and its symptoms and complications, and finally nominates a possible vaccine approach against CHIKV infection.
CHIKV has been isolated into three genotypes based on phylogenetic studies. These genotypes, based on the gene sequences of an Envelope protein (E1), are Asian, East/Central/ South African, and West African [4, 11, 15] . Using phylogenetic models, Cherian et al. estimate that the Asian genotype of CHIKV emerged between 50 and 310 y ago, and the West and East African genotypes diverged between 100 and 840 y ago [15] . Since then, CHIKV has come a long way, with several mutations incorporated, and has continued to wreak epidemics in several regions. Recent activities of CHIKV include the Indian epidemic in 2005-2006, which was followed by a sudden explosion of cases in 2007. An estimated 1.3 million people across 13 states were reported to be infected in India [12, 16] , and CHIKV was also widespread in Malaysia, Sri Lanka, and Indonesia [17] . In July-August of 2007, CHIKV was reported in Italy, probably brought in by travelers from CHIKV-prone regions of India, Africa, and Indian Ocean islands such as Mauritius, Madagascar, and Seychelles. Few of the Italian isolates were found to have evolved from the Kerala isolate, which was associated with a A226V shift in E1 gene that represents a successful evolutionary adaptation in the mosquito vector similar to the ones observed in Reunion Island [2, 18, 19] .
In recent times, with an increase in global travel, the risk for spreading CHIKV to non-endemic regions has heightened [1] . Several travelers have brought CHIKV home with them after visiting areas with actively infected populations [12, 20] . Such cases have been documented in European countries, Australia, Asia, and the United States [8, 21] . The United States has already reported at least twelve cases of travel-associated CHIKV, while France has reported 850 cases, and the United Kingdom 93 [8, 14] . Beyond this, CHIKV-infected travelers have also been diagnosed in Australia, Belgium, Canada, Czech Republic, French Guiana, Germany, Hong Kong, Italy, Japan, Kenya, Malaysia, Martinique, Norway, Switzerland, and Sri Lanka [21] . Some travelers were viremic, worrying public health officials about the spread of CHIKV to new areas [1, 8] .
The incubation time for CHIKV is relatively short, requiring only 2-6 d with symptoms usually appearing 4-7 d post-infection [22] . Vazeille et al. detected CHIKV in the salivary glands of Ae. albopictus only 2 d after infection [5] . Upon infection, CHIKF tends to present itself in two phases. The first stage is acute, while the second stage, experienced by most but not all, is persistent, causing disabling polyarthritis. Characteristics of the acute phase include an abrupt onset of fever, arthralgia, and in some cases, maculopapular rash [6, 23] . The acute phase causes such intense joint and muscular pain that makes movement very difficult and prostrates its victims [6, 20] .
Ninety-five percent of infected adults are symptomatic after infection, and of these, most become disabled for weeks to months as a result of decreased dexterity, loss of mobility, and delayed reaction. Eighteen months after disease onset, 40% of patients are found to still have anti-CHIKV IgM [6, 18, 23, 24] . The chronic stage of CHIKF is characterized by polyarthralgia that can last from weeks to years beyond the acute stage [6] . CHIKV has been shown to attack fibroblasts, explaining the involvement of muscles, joints, and skin connective tissues. The high number of nociceptive nerve endings found within the joints and muscle connective tissues can explain pain associated with CHIKF [25, 26] .
More than 50% of patients who suffer from severe CHIKF are over 65 y old, and more than 33% of them die. Most adults who suffer from severe CHIKF have underlying medical conditions [6, 24, 27] . The other group that is disproportionately affected by severe CHIKV is children. Other complications associated with CHIKV, from most common to least common, include respiratory failure, cardiovascular decompensation, meningoencephalitis, severe acute hepatitis, severe cutaneous effects, other central nervous system problems, and kidney failure [6, 18, 20, 23, 24, 26, 27] .
CHIKV undertakes a complex replication cycle upon host infection (Figure 2 ), which makes its genome susceptible to mutations [28, 29] . For instance, Ae. aegypti, responsible for epidemics in Kenya, Comoros, and Seychelles, carried CHIKV with an alanine in the 226 position of the E1 gene (E1-A226) [4, 18] . However, when the virus struck La Reunion Islands, a decline in population of Ae. aegypti, due to massive dichlorodiphenyltrichloroethane usage and dearth of Ae. albopictus species' www.plosntds.org population, resulted in an ecological pressure, favoring replacement of alanine at position 226 with valine (E1-A226V) [5] . This mutation allowed CHIKV's secondary vector species, Ae. albopictus, to supplement Ae. aegypti as its primary vector [5] .
Within a year, the E1-A226V mutation was present in La Reunion Island, and Ae. albopictus apparently vectored the large epidemic infecting 34% of La Reunion Island's population [5] . All of the CHIKV strains isolated from Mayotte carried the E1-A226V mutation, and the mutation was also found in Madagascar in 2007 [5] . The E1-A226V mutation was not present at the beginning of the Indian Ocean Islands outbreak (before September 2005). However, more than 90% of later viral strains found there had incorporated the mutation (December-March 2006), indicating a genotype switch during the winter season [5, 18, 20] .
The E1-A226V mutation also enabled an increase in infectivity of Ae. albopictus when compared to its infectivity of Ae. aegypti [4, 11, 18, 30] , and with several factors taken together, Ae. albopictus has become the new preferred and more lethal vector for CHIKV [4, 5, 11] . In fact, Tsetsarkin et al. found that a Green Fluorescent Protein tagged E1-A226V virus was 100 times more infective to Ae. albopictus than it was to Ae. aegypti [4] . In all the Indian Ocean Islands, Ae. albopictus became the main vector for CHIKV within 1-2 y after CHIKV was introduced to the region [31] .
Of note is that Ae. aegypti has most likely been established in North America for over 300 y, while Ae. albopictus has been in many areas of the US, since 1985, primarily in Florida [32] and since then has expanded its range in the country. Reiskind et al. set out to determine if Ae. aegypti and Ae. albopictus mosquitoes captured in Florida were susceptible to CHIKV infection by a La Reunion isolate [32] . Each mosquito tested was highly susceptible to infection by a full-length infectious clone of the La Réunion Island isolate, CHIKV LR2006 OPY1 strain. Even though the Ae. albopictus strains were more susceptible to infection, overall ecology and differences in human biting patterns need to be studied further Characteristically, there are two rounds of translation: (+) sense genomic RNA (49S9 = 11.7 kb) acts directly as mRNA and is partially translated (59 end) to produce non-structural proteins (nsp's). These proteins are responsible for replication and formation of a complementary (2) strand, the template for further (+) strand synthesis. Subgenomic mRNA (26 S = 4.1 kb) replication occurs through the synthesis of full-length (2) intermediate RNA, which is regulated by nsp4 and p123 precursor in early infection and later by mature nsp's. Translation of the newly synthesized sub-genomic RNA results in production of structural proteins such as Capsid and protein E2-6k-E1 (from 39 end of genome). Assembly occurs at the cell surface, and the envelope is acquired as the virus buds from the cell and release and maturation almost simultaneous occurred. Replication occurs in the cytoplasm and is very rapid (,4 h) [28, 29] . doi:10.1371/journal.pntd.0000623.g002 www.plosntds.org to gain a more accurate understanding of a potential CHIKV epidemic in the US [32] .
During the 7 d preceding birth, no human mother has been reported to transmit the disease vertically. However, about 50% of newborns delivered while the mother was infected with CHIKV contracted the disease from their mother, despite the method of delivery. Furthermore, there have been instances of CHIKV transmission from mother to fetus causing congenital illness and fetal death [33] .
During the 2005-2006 La Reunion Island outbreaks, Ramful et al. discovered that mothers could transmit CHIKV to their progeny during the perinatal period (Day 24 to Day +1) [33, 34] , and it is associated with a high degree of morbidity. By mean Day 4 of life, all of the neonates were symptomatic for CHIKV, exhibiting common CHIKF symptoms. Six neonates were confirmed to have contracted CHIKV and developed mengoencephalitis. Of those mothers who, during the La Reunion Island epidemic, were infected long before delivery, only three fetal deaths were reported [12, 33] . Ramful et al. theorized that motherto-child transmission most likely happens transplacentally shortly before delivery [33] . A similar study by Gerardin et al. reported nineteen cases of neonatal infection associated with intrapartum maternal viremia that progressed to develop encephalitis owing to vertical transmission from infected mothers [34] .
Clinical and epidemiological similarities with dengue fever make CHIKV diagnosis difficult, which may lead physicians to misdiagnose CHIKV as dengue fever; therefore, the incidence of CHIKV may actually be higher than currently believed (Table 1 ) [6, 12, 35] .
The amount of time elapsed since disease onset is the most critical parameter when choosing a diagnostic test. CHIKV can be detected and isolated by culturing with mosquito cells (C6/36), Vero cells (mammalian), or in mice [26] . However, this method can take at least a week and only achieves a high sensitivity during the viremic phase, which usually only lasts up to 48 h after the bite. Five days post-infection, the viral isolation approach has a low sensitivity but is still the preferred method for detecting the CHIKV strain [12, 26, 31, 35] . RT-PCR on the other hand is a faster and more sensitive method that can be used within the first week of disease onset [26] , and it is currently the most sensitive method for detecting and quantifying viral mRNA [4, 36] .
Classic serological detection, by assays such as ELISA [37] , immunofluorescence [5, 38] , complement binding, and haemagglutination inhibition [39] , constitutes the second diagnostic tool used for biological diagnosis of CHIKV infection. These proven techniques are useful for detection of Antigen in mosquitoes during epidemiological studies. These assays detect virus-specific IgM and IgG, however the sensitivity and specificity of these assays has been poorly characterized. Viral competence, or the potential of viral infection and transmission, is an important parameter that can be quantified by ELISA, viral culture, and PCR.
A study by Ng et al. showed biomarkers indicative of severe CHIKV infection [40] . They found decreased levels of RANTES and increased levels of Interleukin-6 (IL-6) and Interleukin-1b (IL-1b) that could be sued for CHIKV detection in patients as indicators of CHIKV-driven cytokine storm. Couderc et al. demonstrate another cytokine, type-I IFN, as a key player in the progression to CHIKV infection [26] . Using an IFN-a/b null mouse model, they demonstrated evidence of muscles, joints, and skin as privileged CHIKV targets, which is consistent with human pathology. Although Ng et al. concluded that RANTES levels were significantly suppressed in severe CHIKF patients [40] , interestingly, an increase in levels of RANTES has been observed in dengue infection [41] . Since the symptoms of CHIKF mimic those of dengue fever, results obtained from this study strongly suggest that RANTES could be a potential distinctive biomarker that differentiates between these two clinically similar diseases.
There are no approved antiviral treatments currently available for CHIKV [1, 3, 12, 42] . Currently, CHIKF is treated symptomatically, usually with non-steroidal anti-inflammatory drugs or steroids, bed rest, and fluids. Movement and mild exercise are thought to decrease stiffness and morning arthralgia, but heavy exercise may exacerbate rheumatic symptoms. Corticosteroids may be used in cases of debilitating chronic CHIKV infection. There is a debate about the appropriateness of chloroquine as treatment for unresolved, non-steroidal anti-inflammatory drugresistant arthritis [43] . A study showed that viral production was www.plosntds.org drastically reduced at 16 h post-infection after treatment with 100 mM dec-RVKR-cmk (Decanoyl-Arg-Val-Lys-Arg-chloromethylketone), a furine inhibitor [42, 44] . Chloroquine acted by raising the pH, blocking low pH-dependent entry of virus into the cell. It is important to note that dec-RVKR-cmk or chloroquine only inhibited viral spreading from cell to cell, not CHIKV replication once it had entered the cell [43] . However, most would agree that the best weapon against CHIKV is prevention. A live CHIKV vaccine developed by the United States reached phase II clinical trial encompassing 59 healthy volunteers [45] . Eight percent of the volunteers experienced transient arthralgia, while 98% of the volunteers had seroconversion [45] . However, live CHIKV vaccines are still questionable. One cannot discount the risk of a live vaccine possibly inducing chronic rheumatism. Also, there is the question as to whether widespread use among the public could trigger mosquito transmission or lead to chronic infection or viral reversion [1] .
An alternative approach would be to produce a chimeric vaccine against CHIKV. Wang et al. developed a chimeric alphavirus vaccine that is uniformly attenuated and does not cause reactogenicity in mice [3] . Three different versions of this vaccine were made using three different backbone vectors: Venezuelan equine encephalitis virus (VEEV) attenuated vaccine strain T-83, naturally attenuated eastern equine encephalitis virus (EEEV), and attenuated Sindbis virus (SINV). In short, CHIKV structural proteins were engineered into the backbones of the aforementioned vaccines to produce the chimeras [3] . These chimeras were found to stimulate a strong humoral immunity, and even at doses of 5.3-5.8 log 10 PFU, they did not trigger reactogenicity. When vaccinated mice were challenged with CHIKV, neither adult nor neonatal mice gained weight, had fever, or displayed signs of neurological illness. Upon comparison of the chimeras with the Army181/25 vaccine, the Army vaccine resulted in higher levels of viremia and replication in the joints of neonatal mice. Because the joints are known targets of CHIKV, Wang et al. noted their vaccine might avoid the negative reactogenic side effects of the Army vaccine. After being subcutaneously vaccinated with 5.3-5.8 log 10 PFU of the chimeric vaccines, mice produced strong neutralizing antibody titers. The VEEV and EEEV chimeras yielded higher neutralizing antibody titers than the SINV chimera without being more virulent. On top of this, the VEEV and EEEV CHIKV chimeras seemed to be more immunogenic than the Army vaccine despite the chimeras' lower viremia and replication in the joints of neonatal mice [3] .
Tiwari et al. [46] adopted a different strategy using formalin inactivated CHIKV in combination with alhydrogel (Aluminum Hydroxide) as an adjuvant. This study clearly suggests that this vaccine elicits both humoral and cell-mediated immune responses in mice, providing its immunogenic potential. A recent study by Couderc et al. [47] showed passive immunization as a potential treatment for CHIKV infection. Using purified immunoglobulin extracted from convalescent CHIKV patients, they demonstrated effective neutralizing activity against CHIKV infection both in vitro and in vivo. This thereby establishes a potential preventive and therapeutic approach to combat CHIKV infection. Pathogenesis studies conducted with related alpha virus, like RRV, have shown the role of macrophages in persistence on infection [48] . They also demonstrated the role of RRV-specific CD8 T cells in clearing viral load in infected patients, thereby warranting similar investigations with CHIKV and the importance of investigating a cell-mediated immune response-based vaccine against CHIKV [49] .
There are always certain risks associated with live attenuated or inactivated viral vaccines [50] . One way to avoid these potential problems is to construct a consensus-based DNA vaccine. DNA based vaccines have an improved safety profile as compared to live or attenuated vaccines [51, 52] . A consequence of CHIKV's rapid evolution is difficulty in constructing a vaccine that will be able to Figure 3 . Levels of CHIKV-specific IgG in mice immunized with CHIKV vaccines. Each group of C57BL/6 mice (n = 5) was immunized with 12.5 mg of pVax1 control vector or CHIKV vaccine plasmids as indicated at 0 and 2 wk. Mice were bled 2 wk after each immunization, and each group's serum pool was diluted to 1:100 and 1:500 for reaction with specific vaccine constructs. Serum was incubated for 1 h at 37uC on 96-well plates coated with 2 mg/ml of respective CHIKV peptides, and antibody was detected using anti-mouse IgG-HRP and OD was measured at 405 nm. doi:10.1371/journal.pntd.0000623.g003 www.plosntds.org effectively protect large populations from multiple strains of the virus. One of the strengths of DNA consensus vaccines is its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines.
Recently, Muthumani et al. constructed a vaccine that was shown to induce both humoral and cellular immunity in vivo in 3-4-wk-old female C57/BL6 mice [49] . These mice were immunized using an in vivo electroporation method to deliver the vaccine into the quadriceps muscle. The consensus construct was designed against E1, E2, and the core protein capsid. To design the construct, they aligned 21 sequences of CHIKV isolated between 1952 and 2006, using strains from differing countries, including La Reunion Island. The most common nucleotide among the sequences was chosen at each position to be used in the consensus construct, taking care not to alter the reading frame. They conducted codon and RNA optimization, added a strong Kozak sequence, and substituted signal peptide with an immunoglobulin E leader sequence to improve vaccine efficacy.
After immunizing the mice, spleens were harvested along with serum and tested to determine antibody titer. After three immunizations, consensus E1, E2, and C vaccines were shown to induce T-cell immune responses leading to strong IFN-c responses and proliferation in C57/BL6 mice. Furthermore, when compared with control mice, immunized mice had higher total IgG levels as well as higher anti-E1 specific, anti-E2 specific, and anti-C specific IgG antibodies, suggesting a strong humoral immune response ( Figure 3 ) and also specificity for the antigens encoded in the vaccine constructs ( Figure 4 ). Because of its promising results and the need for a safer vaccine, this consensus DNA vaccine deserves further investigation. Determining longevity of protective effects of the vaccine and persistence of antibody and IFN-c responses could be the next step of investigation. Challenged studies of immunized mice must also be carried out.
CHIKV mosquito-borne disease has caused massive outbreaks for at least half a century but is no longer confined to the www.plosntds.org developing nations. It began to encroach into the boundaries of the developing world. As a result, the NIAID has designated CHIKV as a Category C pathogen alongside the influenza and SARS-CoV viruses [3] . Realization of the potential severity of this disease is exigent; for instance, if used as a biological weapon, the world economy could be severely crippled; if enough members of the armed forces were to become infected during a military deployment, military operations could be significantly affected. Efforts to monitor the disease will only provide minimal warning in a global society, and steps to prevent the morbidity and mortality associated with pandemic are imperative [21, 31] . Despite the gravity of its infectious potency and the fear of it being a potential biological weapon, there is currently no vaccine for CHIKV infections. Live attenuated vaccine trials were carried out in 2000, but funding for the project was discontinued. Newer approaches such as DNA vaccines appear promising over conventional strategies like live attenuated or inactivated virus and thus call for further investigation. Recent advances such electroporation delivery and incorporation of adjuvants has boosted DNA vaccine efficacy [51, 53] . Despite the low antibody response to DNA vaccines, other numerous advantages have overshadowed these minor drawbacks (Table 2) , the most important one being the ability to induce both humoral and cellular immune responses [51, 54] .
Judging by recent success, such as the immunogenic construct developed by Muthumani et al., DNA vaccines could play a major role in combating CHIKV [49] . Vaccines are literally a critical component of CHIKV disease control and therefore research in this area is highly encouraged. The dramatic spread of dengue viruses (DENV) throughout tropical America since 1980 via the same vectors and human hosts underscores the risk to public health in the Americas. The adverse events associated with the current live vaccine are well documented [55] . Realizing these drawbacks, earnest efforts should be taken to develop new strategies to forestall further spread and complications. | What are the strengths and advantages of DNA based vaccine? | 2,529 | its ability to induce cross-reactive immune responses against the three distinct phylogenetic groups of CHIKV. Also DNA-based vaccines can be produced more rapidly than protein-based vaccines. | 20,667 |
1,688 | Health care workers indicate ill preparedness for Ebola Virus Disease outbreak in Ashanti Region of Ghana
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5461762/
SHA: f8efe7295a7cf875c8a695df3e87a42e651ce60d
Authors: Annan, Augustina Angelina; Yar, Denis Dekugmen; Owusu, Michael; Biney, Eno Akua; Forson, Paa Kobina; Okyere, Portia Boakye; Gyimah, Akosua Adumea; Owusu-Dabo, Ellis
Date: 2017-06-06
DOI: 10.1186/s12889-017-4474-6
License: cc-by
Abstract: BACKGROUND: The recent Ebola Virus Disease (EVD) epidemic that hit some countries in West Africa underscores the need to train front line high-risk health workers on disease prevention skills. Although Ghana did not record (and is yet to) any case, and several health workers have received numerous training schemes, there is no record of any study that assessed preparedness of healthcare workers (HCWS) regarding EVD and any emergency prone disease in Ghana. We therefore conducted a hospital based cross sectional study involving 101 HCWs from two facilities in Kumasi, Ghana to assess the level of preparedness of HCWs to respond to any possible EVD. METHODS: We administered a face-to-face questionnaire using an adapted WHO (2015) and CDC (2014) Checklist for Ebola Preparedness and assessed overall knowledge gaps, and preparedness of the Ghanaian HCWs in selected health facilities of the Ashanti Region of Ghana from October to December 2015. RESULTS: A total 92 (91.09%) HCWs indicated they were not adequately trained to handle an EVD suspected case. Only 25.74% (n = 26) considered their facilities sufficiently equipped to handle and manage EVD patients. When asked which disinfectant to use after attending to and caring for a suspected patient with EVD, only 8.91% (n = 9) could correctly identify the right disinfectant (χ(2) = 28.52, p = 0.001). CONCLUSION: Our study demonstrates poor knowledge and ill preparedness and unwillingness of many HCWs to attend to EVD. Beyond knowledge acquisition, there is the need for more training from time to time to fully prepare HCWs to handle any possible EVD case.
Text: During the last outbreak of Ebola Virus Disease (EVD) and its consequential massive epidemic with very high mortality [1] , many health systems and services in West Africa were overwhelmed and disrupted. This was partly due to the poor and weak health systems coupled with unprepared and unskilled frontline healthcare workers (HCWs) compounded by poor understanding of the disease dynamics tied to lack of requisite resources. During the early part of 2014, the emergence of EVD [1] in Guinea, jolted the health care systems of West African sub-region claiming over 9800 lives [2] including more than 491 (58.7%) deaths of HCWs from 839 infections [2] . This epidemic therefore reinforced the fact that HCWs are at high-risk of being infected with the disease in line with their core duties. Empirical data generated during and after the epidemic indicated how unprepared most HCWs were in the face of the crisis. Studies in Nigeria, Guinea and India indicate the low level of knowledge, negative attitude and sub-standard practices that can be eliminated through continued training of HCWs as well as provision of needed and adequate resources in their line of duties [3] [4] [5] [6] .
The countries worst hit were Liberia, Sierra Leone, Guinea and several other countries with imported cases [7] . Like most West African nations, Ghana was on high alert and was number one on the list of countries deemed to be at high risk of EVD. Thus, the country tried to make some preparations in the wake of the epidemic [8] . The government with support from donor organizations such as the World Health Organization (WHO), Médecins sans frontières (MSF) put in place resources for training of health professionals and some level of retooling of hospitals and preparedness of health workers in the face of any possible emergency scenarios. Various HCWs received both theoretical and practical training on how to manage infected and affected persons. These training sessions took the form of onsite and off site coaching as well as workshops. Simulation exercises were also conducted to bring to bear how HCWs would react during EVD emergency scenarios. For example, the German government through the Kumasi Centre for Collaborative Research in Tropical Medicine organized hands on training for several West African nationals on sample taking, sample testing and donning and doffing personal protective equipment (http://kccr.org). More importantly, there was the construction of three treatment centres and as well, a standby ambulance service for transportation of confirmed cases to the treatment centres. Incidentally, Ghana did not record any case in the wake of the epidemic and has so far not recorded any case of EVD. However, the response of HCWs to the scenarios identified several gaps. Following a series of training for HCWs, one could easily assume that health care workers are adequately prepared and equipped with the requisite knowledge and skills to deal with any possible EVD outbreak. It is unclear for example to what extent these exercises were practiced and for how long they were made a part of routine hospital activities. One therefore wonders how well prepared HCWs within Ghana are to responding not only to EVD but other epidemic prone diseases (EPDs) requiring a concerted approach to preparedness and management.
Even though some resources have been invested in response to the EVD scare in Ghana, there has not been any assessment on the preparedness of health workers in the face of any possible emergency scenarios. Simply providing the tools such as medications, personnel protective equipment (PPE) and other logistics is no panacea for adequately and appropriately responding to EPDs. Consequently, if healthcare staff lack the basic knowledge, practical and organizational skills to plan and respond to such emergency situations, it would not only spell doom for themselves but also for the general population as was the case of the recent epidemic in West Africa. It is important for example to understand the dynamics of what will propel a HCW to be willing to put his or her life in the line of duty for a case of EVD. It is therefore critical to understand current preparedness of the healthcare worker in Ghana in order to make recommendations for future training and planning for any epidemics situation. The need for Ghana to therefore have empirical data on general emergency preparedness to determine and understand knowledge gaps, and to assess knowledge versus practice in a bid to provide direction for policy cannot be overemphasized. In view of this, we therefore assessed the level of preparedness, readiness and knowledge of EVD emergency response among a population of healthcare workers (HCWs) in the Kumasi Metropolis of Ashanti Region, Ghana.
We conducted a hospital based cross-sectional study among healthcare workers at the Kumasi South and Suntreso Government Hospitals designated as "advanced Ebola holding unit" and "Ebola standing team" respectively, in the Kumasi Metropolis. The Kumasi South and Suntreso hospitals have an average monthly Out Patient Department (OPD) attendance of about 20,603 and 11,712 patients respectively and health staff of approximately 450 each. Similar to most facilities, there are more females nurses than males.
We organized a day's training for our research assistants on how to use Personal Digital Assistant device (PDAs) Samsung Galaxy note 8 GT-N5100 (Samsung Electronics Co. Ltd., Seoul, Korea) in capturing data.
The original version of the questionnaire was pretested, with five healthcare workers who were similar in their characteristics to the members of the study population but outside the area of jurisdiction and study to ensure validity and measurement bias. The questionnaire was revised based on the suggestions and comments (mainly on how the questions had been constructed) obtained from the pilot. This was the final and validated data capturing tool used during the study.
At both facilities, we contacted the Medical Superintendents to obtain permission to attend their morning meetings to explain the aims and objectives of the work to HCWs. During this time, HCWs were given the opportunity to ask questions. Two field assistants were stationed at each of the study sites for data capture. Some of the questions asked included the organism responsible for EVD, the mode of transmission of the disease, HCW preparedness to handle any EVD case and among other things early clinical features of the infection.
In estimating the sample size for this study, previous data from the hospital indicates that there are approximately 900 HCWs at the two facilities. Assuming a 95% confidence interval and if 70% of these HCWs would come into contact with an EVD suspected case, allowing an error rate of 10%, approximately 87 HCWs would provide a default study power of 80% and an alpha of 5%. With approximately a non-response rate of 15% allowing us to sample 101 HCWs from the two facilities providing emergency services within the Ashanti Region of Ghana.
Any healthcare worker attending directly to patients in emergency situation was therefore eligible for inclusion in the study. Our sampling frame consisted of a list of a total of 200. From this list, we then took a systematic random sample of all eligible health workers to represent the sample size. After obtaining written informed consent indicated by signature and or thumbprint of participants, we then administered the questionnaires within the two facilities.
We used the WHO (2015) and CDC (2014) Checklist for Ebola Preparedness that provides practical and specific suggestions to ensure that health facilities are able to help their personnel detect possible Ebola cases, protect personnel, and respond appropriately [9, 10] . This checklist included facility evaluation, knowledge and preparedness of HCWs. Based on these checklists we developed a questionnaire to ascertain the overall knowledge and preparedness of Ghanaian HCWs on EVD outbreak. Our questionnaire was administered from a PDA and recorded each participant's demographics, preparedness, form of compensation HCWs think would be appropriate when taking care of EVD case, and knowledge of EVD during the period October to December 2015. Answers to these questions were needed from HCWs to determine information access on EVD among HCWs, their knowledge about EVD and the form of compensation HCWs think would be appropriate when taking care of EVD case among others.
Data were collected electronically using tablets for cloud storage through CommCare ODK version 2.27.2, aggregated into Microsoft Excel file, exported into STATA version 14 and analyzed. Descriptive statistics was used to summarize the distribution of various variables into tables and figures. Categorical variables were analyzed using chisquare tests and logistic regression for associations.
Background of the study participants Table 1 shows the background characteristics of the study participants. A total of 101 study participants were interviewed, of which 85 (84.16%) were females. Respondents were categorized into three main groups by occupation: Nurses (76.24%), Medical Doctors (19.80%) and Physician Assistants (PA) (3.96%). Majority (54.46%) of the respondents were married. A total 52.48% (53) had been practicing their profession for less than 5 years (SD = 9.22 ± 10.52 years). At both facilities, 75.25% (76) of the respondents had been working in the facility for less than 5 years (SD = 4.04 ± 4.07 years). Table 2 shows the participants knowledge and awareness of EVD. Of the 101 HCWs interviewed, 83.17% (n = 84) correctly identified the cause of EVD, 13.86% (n = 14) did not know the cause, while 2.97% (n = 3) incorrectly labeled the cause to be a bacterium. Even though one (0.99%) Doctor and 16 (15.84%) Nurses were unable to correctly identify the cause; no group was significantly likely to incorrectly label the cause of EVD (χ 2 = 5.41, p = 0.247).
A total of 72 (71.29%) HCWs indicated media especially radio as the main source of information when asked where they first heard of EVD. This was significantly more than other sources (χ 2 = 45.44, p < 0.05). When asked which biosafety level laboratory (BSL) is required to test sample from suspected patient with EVD, a total 19 (18.81%) indicated BSL-3 of which 11 (10.89%) were Medical Doctors, while 8 (7.92) and 1 (0.99%) were Nurses and Physician Assistants, respectively. A further 76 (75.25%), of which 9 (8.91%) were doctors, 62 (61.39%) Nurses When asked which disinfectant to use after attending to and caring for a suspected patient with EVD, only 8.91% (n = 9) could correctly identify bleach (0.5% Sodium Hypochlorite) which disinfectant to use (χ 2 = 28.52, p = 0.001).
Preparedness for an EVD outbreak by HCW category Table 3 shows the levels of preparedness of HCWs to handle and manage EVD outbreak. When HCWs were asked if they considered their facilities sufficiently equipped to handle and manage EVD patients, 25.74% (n = 26) responded in the affirmative, while 54.46% (55) indicated otherwise. Of this, 14 (13.86%) were Medical Doctors, 39 (38.61%) Nurses and 2 (1.98%) were PA (χ 2 = 2.66, p = 0.62). If they became accidentally infected with EVD after attending to a patient with EVD, 98 (97.03%) of those surveyed indicated they would accept to be isolated (χ 2 = 4.69, p = 0.321). Meanwhile, 44.55% (n = 45) of HCWs would willingly attend to an EVD suspected patient (χ 2 = 8.03, p = 0.09).
A total of 92 (91.09%) HCWs surveyed indicated they were not adequately trained to handle an EVD suspected case. When asked to rate their competence in handling an EVD suspected patient, 18.81% (n = 19) indicated they had little confidence and competence, while 6.93% (n = 7) indicated they were extremely confident to handle a suspected case of EVD (χ 2 = 13.09, p = 0.11).
Beyond EVD, we asked our survey population to name other epidemic prone diseases. Of the total number of HCWs interviewed, 56.43% (57/101) mentioned epidemic diseases of bacteria origin such as tuberculosis and cholera. A further 33.70% (34/101) named diseases of viral origin such as SARS, Flu, HIV, Lassa fever and dengue, while 9.90% (10) mentioned others referring to malaria. When asked the form of compensation HCWs thought would be appropriate when taking care of an Ebola suspected patient, responses given included personal insurance (32/101), family compensation in case of death (31/101), money (30/101) and awards (8/101) while others also suggested job promotion (7/101), and others (18/101).
Our survey population recommended the provision of logistics and training as two key issues in the way forward in adequately preparing HCWs towards any epidemic prone diseases.
Many issues surrounding the preparedness of HCWs have been extensively discussed globally especially in the aftermath of the Severe Acute Respiratory Syndrome Coronavirus (SARS-CoV) and the Middle East Respiratory Syndrome (MERS)-CoV epidemic. While it is on record that the recent EVD outbreak recorded very high mortality among HCWs, to the best of our knowledge, only few studies have addressed these issues in anticipation of an EVD outbreak particularly in countries not hit by the EVD epidemic and especially in sub Saharan Africa, such a study is almost non-existent. Our study therefore assessed how prepared HCWs are in the face of a possible EVD epidemic.
The results of this survey showed that more than half (54.46%) HCWs indicated that their facilities were not ready to handle EVD cases. Nearly 92% indicated they were not adequately trained to handle an EVD suspected case and it is not surprising that less than 50% indicated they would willingly attend to a suspected patient. Moreover, nearly a third of HCWs would also want insurance for themselves and their families in case they were infected with EVD.
These results are clearly indicative of how ill-prepared the HCWs surveyed are in the face a potentially life threatening epidemic prone diseases, such as EVD in Ghana. In this study, only 25.7% of HCWs said their facility was sufficiently equipped to handle an EVD outbreak. Such low ratings of the hospitals by majority of HCWs is a mark of lack of confidence in their facilities preparedness and this may actually indicate a real lack of preparedness and readiness of the hospitals to handle not only EVD cases but potentially other epidemic prone diseases. Alternatively, it could also mean that HCWs were probably unaware of preparatory work and retooling of their facilities to handle EVD outbreak situation.
Willingness to work during outbreaks and emergencies is deemed a sense of duty even in the face of risk. In this study, less than 50% of HCWs indicated their willingness to work in the event of an EVD outbreak. Additionally, over one third indicated various forms of compensation for themselves or families in case of death or while taking care of an EVD case. This implies that if HCWs are assured or guaranteed that they and or their families would be taken care of in case of death or while taking care of an EVD case, they will willingly work in the face of any emergency scenario. The assumption is that HCWs would willingly work in the face of an infectious diseases emergency and respond appropriately; however, there are evidences of HCWs avoiding this "sacred duty" in caring for patients and would leave patients vulnerable in times of crisis [11] . In order to prevent HCWs from being infected while obliged to work even in the face of personal risk as required by their codes of ethics and professionalism, it is imperative to ensure that appropriate conventional standards, guarantees and effective public health practices are met to enable HCWs respond to such outbreaks so that they are not infected and or affected despite the risks they might face and continue to face [12] . Thus, appropriate training of HCWs as indicated by those surveyed during the study, coupled with retooling of some health facilities preparation is very critical in ensuring that they are equipped with the needed knowledge and tools needed to work with in the face of any epidemic.
General knowledge of EVD is crucial to adequately respond to and care for patients. Nearly 17% of our study population could not identify that EVD as caused by a virus. Arguably, infection control measures would be difficult and problematic for such HCWs. Less than 10% could correctly identify 0.5% Sodium Hypochlorite as the best disinfectant out of the many options provided. This strongly contradicts a similar study in Conakry conducted during the peak of the epidemic where 68% of HCWs knew the correct concentration of disinfectant [5] . While not trying to compare these two scenarios, this information may be vital in the realization that knowledge of HCWs in infection prevention and control measures is critical in their line of duty.
This study showed that most HCWs first heard of EVD through the media especially radio. This establishes the crucial role media plays in informing the general populace in such disease outbreaks. In Ghana, there are over 350 media outlets (radio and television put together) and majority of households either own a radio, television or have access to internet. Notwithstanding the media pluralism, it is still incumbent upon health institutions and facilities to organize special training on any emerging infectious disease that occurs globally to update the knowledge of HCWs.
Isolation is a key public health measure to prevent the spread of infectious diseases. In this study, over 97% of HCW indicated their willingness to comply and accept to be isolated in case they became infected after attending to suspected EVD patient. However, a small proportion of HCWs surveyed stated that they would be very unhappy, and this could ultimately affect compliance. Isolation is one of the oldest methods of controlling communicable disease outbreaks for patients [13] . However, it is worthy of note that less that 50% said they would be willing to attend to an EVD suspected patient and we suspected that this could be related to fear of personal safety [14] . Emergency response from an epidemic prone disease from an exotic virulent virus or pathogen will naturally spark some level of fear and skepticism among any group of individuals especially when their knowledge about the dynamics of the disease outbreak is low. There are stories of HCWs who have avoided the responsibility of treating patients [15] and this was apparent in the HIV/AIDS and Severe Acute Respiratory Syndrome-Coronavirus (SARS-CoV) during the 1980s and 2003, respectively where the fear of contact with suspected and infected patients gripped some HCWs [16, 17] . In the long run, this fear would likely affect their confidence and commitment to professionalism.
The results of this study point to the fact that knowledge and the provision of tools such as personnel protective equipment (PPE) and other logistics alone is not good enough strategy. There might be the need to as well address issues related to myth, and culture as well as assurances of upkeep should one be infected. The general outlook one's country's devotion to their health staff might be a contributory factor in all of this and cannot be ignored. However, getting HCWs inspired and feel safe in caring for such highly infectious disease outbreaks is critical. During our study, HCWs indicated various forms of compensation to be paid to them should they be affected in the case of EVD attack.
This study had some inherent limitations. This was an exploratory study and our sample size was limited. Therefore, while not trying to generalize the results, we are of the opinion that this may be a reflection of HCWs in general. Additionally, since our study focused mainly on two health facilities, we are again careful in extrapolating these to other to reflect other facilities. Moreover, since this has not been a real experience, and a questionnaire-based survey, responses may not accurately reflect real-life experiences in the event of an EVD epidemic. Despite these limitations, the need for training was strong among HCWs. The results further demonstrate the ill-preparedness of health facilities, and the large proportion of HCWs unwillingness to attend to a suspected case of EVD. This thus calls for concerted efforts of health institutions and facilities to fully equip and prepare HCWs with the requisite tools and knowledge and ensuring competency to handle any epidemic prone disease. | How many facilities were monitored in this study? | 4,250 | two | 1,004 |
1,688 | Health care workers indicate ill preparedness for Ebola Virus Disease outbreak in Ashanti Region of Ghana
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5461762/
SHA: f8efe7295a7cf875c8a695df3e87a42e651ce60d
Authors: Annan, Augustina Angelina; Yar, Denis Dekugmen; Owusu, Michael; Biney, Eno Akua; Forson, Paa Kobina; Okyere, Portia Boakye; Gyimah, Akosua Adumea; Owusu-Dabo, Ellis
Date: 2017-06-06
DOI: 10.1186/s12889-017-4474-6
License: cc-by
Abstract: BACKGROUND: The recent Ebola Virus Disease (EVD) epidemic that hit some countries in West Africa underscores the need to train front line high-risk health workers on disease prevention skills. Although Ghana did not record (and is yet to) any case, and several health workers have received numerous training schemes, there is no record of any study that assessed preparedness of healthcare workers (HCWS) regarding EVD and any emergency prone disease in Ghana. We therefore conducted a hospital based cross sectional study involving 101 HCWs from two facilities in Kumasi, Ghana to assess the level of preparedness of HCWs to respond to any possible EVD. METHODS: We administered a face-to-face questionnaire using an adapted WHO (2015) and CDC (2014) Checklist for Ebola Preparedness and assessed overall knowledge gaps, and preparedness of the Ghanaian HCWs in selected health facilities of the Ashanti Region of Ghana from October to December 2015. RESULTS: A total 92 (91.09%) HCWs indicated they were not adequately trained to handle an EVD suspected case. Only 25.74% (n = 26) considered their facilities sufficiently equipped to handle and manage EVD patients. When asked which disinfectant to use after attending to and caring for a suspected patient with EVD, only 8.91% (n = 9) could correctly identify the right disinfectant (χ(2) = 28.52, p = 0.001). CONCLUSION: Our study demonstrates poor knowledge and ill preparedness and unwillingness of many HCWs to attend to EVD. Beyond knowledge acquisition, there is the need for more training from time to time to fully prepare HCWs to handle any possible EVD case.
Text: During the last outbreak of Ebola Virus Disease (EVD) and its consequential massive epidemic with very high mortality [1] , many health systems and services in West Africa were overwhelmed and disrupted. This was partly due to the poor and weak health systems coupled with unprepared and unskilled frontline healthcare workers (HCWs) compounded by poor understanding of the disease dynamics tied to lack of requisite resources. During the early part of 2014, the emergence of EVD [1] in Guinea, jolted the health care systems of West African sub-region claiming over 9800 lives [2] including more than 491 (58.7%) deaths of HCWs from 839 infections [2] . This epidemic therefore reinforced the fact that HCWs are at high-risk of being infected with the disease in line with their core duties. Empirical data generated during and after the epidemic indicated how unprepared most HCWs were in the face of the crisis. Studies in Nigeria, Guinea and India indicate the low level of knowledge, negative attitude and sub-standard practices that can be eliminated through continued training of HCWs as well as provision of needed and adequate resources in their line of duties [3] [4] [5] [6] .
The countries worst hit were Liberia, Sierra Leone, Guinea and several other countries with imported cases [7] . Like most West African nations, Ghana was on high alert and was number one on the list of countries deemed to be at high risk of EVD. Thus, the country tried to make some preparations in the wake of the epidemic [8] . The government with support from donor organizations such as the World Health Organization (WHO), Médecins sans frontières (MSF) put in place resources for training of health professionals and some level of retooling of hospitals and preparedness of health workers in the face of any possible emergency scenarios. Various HCWs received both theoretical and practical training on how to manage infected and affected persons. These training sessions took the form of onsite and off site coaching as well as workshops. Simulation exercises were also conducted to bring to bear how HCWs would react during EVD emergency scenarios. For example, the German government through the Kumasi Centre for Collaborative Research in Tropical Medicine organized hands on training for several West African nationals on sample taking, sample testing and donning and doffing personal protective equipment (http://kccr.org). More importantly, there was the construction of three treatment centres and as well, a standby ambulance service for transportation of confirmed cases to the treatment centres. Incidentally, Ghana did not record any case in the wake of the epidemic and has so far not recorded any case of EVD. However, the response of HCWs to the scenarios identified several gaps. Following a series of training for HCWs, one could easily assume that health care workers are adequately prepared and equipped with the requisite knowledge and skills to deal with any possible EVD outbreak. It is unclear for example to what extent these exercises were practiced and for how long they were made a part of routine hospital activities. One therefore wonders how well prepared HCWs within Ghana are to responding not only to EVD but other epidemic prone diseases (EPDs) requiring a concerted approach to preparedness and management.
Even though some resources have been invested in response to the EVD scare in Ghana, there has not been any assessment on the preparedness of health workers in the face of any possible emergency scenarios. Simply providing the tools such as medications, personnel protective equipment (PPE) and other logistics is no panacea for adequately and appropriately responding to EPDs. Consequently, if healthcare staff lack the basic knowledge, practical and organizational skills to plan and respond to such emergency situations, it would not only spell doom for themselves but also for the general population as was the case of the recent epidemic in West Africa. It is important for example to understand the dynamics of what will propel a HCW to be willing to put his or her life in the line of duty for a case of EVD. It is therefore critical to understand current preparedness of the healthcare worker in Ghana in order to make recommendations for future training and planning for any epidemics situation. The need for Ghana to therefore have empirical data on general emergency preparedness to determine and understand knowledge gaps, and to assess knowledge versus practice in a bid to provide direction for policy cannot be overemphasized. In view of this, we therefore assessed the level of preparedness, readiness and knowledge of EVD emergency response among a population of healthcare workers (HCWs) in the Kumasi Metropolis of Ashanti Region, Ghana.
We conducted a hospital based cross-sectional study among healthcare workers at the Kumasi South and Suntreso Government Hospitals designated as "advanced Ebola holding unit" and "Ebola standing team" respectively, in the Kumasi Metropolis. The Kumasi South and Suntreso hospitals have an average monthly Out Patient Department (OPD) attendance of about 20,603 and 11,712 patients respectively and health staff of approximately 450 each. Similar to most facilities, there are more females nurses than males.
We organized a day's training for our research assistants on how to use Personal Digital Assistant device (PDAs) Samsung Galaxy note 8 GT-N5100 (Samsung Electronics Co. Ltd., Seoul, Korea) in capturing data.
The original version of the questionnaire was pretested, with five healthcare workers who were similar in their characteristics to the members of the study population but outside the area of jurisdiction and study to ensure validity and measurement bias. The questionnaire was revised based on the suggestions and comments (mainly on how the questions had been constructed) obtained from the pilot. This was the final and validated data capturing tool used during the study.
At both facilities, we contacted the Medical Superintendents to obtain permission to attend their morning meetings to explain the aims and objectives of the work to HCWs. During this time, HCWs were given the opportunity to ask questions. Two field assistants were stationed at each of the study sites for data capture. Some of the questions asked included the organism responsible for EVD, the mode of transmission of the disease, HCW preparedness to handle any EVD case and among other things early clinical features of the infection.
In estimating the sample size for this study, previous data from the hospital indicates that there are approximately 900 HCWs at the two facilities. Assuming a 95% confidence interval and if 70% of these HCWs would come into contact with an EVD suspected case, allowing an error rate of 10%, approximately 87 HCWs would provide a default study power of 80% and an alpha of 5%. With approximately a non-response rate of 15% allowing us to sample 101 HCWs from the two facilities providing emergency services within the Ashanti Region of Ghana.
Any healthcare worker attending directly to patients in emergency situation was therefore eligible for inclusion in the study. Our sampling frame consisted of a list of a total of 200. From this list, we then took a systematic random sample of all eligible health workers to represent the sample size. After obtaining written informed consent indicated by signature and or thumbprint of participants, we then administered the questionnaires within the two facilities.
We used the WHO (2015) and CDC (2014) Checklist for Ebola Preparedness that provides practical and specific suggestions to ensure that health facilities are able to help their personnel detect possible Ebola cases, protect personnel, and respond appropriately [9, 10] . This checklist included facility evaluation, knowledge and preparedness of HCWs. Based on these checklists we developed a questionnaire to ascertain the overall knowledge and preparedness of Ghanaian HCWs on EVD outbreak. Our questionnaire was administered from a PDA and recorded each participant's demographics, preparedness, form of compensation HCWs think would be appropriate when taking care of EVD case, and knowledge of EVD during the period October to December 2015. Answers to these questions were needed from HCWs to determine information access on EVD among HCWs, their knowledge about EVD and the form of compensation HCWs think would be appropriate when taking care of EVD case among others.
Data were collected electronically using tablets for cloud storage through CommCare ODK version 2.27.2, aggregated into Microsoft Excel file, exported into STATA version 14 and analyzed. Descriptive statistics was used to summarize the distribution of various variables into tables and figures. Categorical variables were analyzed using chisquare tests and logistic regression for associations.
Background of the study participants Table 1 shows the background characteristics of the study participants. A total of 101 study participants were interviewed, of which 85 (84.16%) were females. Respondents were categorized into three main groups by occupation: Nurses (76.24%), Medical Doctors (19.80%) and Physician Assistants (PA) (3.96%). Majority (54.46%) of the respondents were married. A total 52.48% (53) had been practicing their profession for less than 5 years (SD = 9.22 ± 10.52 years). At both facilities, 75.25% (76) of the respondents had been working in the facility for less than 5 years (SD = 4.04 ± 4.07 years). Table 2 shows the participants knowledge and awareness of EVD. Of the 101 HCWs interviewed, 83.17% (n = 84) correctly identified the cause of EVD, 13.86% (n = 14) did not know the cause, while 2.97% (n = 3) incorrectly labeled the cause to be a bacterium. Even though one (0.99%) Doctor and 16 (15.84%) Nurses were unable to correctly identify the cause; no group was significantly likely to incorrectly label the cause of EVD (χ 2 = 5.41, p = 0.247).
A total of 72 (71.29%) HCWs indicated media especially radio as the main source of information when asked where they first heard of EVD. This was significantly more than other sources (χ 2 = 45.44, p < 0.05). When asked which biosafety level laboratory (BSL) is required to test sample from suspected patient with EVD, a total 19 (18.81%) indicated BSL-3 of which 11 (10.89%) were Medical Doctors, while 8 (7.92) and 1 (0.99%) were Nurses and Physician Assistants, respectively. A further 76 (75.25%), of which 9 (8.91%) were doctors, 62 (61.39%) Nurses When asked which disinfectant to use after attending to and caring for a suspected patient with EVD, only 8.91% (n = 9) could correctly identify bleach (0.5% Sodium Hypochlorite) which disinfectant to use (χ 2 = 28.52, p = 0.001).
Preparedness for an EVD outbreak by HCW category Table 3 shows the levels of preparedness of HCWs to handle and manage EVD outbreak. When HCWs were asked if they considered their facilities sufficiently equipped to handle and manage EVD patients, 25.74% (n = 26) responded in the affirmative, while 54.46% (55) indicated otherwise. Of this, 14 (13.86%) were Medical Doctors, 39 (38.61%) Nurses and 2 (1.98%) were PA (χ 2 = 2.66, p = 0.62). If they became accidentally infected with EVD after attending to a patient with EVD, 98 (97.03%) of those surveyed indicated they would accept to be isolated (χ 2 = 4.69, p = 0.321). Meanwhile, 44.55% (n = 45) of HCWs would willingly attend to an EVD suspected patient (χ 2 = 8.03, p = 0.09).
A total of 92 (91.09%) HCWs surveyed indicated they were not adequately trained to handle an EVD suspected case. When asked to rate their competence in handling an EVD suspected patient, 18.81% (n = 19) indicated they had little confidence and competence, while 6.93% (n = 7) indicated they were extremely confident to handle a suspected case of EVD (χ 2 = 13.09, p = 0.11).
Beyond EVD, we asked our survey population to name other epidemic prone diseases. Of the total number of HCWs interviewed, 56.43% (57/101) mentioned epidemic diseases of bacteria origin such as tuberculosis and cholera. A further 33.70% (34/101) named diseases of viral origin such as SARS, Flu, HIV, Lassa fever and dengue, while 9.90% (10) mentioned others referring to malaria. When asked the form of compensation HCWs thought would be appropriate when taking care of an Ebola suspected patient, responses given included personal insurance (32/101), family compensation in case of death (31/101), money (30/101) and awards (8/101) while others also suggested job promotion (7/101), and others (18/101).
Our survey population recommended the provision of logistics and training as two key issues in the way forward in adequately preparing HCWs towards any epidemic prone diseases.
Many issues surrounding the preparedness of HCWs have been extensively discussed globally especially in the aftermath of the Severe Acute Respiratory Syndrome Coronavirus (SARS-CoV) and the Middle East Respiratory Syndrome (MERS)-CoV epidemic. While it is on record that the recent EVD outbreak recorded very high mortality among HCWs, to the best of our knowledge, only few studies have addressed these issues in anticipation of an EVD outbreak particularly in countries not hit by the EVD epidemic and especially in sub Saharan Africa, such a study is almost non-existent. Our study therefore assessed how prepared HCWs are in the face of a possible EVD epidemic.
The results of this survey showed that more than half (54.46%) HCWs indicated that their facilities were not ready to handle EVD cases. Nearly 92% indicated they were not adequately trained to handle an EVD suspected case and it is not surprising that less than 50% indicated they would willingly attend to a suspected patient. Moreover, nearly a third of HCWs would also want insurance for themselves and their families in case they were infected with EVD.
These results are clearly indicative of how ill-prepared the HCWs surveyed are in the face a potentially life threatening epidemic prone diseases, such as EVD in Ghana. In this study, only 25.7% of HCWs said their facility was sufficiently equipped to handle an EVD outbreak. Such low ratings of the hospitals by majority of HCWs is a mark of lack of confidence in their facilities preparedness and this may actually indicate a real lack of preparedness and readiness of the hospitals to handle not only EVD cases but potentially other epidemic prone diseases. Alternatively, it could also mean that HCWs were probably unaware of preparatory work and retooling of their facilities to handle EVD outbreak situation.
Willingness to work during outbreaks and emergencies is deemed a sense of duty even in the face of risk. In this study, less than 50% of HCWs indicated their willingness to work in the event of an EVD outbreak. Additionally, over one third indicated various forms of compensation for themselves or families in case of death or while taking care of an EVD case. This implies that if HCWs are assured or guaranteed that they and or their families would be taken care of in case of death or while taking care of an EVD case, they will willingly work in the face of any emergency scenario. The assumption is that HCWs would willingly work in the face of an infectious diseases emergency and respond appropriately; however, there are evidences of HCWs avoiding this "sacred duty" in caring for patients and would leave patients vulnerable in times of crisis [11] . In order to prevent HCWs from being infected while obliged to work even in the face of personal risk as required by their codes of ethics and professionalism, it is imperative to ensure that appropriate conventional standards, guarantees and effective public health practices are met to enable HCWs respond to such outbreaks so that they are not infected and or affected despite the risks they might face and continue to face [12] . Thus, appropriate training of HCWs as indicated by those surveyed during the study, coupled with retooling of some health facilities preparation is very critical in ensuring that they are equipped with the needed knowledge and tools needed to work with in the face of any epidemic.
General knowledge of EVD is crucial to adequately respond to and care for patients. Nearly 17% of our study population could not identify that EVD as caused by a virus. Arguably, infection control measures would be difficult and problematic for such HCWs. Less than 10% could correctly identify 0.5% Sodium Hypochlorite as the best disinfectant out of the many options provided. This strongly contradicts a similar study in Conakry conducted during the peak of the epidemic where 68% of HCWs knew the correct concentration of disinfectant [5] . While not trying to compare these two scenarios, this information may be vital in the realization that knowledge of HCWs in infection prevention and control measures is critical in their line of duty.
This study showed that most HCWs first heard of EVD through the media especially radio. This establishes the crucial role media plays in informing the general populace in such disease outbreaks. In Ghana, there are over 350 media outlets (radio and television put together) and majority of households either own a radio, television or have access to internet. Notwithstanding the media pluralism, it is still incumbent upon health institutions and facilities to organize special training on any emerging infectious disease that occurs globally to update the knowledge of HCWs.
Isolation is a key public health measure to prevent the spread of infectious diseases. In this study, over 97% of HCW indicated their willingness to comply and accept to be isolated in case they became infected after attending to suspected EVD patient. However, a small proportion of HCWs surveyed stated that they would be very unhappy, and this could ultimately affect compliance. Isolation is one of the oldest methods of controlling communicable disease outbreaks for patients [13] . However, it is worthy of note that less that 50% said they would be willing to attend to an EVD suspected patient and we suspected that this could be related to fear of personal safety [14] . Emergency response from an epidemic prone disease from an exotic virulent virus or pathogen will naturally spark some level of fear and skepticism among any group of individuals especially when their knowledge about the dynamics of the disease outbreak is low. There are stories of HCWs who have avoided the responsibility of treating patients [15] and this was apparent in the HIV/AIDS and Severe Acute Respiratory Syndrome-Coronavirus (SARS-CoV) during the 1980s and 2003, respectively where the fear of contact with suspected and infected patients gripped some HCWs [16, 17] . In the long run, this fear would likely affect their confidence and commitment to professionalism.
The results of this study point to the fact that knowledge and the provision of tools such as personnel protective equipment (PPE) and other logistics alone is not good enough strategy. There might be the need to as well address issues related to myth, and culture as well as assurances of upkeep should one be infected. The general outlook one's country's devotion to their health staff might be a contributory factor in all of this and cannot be ignored. However, getting HCWs inspired and feel safe in caring for such highly infectious disease outbreaks is critical. During our study, HCWs indicated various forms of compensation to be paid to them should they be affected in the case of EVD attack.
This study had some inherent limitations. This was an exploratory study and our sample size was limited. Therefore, while not trying to generalize the results, we are of the opinion that this may be a reflection of HCWs in general. Additionally, since our study focused mainly on two health facilities, we are again careful in extrapolating these to other to reflect other facilities. Moreover, since this has not been a real experience, and a questionnaire-based survey, responses may not accurately reflect real-life experiences in the event of an EVD epidemic. Despite these limitations, the need for training was strong among HCWs. The results further demonstrate the ill-preparedness of health facilities, and the large proportion of HCWs unwillingness to attend to a suspected case of EVD. This thus calls for concerted efforts of health institutions and facilities to fully equip and prepare HCWs with the requisite tools and knowledge and ensuring competency to handle any epidemic prone disease. | What percentage of facilities believed they were adequately equipped to handle Ebola virus disease? | 4,251 | 25.74% | 1,524 |
1,688 | Health care workers indicate ill preparedness for Ebola Virus Disease outbreak in Ashanti Region of Ghana
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5461762/
SHA: f8efe7295a7cf875c8a695df3e87a42e651ce60d
Authors: Annan, Augustina Angelina; Yar, Denis Dekugmen; Owusu, Michael; Biney, Eno Akua; Forson, Paa Kobina; Okyere, Portia Boakye; Gyimah, Akosua Adumea; Owusu-Dabo, Ellis
Date: 2017-06-06
DOI: 10.1186/s12889-017-4474-6
License: cc-by
Abstract: BACKGROUND: The recent Ebola Virus Disease (EVD) epidemic that hit some countries in West Africa underscores the need to train front line high-risk health workers on disease prevention skills. Although Ghana did not record (and is yet to) any case, and several health workers have received numerous training schemes, there is no record of any study that assessed preparedness of healthcare workers (HCWS) regarding EVD and any emergency prone disease in Ghana. We therefore conducted a hospital based cross sectional study involving 101 HCWs from two facilities in Kumasi, Ghana to assess the level of preparedness of HCWs to respond to any possible EVD. METHODS: We administered a face-to-face questionnaire using an adapted WHO (2015) and CDC (2014) Checklist for Ebola Preparedness and assessed overall knowledge gaps, and preparedness of the Ghanaian HCWs in selected health facilities of the Ashanti Region of Ghana from October to December 2015. RESULTS: A total 92 (91.09%) HCWs indicated they were not adequately trained to handle an EVD suspected case. Only 25.74% (n = 26) considered their facilities sufficiently equipped to handle and manage EVD patients. When asked which disinfectant to use after attending to and caring for a suspected patient with EVD, only 8.91% (n = 9) could correctly identify the right disinfectant (χ(2) = 28.52, p = 0.001). CONCLUSION: Our study demonstrates poor knowledge and ill preparedness and unwillingness of many HCWs to attend to EVD. Beyond knowledge acquisition, there is the need for more training from time to time to fully prepare HCWs to handle any possible EVD case.
Text: During the last outbreak of Ebola Virus Disease (EVD) and its consequential massive epidemic with very high mortality [1] , many health systems and services in West Africa were overwhelmed and disrupted. This was partly due to the poor and weak health systems coupled with unprepared and unskilled frontline healthcare workers (HCWs) compounded by poor understanding of the disease dynamics tied to lack of requisite resources. During the early part of 2014, the emergence of EVD [1] in Guinea, jolted the health care systems of West African sub-region claiming over 9800 lives [2] including more than 491 (58.7%) deaths of HCWs from 839 infections [2] . This epidemic therefore reinforced the fact that HCWs are at high-risk of being infected with the disease in line with their core duties. Empirical data generated during and after the epidemic indicated how unprepared most HCWs were in the face of the crisis. Studies in Nigeria, Guinea and India indicate the low level of knowledge, negative attitude and sub-standard practices that can be eliminated through continued training of HCWs as well as provision of needed and adequate resources in their line of duties [3] [4] [5] [6] .
The countries worst hit were Liberia, Sierra Leone, Guinea and several other countries with imported cases [7] . Like most West African nations, Ghana was on high alert and was number one on the list of countries deemed to be at high risk of EVD. Thus, the country tried to make some preparations in the wake of the epidemic [8] . The government with support from donor organizations such as the World Health Organization (WHO), Médecins sans frontières (MSF) put in place resources for training of health professionals and some level of retooling of hospitals and preparedness of health workers in the face of any possible emergency scenarios. Various HCWs received both theoretical and practical training on how to manage infected and affected persons. These training sessions took the form of onsite and off site coaching as well as workshops. Simulation exercises were also conducted to bring to bear how HCWs would react during EVD emergency scenarios. For example, the German government through the Kumasi Centre for Collaborative Research in Tropical Medicine organized hands on training for several West African nationals on sample taking, sample testing and donning and doffing personal protective equipment (http://kccr.org). More importantly, there was the construction of three treatment centres and as well, a standby ambulance service for transportation of confirmed cases to the treatment centres. Incidentally, Ghana did not record any case in the wake of the epidemic and has so far not recorded any case of EVD. However, the response of HCWs to the scenarios identified several gaps. Following a series of training for HCWs, one could easily assume that health care workers are adequately prepared and equipped with the requisite knowledge and skills to deal with any possible EVD outbreak. It is unclear for example to what extent these exercises were practiced and for how long they were made a part of routine hospital activities. One therefore wonders how well prepared HCWs within Ghana are to responding not only to EVD but other epidemic prone diseases (EPDs) requiring a concerted approach to preparedness and management.
Even though some resources have been invested in response to the EVD scare in Ghana, there has not been any assessment on the preparedness of health workers in the face of any possible emergency scenarios. Simply providing the tools such as medications, personnel protective equipment (PPE) and other logistics is no panacea for adequately and appropriately responding to EPDs. Consequently, if healthcare staff lack the basic knowledge, practical and organizational skills to plan and respond to such emergency situations, it would not only spell doom for themselves but also for the general population as was the case of the recent epidemic in West Africa. It is important for example to understand the dynamics of what will propel a HCW to be willing to put his or her life in the line of duty for a case of EVD. It is therefore critical to understand current preparedness of the healthcare worker in Ghana in order to make recommendations for future training and planning for any epidemics situation. The need for Ghana to therefore have empirical data on general emergency preparedness to determine and understand knowledge gaps, and to assess knowledge versus practice in a bid to provide direction for policy cannot be overemphasized. In view of this, we therefore assessed the level of preparedness, readiness and knowledge of EVD emergency response among a population of healthcare workers (HCWs) in the Kumasi Metropolis of Ashanti Region, Ghana.
We conducted a hospital based cross-sectional study among healthcare workers at the Kumasi South and Suntreso Government Hospitals designated as "advanced Ebola holding unit" and "Ebola standing team" respectively, in the Kumasi Metropolis. The Kumasi South and Suntreso hospitals have an average monthly Out Patient Department (OPD) attendance of about 20,603 and 11,712 patients respectively and health staff of approximately 450 each. Similar to most facilities, there are more females nurses than males.
We organized a day's training for our research assistants on how to use Personal Digital Assistant device (PDAs) Samsung Galaxy note 8 GT-N5100 (Samsung Electronics Co. Ltd., Seoul, Korea) in capturing data.
The original version of the questionnaire was pretested, with five healthcare workers who were similar in their characteristics to the members of the study population but outside the area of jurisdiction and study to ensure validity and measurement bias. The questionnaire was revised based on the suggestions and comments (mainly on how the questions had been constructed) obtained from the pilot. This was the final and validated data capturing tool used during the study.
At both facilities, we contacted the Medical Superintendents to obtain permission to attend their morning meetings to explain the aims and objectives of the work to HCWs. During this time, HCWs were given the opportunity to ask questions. Two field assistants were stationed at each of the study sites for data capture. Some of the questions asked included the organism responsible for EVD, the mode of transmission of the disease, HCW preparedness to handle any EVD case and among other things early clinical features of the infection.
In estimating the sample size for this study, previous data from the hospital indicates that there are approximately 900 HCWs at the two facilities. Assuming a 95% confidence interval and if 70% of these HCWs would come into contact with an EVD suspected case, allowing an error rate of 10%, approximately 87 HCWs would provide a default study power of 80% and an alpha of 5%. With approximately a non-response rate of 15% allowing us to sample 101 HCWs from the two facilities providing emergency services within the Ashanti Region of Ghana.
Any healthcare worker attending directly to patients in emergency situation was therefore eligible for inclusion in the study. Our sampling frame consisted of a list of a total of 200. From this list, we then took a systematic random sample of all eligible health workers to represent the sample size. After obtaining written informed consent indicated by signature and or thumbprint of participants, we then administered the questionnaires within the two facilities.
We used the WHO (2015) and CDC (2014) Checklist for Ebola Preparedness that provides practical and specific suggestions to ensure that health facilities are able to help their personnel detect possible Ebola cases, protect personnel, and respond appropriately [9, 10] . This checklist included facility evaluation, knowledge and preparedness of HCWs. Based on these checklists we developed a questionnaire to ascertain the overall knowledge and preparedness of Ghanaian HCWs on EVD outbreak. Our questionnaire was administered from a PDA and recorded each participant's demographics, preparedness, form of compensation HCWs think would be appropriate when taking care of EVD case, and knowledge of EVD during the period October to December 2015. Answers to these questions were needed from HCWs to determine information access on EVD among HCWs, their knowledge about EVD and the form of compensation HCWs think would be appropriate when taking care of EVD case among others.
Data were collected electronically using tablets for cloud storage through CommCare ODK version 2.27.2, aggregated into Microsoft Excel file, exported into STATA version 14 and analyzed. Descriptive statistics was used to summarize the distribution of various variables into tables and figures. Categorical variables were analyzed using chisquare tests and logistic regression for associations.
Background of the study participants Table 1 shows the background characteristics of the study participants. A total of 101 study participants were interviewed, of which 85 (84.16%) were females. Respondents were categorized into three main groups by occupation: Nurses (76.24%), Medical Doctors (19.80%) and Physician Assistants (PA) (3.96%). Majority (54.46%) of the respondents were married. A total 52.48% (53) had been practicing their profession for less than 5 years (SD = 9.22 ± 10.52 years). At both facilities, 75.25% (76) of the respondents had been working in the facility for less than 5 years (SD = 4.04 ± 4.07 years). Table 2 shows the participants knowledge and awareness of EVD. Of the 101 HCWs interviewed, 83.17% (n = 84) correctly identified the cause of EVD, 13.86% (n = 14) did not know the cause, while 2.97% (n = 3) incorrectly labeled the cause to be a bacterium. Even though one (0.99%) Doctor and 16 (15.84%) Nurses were unable to correctly identify the cause; no group was significantly likely to incorrectly label the cause of EVD (χ 2 = 5.41, p = 0.247).
A total of 72 (71.29%) HCWs indicated media especially radio as the main source of information when asked where they first heard of EVD. This was significantly more than other sources (χ 2 = 45.44, p < 0.05). When asked which biosafety level laboratory (BSL) is required to test sample from suspected patient with EVD, a total 19 (18.81%) indicated BSL-3 of which 11 (10.89%) were Medical Doctors, while 8 (7.92) and 1 (0.99%) were Nurses and Physician Assistants, respectively. A further 76 (75.25%), of which 9 (8.91%) were doctors, 62 (61.39%) Nurses When asked which disinfectant to use after attending to and caring for a suspected patient with EVD, only 8.91% (n = 9) could correctly identify bleach (0.5% Sodium Hypochlorite) which disinfectant to use (χ 2 = 28.52, p = 0.001).
Preparedness for an EVD outbreak by HCW category Table 3 shows the levels of preparedness of HCWs to handle and manage EVD outbreak. When HCWs were asked if they considered their facilities sufficiently equipped to handle and manage EVD patients, 25.74% (n = 26) responded in the affirmative, while 54.46% (55) indicated otherwise. Of this, 14 (13.86%) were Medical Doctors, 39 (38.61%) Nurses and 2 (1.98%) were PA (χ 2 = 2.66, p = 0.62). If they became accidentally infected with EVD after attending to a patient with EVD, 98 (97.03%) of those surveyed indicated they would accept to be isolated (χ 2 = 4.69, p = 0.321). Meanwhile, 44.55% (n = 45) of HCWs would willingly attend to an EVD suspected patient (χ 2 = 8.03, p = 0.09).
A total of 92 (91.09%) HCWs surveyed indicated they were not adequately trained to handle an EVD suspected case. When asked to rate their competence in handling an EVD suspected patient, 18.81% (n = 19) indicated they had little confidence and competence, while 6.93% (n = 7) indicated they were extremely confident to handle a suspected case of EVD (χ 2 = 13.09, p = 0.11).
Beyond EVD, we asked our survey population to name other epidemic prone diseases. Of the total number of HCWs interviewed, 56.43% (57/101) mentioned epidemic diseases of bacteria origin such as tuberculosis and cholera. A further 33.70% (34/101) named diseases of viral origin such as SARS, Flu, HIV, Lassa fever and dengue, while 9.90% (10) mentioned others referring to malaria. When asked the form of compensation HCWs thought would be appropriate when taking care of an Ebola suspected patient, responses given included personal insurance (32/101), family compensation in case of death (31/101), money (30/101) and awards (8/101) while others also suggested job promotion (7/101), and others (18/101).
Our survey population recommended the provision of logistics and training as two key issues in the way forward in adequately preparing HCWs towards any epidemic prone diseases.
Many issues surrounding the preparedness of HCWs have been extensively discussed globally especially in the aftermath of the Severe Acute Respiratory Syndrome Coronavirus (SARS-CoV) and the Middle East Respiratory Syndrome (MERS)-CoV epidemic. While it is on record that the recent EVD outbreak recorded very high mortality among HCWs, to the best of our knowledge, only few studies have addressed these issues in anticipation of an EVD outbreak particularly in countries not hit by the EVD epidemic and especially in sub Saharan Africa, such a study is almost non-existent. Our study therefore assessed how prepared HCWs are in the face of a possible EVD epidemic.
The results of this survey showed that more than half (54.46%) HCWs indicated that their facilities were not ready to handle EVD cases. Nearly 92% indicated they were not adequately trained to handle an EVD suspected case and it is not surprising that less than 50% indicated they would willingly attend to a suspected patient. Moreover, nearly a third of HCWs would also want insurance for themselves and their families in case they were infected with EVD.
These results are clearly indicative of how ill-prepared the HCWs surveyed are in the face a potentially life threatening epidemic prone diseases, such as EVD in Ghana. In this study, only 25.7% of HCWs said their facility was sufficiently equipped to handle an EVD outbreak. Such low ratings of the hospitals by majority of HCWs is a mark of lack of confidence in their facilities preparedness and this may actually indicate a real lack of preparedness and readiness of the hospitals to handle not only EVD cases but potentially other epidemic prone diseases. Alternatively, it could also mean that HCWs were probably unaware of preparatory work and retooling of their facilities to handle EVD outbreak situation.
Willingness to work during outbreaks and emergencies is deemed a sense of duty even in the face of risk. In this study, less than 50% of HCWs indicated their willingness to work in the event of an EVD outbreak. Additionally, over one third indicated various forms of compensation for themselves or families in case of death or while taking care of an EVD case. This implies that if HCWs are assured or guaranteed that they and or their families would be taken care of in case of death or while taking care of an EVD case, they will willingly work in the face of any emergency scenario. The assumption is that HCWs would willingly work in the face of an infectious diseases emergency and respond appropriately; however, there are evidences of HCWs avoiding this "sacred duty" in caring for patients and would leave patients vulnerable in times of crisis [11] . In order to prevent HCWs from being infected while obliged to work even in the face of personal risk as required by their codes of ethics and professionalism, it is imperative to ensure that appropriate conventional standards, guarantees and effective public health practices are met to enable HCWs respond to such outbreaks so that they are not infected and or affected despite the risks they might face and continue to face [12] . Thus, appropriate training of HCWs as indicated by those surveyed during the study, coupled with retooling of some health facilities preparation is very critical in ensuring that they are equipped with the needed knowledge and tools needed to work with in the face of any epidemic.
General knowledge of EVD is crucial to adequately respond to and care for patients. Nearly 17% of our study population could not identify that EVD as caused by a virus. Arguably, infection control measures would be difficult and problematic for such HCWs. Less than 10% could correctly identify 0.5% Sodium Hypochlorite as the best disinfectant out of the many options provided. This strongly contradicts a similar study in Conakry conducted during the peak of the epidemic where 68% of HCWs knew the correct concentration of disinfectant [5] . While not trying to compare these two scenarios, this information may be vital in the realization that knowledge of HCWs in infection prevention and control measures is critical in their line of duty.
This study showed that most HCWs first heard of EVD through the media especially radio. This establishes the crucial role media plays in informing the general populace in such disease outbreaks. In Ghana, there are over 350 media outlets (radio and television put together) and majority of households either own a radio, television or have access to internet. Notwithstanding the media pluralism, it is still incumbent upon health institutions and facilities to organize special training on any emerging infectious disease that occurs globally to update the knowledge of HCWs.
Isolation is a key public health measure to prevent the spread of infectious diseases. In this study, over 97% of HCW indicated their willingness to comply and accept to be isolated in case they became infected after attending to suspected EVD patient. However, a small proportion of HCWs surveyed stated that they would be very unhappy, and this could ultimately affect compliance. Isolation is one of the oldest methods of controlling communicable disease outbreaks for patients [13] . However, it is worthy of note that less that 50% said they would be willing to attend to an EVD suspected patient and we suspected that this could be related to fear of personal safety [14] . Emergency response from an epidemic prone disease from an exotic virulent virus or pathogen will naturally spark some level of fear and skepticism among any group of individuals especially when their knowledge about the dynamics of the disease outbreak is low. There are stories of HCWs who have avoided the responsibility of treating patients [15] and this was apparent in the HIV/AIDS and Severe Acute Respiratory Syndrome-Coronavirus (SARS-CoV) during the 1980s and 2003, respectively where the fear of contact with suspected and infected patients gripped some HCWs [16, 17] . In the long run, this fear would likely affect their confidence and commitment to professionalism.
The results of this study point to the fact that knowledge and the provision of tools such as personnel protective equipment (PPE) and other logistics alone is not good enough strategy. There might be the need to as well address issues related to myth, and culture as well as assurances of upkeep should one be infected. The general outlook one's country's devotion to their health staff might be a contributory factor in all of this and cannot be ignored. However, getting HCWs inspired and feel safe in caring for such highly infectious disease outbreaks is critical. During our study, HCWs indicated various forms of compensation to be paid to them should they be affected in the case of EVD attack.
This study had some inherent limitations. This was an exploratory study and our sample size was limited. Therefore, while not trying to generalize the results, we are of the opinion that this may be a reflection of HCWs in general. Additionally, since our study focused mainly on two health facilities, we are again careful in extrapolating these to other to reflect other facilities. Moreover, since this has not been a real experience, and a questionnaire-based survey, responses may not accurately reflect real-life experiences in the event of an EVD epidemic. Despite these limitations, the need for training was strong among HCWs. The results further demonstrate the ill-preparedness of health facilities, and the large proportion of HCWs unwillingness to attend to a suspected case of EVD. This thus calls for concerted efforts of health institutions and facilities to fully equip and prepare HCWs with the requisite tools and knowledge and ensuring competency to handle any epidemic prone disease. | How many facilities believed they were adequately equipped to handle Ebla virus disease? | 4,252 | 26 | 1,536 |
1,688 | Health care workers indicate ill preparedness for Ebola Virus Disease outbreak in Ashanti Region of Ghana
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5461762/
SHA: f8efe7295a7cf875c8a695df3e87a42e651ce60d
Authors: Annan, Augustina Angelina; Yar, Denis Dekugmen; Owusu, Michael; Biney, Eno Akua; Forson, Paa Kobina; Okyere, Portia Boakye; Gyimah, Akosua Adumea; Owusu-Dabo, Ellis
Date: 2017-06-06
DOI: 10.1186/s12889-017-4474-6
License: cc-by
Abstract: BACKGROUND: The recent Ebola Virus Disease (EVD) epidemic that hit some countries in West Africa underscores the need to train front line high-risk health workers on disease prevention skills. Although Ghana did not record (and is yet to) any case, and several health workers have received numerous training schemes, there is no record of any study that assessed preparedness of healthcare workers (HCWS) regarding EVD and any emergency prone disease in Ghana. We therefore conducted a hospital based cross sectional study involving 101 HCWs from two facilities in Kumasi, Ghana to assess the level of preparedness of HCWs to respond to any possible EVD. METHODS: We administered a face-to-face questionnaire using an adapted WHO (2015) and CDC (2014) Checklist for Ebola Preparedness and assessed overall knowledge gaps, and preparedness of the Ghanaian HCWs in selected health facilities of the Ashanti Region of Ghana from October to December 2015. RESULTS: A total 92 (91.09%) HCWs indicated they were not adequately trained to handle an EVD suspected case. Only 25.74% (n = 26) considered their facilities sufficiently equipped to handle and manage EVD patients. When asked which disinfectant to use after attending to and caring for a suspected patient with EVD, only 8.91% (n = 9) could correctly identify the right disinfectant (χ(2) = 28.52, p = 0.001). CONCLUSION: Our study demonstrates poor knowledge and ill preparedness and unwillingness of many HCWs to attend to EVD. Beyond knowledge acquisition, there is the need for more training from time to time to fully prepare HCWs to handle any possible EVD case.
Text: During the last outbreak of Ebola Virus Disease (EVD) and its consequential massive epidemic with very high mortality [1] , many health systems and services in West Africa were overwhelmed and disrupted. This was partly due to the poor and weak health systems coupled with unprepared and unskilled frontline healthcare workers (HCWs) compounded by poor understanding of the disease dynamics tied to lack of requisite resources. During the early part of 2014, the emergence of EVD [1] in Guinea, jolted the health care systems of West African sub-region claiming over 9800 lives [2] including more than 491 (58.7%) deaths of HCWs from 839 infections [2] . This epidemic therefore reinforced the fact that HCWs are at high-risk of being infected with the disease in line with their core duties. Empirical data generated during and after the epidemic indicated how unprepared most HCWs were in the face of the crisis. Studies in Nigeria, Guinea and India indicate the low level of knowledge, negative attitude and sub-standard practices that can be eliminated through continued training of HCWs as well as provision of needed and adequate resources in their line of duties [3] [4] [5] [6] .
The countries worst hit were Liberia, Sierra Leone, Guinea and several other countries with imported cases [7] . Like most West African nations, Ghana was on high alert and was number one on the list of countries deemed to be at high risk of EVD. Thus, the country tried to make some preparations in the wake of the epidemic [8] . The government with support from donor organizations such as the World Health Organization (WHO), Médecins sans frontières (MSF) put in place resources for training of health professionals and some level of retooling of hospitals and preparedness of health workers in the face of any possible emergency scenarios. Various HCWs received both theoretical and practical training on how to manage infected and affected persons. These training sessions took the form of onsite and off site coaching as well as workshops. Simulation exercises were also conducted to bring to bear how HCWs would react during EVD emergency scenarios. For example, the German government through the Kumasi Centre for Collaborative Research in Tropical Medicine organized hands on training for several West African nationals on sample taking, sample testing and donning and doffing personal protective equipment (http://kccr.org). More importantly, there was the construction of three treatment centres and as well, a standby ambulance service for transportation of confirmed cases to the treatment centres. Incidentally, Ghana did not record any case in the wake of the epidemic and has so far not recorded any case of EVD. However, the response of HCWs to the scenarios identified several gaps. Following a series of training for HCWs, one could easily assume that health care workers are adequately prepared and equipped with the requisite knowledge and skills to deal with any possible EVD outbreak. It is unclear for example to what extent these exercises were practiced and for how long they were made a part of routine hospital activities. One therefore wonders how well prepared HCWs within Ghana are to responding not only to EVD but other epidemic prone diseases (EPDs) requiring a concerted approach to preparedness and management.
Even though some resources have been invested in response to the EVD scare in Ghana, there has not been any assessment on the preparedness of health workers in the face of any possible emergency scenarios. Simply providing the tools such as medications, personnel protective equipment (PPE) and other logistics is no panacea for adequately and appropriately responding to EPDs. Consequently, if healthcare staff lack the basic knowledge, practical and organizational skills to plan and respond to such emergency situations, it would not only spell doom for themselves but also for the general population as was the case of the recent epidemic in West Africa. It is important for example to understand the dynamics of what will propel a HCW to be willing to put his or her life in the line of duty for a case of EVD. It is therefore critical to understand current preparedness of the healthcare worker in Ghana in order to make recommendations for future training and planning for any epidemics situation. The need for Ghana to therefore have empirical data on general emergency preparedness to determine and understand knowledge gaps, and to assess knowledge versus practice in a bid to provide direction for policy cannot be overemphasized. In view of this, we therefore assessed the level of preparedness, readiness and knowledge of EVD emergency response among a population of healthcare workers (HCWs) in the Kumasi Metropolis of Ashanti Region, Ghana.
We conducted a hospital based cross-sectional study among healthcare workers at the Kumasi South and Suntreso Government Hospitals designated as "advanced Ebola holding unit" and "Ebola standing team" respectively, in the Kumasi Metropolis. The Kumasi South and Suntreso hospitals have an average monthly Out Patient Department (OPD) attendance of about 20,603 and 11,712 patients respectively and health staff of approximately 450 each. Similar to most facilities, there are more females nurses than males.
We organized a day's training for our research assistants on how to use Personal Digital Assistant device (PDAs) Samsung Galaxy note 8 GT-N5100 (Samsung Electronics Co. Ltd., Seoul, Korea) in capturing data.
The original version of the questionnaire was pretested, with five healthcare workers who were similar in their characteristics to the members of the study population but outside the area of jurisdiction and study to ensure validity and measurement bias. The questionnaire was revised based on the suggestions and comments (mainly on how the questions had been constructed) obtained from the pilot. This was the final and validated data capturing tool used during the study.
At both facilities, we contacted the Medical Superintendents to obtain permission to attend their morning meetings to explain the aims and objectives of the work to HCWs. During this time, HCWs were given the opportunity to ask questions. Two field assistants were stationed at each of the study sites for data capture. Some of the questions asked included the organism responsible for EVD, the mode of transmission of the disease, HCW preparedness to handle any EVD case and among other things early clinical features of the infection.
In estimating the sample size for this study, previous data from the hospital indicates that there are approximately 900 HCWs at the two facilities. Assuming a 95% confidence interval and if 70% of these HCWs would come into contact with an EVD suspected case, allowing an error rate of 10%, approximately 87 HCWs would provide a default study power of 80% and an alpha of 5%. With approximately a non-response rate of 15% allowing us to sample 101 HCWs from the two facilities providing emergency services within the Ashanti Region of Ghana.
Any healthcare worker attending directly to patients in emergency situation was therefore eligible for inclusion in the study. Our sampling frame consisted of a list of a total of 200. From this list, we then took a systematic random sample of all eligible health workers to represent the sample size. After obtaining written informed consent indicated by signature and or thumbprint of participants, we then administered the questionnaires within the two facilities.
We used the WHO (2015) and CDC (2014) Checklist for Ebola Preparedness that provides practical and specific suggestions to ensure that health facilities are able to help their personnel detect possible Ebola cases, protect personnel, and respond appropriately [9, 10] . This checklist included facility evaluation, knowledge and preparedness of HCWs. Based on these checklists we developed a questionnaire to ascertain the overall knowledge and preparedness of Ghanaian HCWs on EVD outbreak. Our questionnaire was administered from a PDA and recorded each participant's demographics, preparedness, form of compensation HCWs think would be appropriate when taking care of EVD case, and knowledge of EVD during the period October to December 2015. Answers to these questions were needed from HCWs to determine information access on EVD among HCWs, their knowledge about EVD and the form of compensation HCWs think would be appropriate when taking care of EVD case among others.
Data were collected electronically using tablets for cloud storage through CommCare ODK version 2.27.2, aggregated into Microsoft Excel file, exported into STATA version 14 and analyzed. Descriptive statistics was used to summarize the distribution of various variables into tables and figures. Categorical variables were analyzed using chisquare tests and logistic regression for associations.
Background of the study participants Table 1 shows the background characteristics of the study participants. A total of 101 study participants were interviewed, of which 85 (84.16%) were females. Respondents were categorized into three main groups by occupation: Nurses (76.24%), Medical Doctors (19.80%) and Physician Assistants (PA) (3.96%). Majority (54.46%) of the respondents were married. A total 52.48% (53) had been practicing their profession for less than 5 years (SD = 9.22 ± 10.52 years). At both facilities, 75.25% (76) of the respondents had been working in the facility for less than 5 years (SD = 4.04 ± 4.07 years). Table 2 shows the participants knowledge and awareness of EVD. Of the 101 HCWs interviewed, 83.17% (n = 84) correctly identified the cause of EVD, 13.86% (n = 14) did not know the cause, while 2.97% (n = 3) incorrectly labeled the cause to be a bacterium. Even though one (0.99%) Doctor and 16 (15.84%) Nurses were unable to correctly identify the cause; no group was significantly likely to incorrectly label the cause of EVD (χ 2 = 5.41, p = 0.247).
A total of 72 (71.29%) HCWs indicated media especially radio as the main source of information when asked where they first heard of EVD. This was significantly more than other sources (χ 2 = 45.44, p < 0.05). When asked which biosafety level laboratory (BSL) is required to test sample from suspected patient with EVD, a total 19 (18.81%) indicated BSL-3 of which 11 (10.89%) were Medical Doctors, while 8 (7.92) and 1 (0.99%) were Nurses and Physician Assistants, respectively. A further 76 (75.25%), of which 9 (8.91%) were doctors, 62 (61.39%) Nurses When asked which disinfectant to use after attending to and caring for a suspected patient with EVD, only 8.91% (n = 9) could correctly identify bleach (0.5% Sodium Hypochlorite) which disinfectant to use (χ 2 = 28.52, p = 0.001).
Preparedness for an EVD outbreak by HCW category Table 3 shows the levels of preparedness of HCWs to handle and manage EVD outbreak. When HCWs were asked if they considered their facilities sufficiently equipped to handle and manage EVD patients, 25.74% (n = 26) responded in the affirmative, while 54.46% (55) indicated otherwise. Of this, 14 (13.86%) were Medical Doctors, 39 (38.61%) Nurses and 2 (1.98%) were PA (χ 2 = 2.66, p = 0.62). If they became accidentally infected with EVD after attending to a patient with EVD, 98 (97.03%) of those surveyed indicated they would accept to be isolated (χ 2 = 4.69, p = 0.321). Meanwhile, 44.55% (n = 45) of HCWs would willingly attend to an EVD suspected patient (χ 2 = 8.03, p = 0.09).
A total of 92 (91.09%) HCWs surveyed indicated they were not adequately trained to handle an EVD suspected case. When asked to rate their competence in handling an EVD suspected patient, 18.81% (n = 19) indicated they had little confidence and competence, while 6.93% (n = 7) indicated they were extremely confident to handle a suspected case of EVD (χ 2 = 13.09, p = 0.11).
Beyond EVD, we asked our survey population to name other epidemic prone diseases. Of the total number of HCWs interviewed, 56.43% (57/101) mentioned epidemic diseases of bacteria origin such as tuberculosis and cholera. A further 33.70% (34/101) named diseases of viral origin such as SARS, Flu, HIV, Lassa fever and dengue, while 9.90% (10) mentioned others referring to malaria. When asked the form of compensation HCWs thought would be appropriate when taking care of an Ebola suspected patient, responses given included personal insurance (32/101), family compensation in case of death (31/101), money (30/101) and awards (8/101) while others also suggested job promotion (7/101), and others (18/101).
Our survey population recommended the provision of logistics and training as two key issues in the way forward in adequately preparing HCWs towards any epidemic prone diseases.
Many issues surrounding the preparedness of HCWs have been extensively discussed globally especially in the aftermath of the Severe Acute Respiratory Syndrome Coronavirus (SARS-CoV) and the Middle East Respiratory Syndrome (MERS)-CoV epidemic. While it is on record that the recent EVD outbreak recorded very high mortality among HCWs, to the best of our knowledge, only few studies have addressed these issues in anticipation of an EVD outbreak particularly in countries not hit by the EVD epidemic and especially in sub Saharan Africa, such a study is almost non-existent. Our study therefore assessed how prepared HCWs are in the face of a possible EVD epidemic.
The results of this survey showed that more than half (54.46%) HCWs indicated that their facilities were not ready to handle EVD cases. Nearly 92% indicated they were not adequately trained to handle an EVD suspected case and it is not surprising that less than 50% indicated they would willingly attend to a suspected patient. Moreover, nearly a third of HCWs would also want insurance for themselves and their families in case they were infected with EVD.
These results are clearly indicative of how ill-prepared the HCWs surveyed are in the face a potentially life threatening epidemic prone diseases, such as EVD in Ghana. In this study, only 25.7% of HCWs said their facility was sufficiently equipped to handle an EVD outbreak. Such low ratings of the hospitals by majority of HCWs is a mark of lack of confidence in their facilities preparedness and this may actually indicate a real lack of preparedness and readiness of the hospitals to handle not only EVD cases but potentially other epidemic prone diseases. Alternatively, it could also mean that HCWs were probably unaware of preparatory work and retooling of their facilities to handle EVD outbreak situation.
Willingness to work during outbreaks and emergencies is deemed a sense of duty even in the face of risk. In this study, less than 50% of HCWs indicated their willingness to work in the event of an EVD outbreak. Additionally, over one third indicated various forms of compensation for themselves or families in case of death or while taking care of an EVD case. This implies that if HCWs are assured or guaranteed that they and or their families would be taken care of in case of death or while taking care of an EVD case, they will willingly work in the face of any emergency scenario. The assumption is that HCWs would willingly work in the face of an infectious diseases emergency and respond appropriately; however, there are evidences of HCWs avoiding this "sacred duty" in caring for patients and would leave patients vulnerable in times of crisis [11] . In order to prevent HCWs from being infected while obliged to work even in the face of personal risk as required by their codes of ethics and professionalism, it is imperative to ensure that appropriate conventional standards, guarantees and effective public health practices are met to enable HCWs respond to such outbreaks so that they are not infected and or affected despite the risks they might face and continue to face [12] . Thus, appropriate training of HCWs as indicated by those surveyed during the study, coupled with retooling of some health facilities preparation is very critical in ensuring that they are equipped with the needed knowledge and tools needed to work with in the face of any epidemic.
General knowledge of EVD is crucial to adequately respond to and care for patients. Nearly 17% of our study population could not identify that EVD as caused by a virus. Arguably, infection control measures would be difficult and problematic for such HCWs. Less than 10% could correctly identify 0.5% Sodium Hypochlorite as the best disinfectant out of the many options provided. This strongly contradicts a similar study in Conakry conducted during the peak of the epidemic where 68% of HCWs knew the correct concentration of disinfectant [5] . While not trying to compare these two scenarios, this information may be vital in the realization that knowledge of HCWs in infection prevention and control measures is critical in their line of duty.
This study showed that most HCWs first heard of EVD through the media especially radio. This establishes the crucial role media plays in informing the general populace in such disease outbreaks. In Ghana, there are over 350 media outlets (radio and television put together) and majority of households either own a radio, television or have access to internet. Notwithstanding the media pluralism, it is still incumbent upon health institutions and facilities to organize special training on any emerging infectious disease that occurs globally to update the knowledge of HCWs.
Isolation is a key public health measure to prevent the spread of infectious diseases. In this study, over 97% of HCW indicated their willingness to comply and accept to be isolated in case they became infected after attending to suspected EVD patient. However, a small proportion of HCWs surveyed stated that they would be very unhappy, and this could ultimately affect compliance. Isolation is one of the oldest methods of controlling communicable disease outbreaks for patients [13] . However, it is worthy of note that less that 50% said they would be willing to attend to an EVD suspected patient and we suspected that this could be related to fear of personal safety [14] . Emergency response from an epidemic prone disease from an exotic virulent virus or pathogen will naturally spark some level of fear and skepticism among any group of individuals especially when their knowledge about the dynamics of the disease outbreak is low. There are stories of HCWs who have avoided the responsibility of treating patients [15] and this was apparent in the HIV/AIDS and Severe Acute Respiratory Syndrome-Coronavirus (SARS-CoV) during the 1980s and 2003, respectively where the fear of contact with suspected and infected patients gripped some HCWs [16, 17] . In the long run, this fear would likely affect their confidence and commitment to professionalism.
The results of this study point to the fact that knowledge and the provision of tools such as personnel protective equipment (PPE) and other logistics alone is not good enough strategy. There might be the need to as well address issues related to myth, and culture as well as assurances of upkeep should one be infected. The general outlook one's country's devotion to their health staff might be a contributory factor in all of this and cannot be ignored. However, getting HCWs inspired and feel safe in caring for such highly infectious disease outbreaks is critical. During our study, HCWs indicated various forms of compensation to be paid to them should they be affected in the case of EVD attack.
This study had some inherent limitations. This was an exploratory study and our sample size was limited. Therefore, while not trying to generalize the results, we are of the opinion that this may be a reflection of HCWs in general. Additionally, since our study focused mainly on two health facilities, we are again careful in extrapolating these to other to reflect other facilities. Moreover, since this has not been a real experience, and a questionnaire-based survey, responses may not accurately reflect real-life experiences in the event of an EVD epidemic. Despite these limitations, the need for training was strong among HCWs. The results further demonstrate the ill-preparedness of health facilities, and the large proportion of HCWs unwillingness to attend to a suspected case of EVD. This thus calls for concerted efforts of health institutions and facilities to fully equip and prepare HCWs with the requisite tools and knowledge and ensuring competency to handle any epidemic prone disease. | How many healthcare workers would be willing to continue working during the Ebola virus outbreak? | 4,253 | less than 50% | 16,830 |
1,688 | Health care workers indicate ill preparedness for Ebola Virus Disease outbreak in Ashanti Region of Ghana
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5461762/
SHA: f8efe7295a7cf875c8a695df3e87a42e651ce60d
Authors: Annan, Augustina Angelina; Yar, Denis Dekugmen; Owusu, Michael; Biney, Eno Akua; Forson, Paa Kobina; Okyere, Portia Boakye; Gyimah, Akosua Adumea; Owusu-Dabo, Ellis
Date: 2017-06-06
DOI: 10.1186/s12889-017-4474-6
License: cc-by
Abstract: BACKGROUND: The recent Ebola Virus Disease (EVD) epidemic that hit some countries in West Africa underscores the need to train front line high-risk health workers on disease prevention skills. Although Ghana did not record (and is yet to) any case, and several health workers have received numerous training schemes, there is no record of any study that assessed preparedness of healthcare workers (HCWS) regarding EVD and any emergency prone disease in Ghana. We therefore conducted a hospital based cross sectional study involving 101 HCWs from two facilities in Kumasi, Ghana to assess the level of preparedness of HCWs to respond to any possible EVD. METHODS: We administered a face-to-face questionnaire using an adapted WHO (2015) and CDC (2014) Checklist for Ebola Preparedness and assessed overall knowledge gaps, and preparedness of the Ghanaian HCWs in selected health facilities of the Ashanti Region of Ghana from October to December 2015. RESULTS: A total 92 (91.09%) HCWs indicated they were not adequately trained to handle an EVD suspected case. Only 25.74% (n = 26) considered their facilities sufficiently equipped to handle and manage EVD patients. When asked which disinfectant to use after attending to and caring for a suspected patient with EVD, only 8.91% (n = 9) could correctly identify the right disinfectant (χ(2) = 28.52, p = 0.001). CONCLUSION: Our study demonstrates poor knowledge and ill preparedness and unwillingness of many HCWs to attend to EVD. Beyond knowledge acquisition, there is the need for more training from time to time to fully prepare HCWs to handle any possible EVD case.
Text: During the last outbreak of Ebola Virus Disease (EVD) and its consequential massive epidemic with very high mortality [1] , many health systems and services in West Africa were overwhelmed and disrupted. This was partly due to the poor and weak health systems coupled with unprepared and unskilled frontline healthcare workers (HCWs) compounded by poor understanding of the disease dynamics tied to lack of requisite resources. During the early part of 2014, the emergence of EVD [1] in Guinea, jolted the health care systems of West African sub-region claiming over 9800 lives [2] including more than 491 (58.7%) deaths of HCWs from 839 infections [2] . This epidemic therefore reinforced the fact that HCWs are at high-risk of being infected with the disease in line with their core duties. Empirical data generated during and after the epidemic indicated how unprepared most HCWs were in the face of the crisis. Studies in Nigeria, Guinea and India indicate the low level of knowledge, negative attitude and sub-standard practices that can be eliminated through continued training of HCWs as well as provision of needed and adequate resources in their line of duties [3] [4] [5] [6] .
The countries worst hit were Liberia, Sierra Leone, Guinea and several other countries with imported cases [7] . Like most West African nations, Ghana was on high alert and was number one on the list of countries deemed to be at high risk of EVD. Thus, the country tried to make some preparations in the wake of the epidemic [8] . The government with support from donor organizations such as the World Health Organization (WHO), Médecins sans frontières (MSF) put in place resources for training of health professionals and some level of retooling of hospitals and preparedness of health workers in the face of any possible emergency scenarios. Various HCWs received both theoretical and practical training on how to manage infected and affected persons. These training sessions took the form of onsite and off site coaching as well as workshops. Simulation exercises were also conducted to bring to bear how HCWs would react during EVD emergency scenarios. For example, the German government through the Kumasi Centre for Collaborative Research in Tropical Medicine organized hands on training for several West African nationals on sample taking, sample testing and donning and doffing personal protective equipment (http://kccr.org). More importantly, there was the construction of three treatment centres and as well, a standby ambulance service for transportation of confirmed cases to the treatment centres. Incidentally, Ghana did not record any case in the wake of the epidemic and has so far not recorded any case of EVD. However, the response of HCWs to the scenarios identified several gaps. Following a series of training for HCWs, one could easily assume that health care workers are adequately prepared and equipped with the requisite knowledge and skills to deal with any possible EVD outbreak. It is unclear for example to what extent these exercises were practiced and for how long they were made a part of routine hospital activities. One therefore wonders how well prepared HCWs within Ghana are to responding not only to EVD but other epidemic prone diseases (EPDs) requiring a concerted approach to preparedness and management.
Even though some resources have been invested in response to the EVD scare in Ghana, there has not been any assessment on the preparedness of health workers in the face of any possible emergency scenarios. Simply providing the tools such as medications, personnel protective equipment (PPE) and other logistics is no panacea for adequately and appropriately responding to EPDs. Consequently, if healthcare staff lack the basic knowledge, practical and organizational skills to plan and respond to such emergency situations, it would not only spell doom for themselves but also for the general population as was the case of the recent epidemic in West Africa. It is important for example to understand the dynamics of what will propel a HCW to be willing to put his or her life in the line of duty for a case of EVD. It is therefore critical to understand current preparedness of the healthcare worker in Ghana in order to make recommendations for future training and planning for any epidemics situation. The need for Ghana to therefore have empirical data on general emergency preparedness to determine and understand knowledge gaps, and to assess knowledge versus practice in a bid to provide direction for policy cannot be overemphasized. In view of this, we therefore assessed the level of preparedness, readiness and knowledge of EVD emergency response among a population of healthcare workers (HCWs) in the Kumasi Metropolis of Ashanti Region, Ghana.
We conducted a hospital based cross-sectional study among healthcare workers at the Kumasi South and Suntreso Government Hospitals designated as "advanced Ebola holding unit" and "Ebola standing team" respectively, in the Kumasi Metropolis. The Kumasi South and Suntreso hospitals have an average monthly Out Patient Department (OPD) attendance of about 20,603 and 11,712 patients respectively and health staff of approximately 450 each. Similar to most facilities, there are more females nurses than males.
We organized a day's training for our research assistants on how to use Personal Digital Assistant device (PDAs) Samsung Galaxy note 8 GT-N5100 (Samsung Electronics Co. Ltd., Seoul, Korea) in capturing data.
The original version of the questionnaire was pretested, with five healthcare workers who were similar in their characteristics to the members of the study population but outside the area of jurisdiction and study to ensure validity and measurement bias. The questionnaire was revised based on the suggestions and comments (mainly on how the questions had been constructed) obtained from the pilot. This was the final and validated data capturing tool used during the study.
At both facilities, we contacted the Medical Superintendents to obtain permission to attend their morning meetings to explain the aims and objectives of the work to HCWs. During this time, HCWs were given the opportunity to ask questions. Two field assistants were stationed at each of the study sites for data capture. Some of the questions asked included the organism responsible for EVD, the mode of transmission of the disease, HCW preparedness to handle any EVD case and among other things early clinical features of the infection.
In estimating the sample size for this study, previous data from the hospital indicates that there are approximately 900 HCWs at the two facilities. Assuming a 95% confidence interval and if 70% of these HCWs would come into contact with an EVD suspected case, allowing an error rate of 10%, approximately 87 HCWs would provide a default study power of 80% and an alpha of 5%. With approximately a non-response rate of 15% allowing us to sample 101 HCWs from the two facilities providing emergency services within the Ashanti Region of Ghana.
Any healthcare worker attending directly to patients in emergency situation was therefore eligible for inclusion in the study. Our sampling frame consisted of a list of a total of 200. From this list, we then took a systematic random sample of all eligible health workers to represent the sample size. After obtaining written informed consent indicated by signature and or thumbprint of participants, we then administered the questionnaires within the two facilities.
We used the WHO (2015) and CDC (2014) Checklist for Ebola Preparedness that provides practical and specific suggestions to ensure that health facilities are able to help their personnel detect possible Ebola cases, protect personnel, and respond appropriately [9, 10] . This checklist included facility evaluation, knowledge and preparedness of HCWs. Based on these checklists we developed a questionnaire to ascertain the overall knowledge and preparedness of Ghanaian HCWs on EVD outbreak. Our questionnaire was administered from a PDA and recorded each participant's demographics, preparedness, form of compensation HCWs think would be appropriate when taking care of EVD case, and knowledge of EVD during the period October to December 2015. Answers to these questions were needed from HCWs to determine information access on EVD among HCWs, their knowledge about EVD and the form of compensation HCWs think would be appropriate when taking care of EVD case among others.
Data were collected electronically using tablets for cloud storage through CommCare ODK version 2.27.2, aggregated into Microsoft Excel file, exported into STATA version 14 and analyzed. Descriptive statistics was used to summarize the distribution of various variables into tables and figures. Categorical variables were analyzed using chisquare tests and logistic regression for associations.
Background of the study participants Table 1 shows the background characteristics of the study participants. A total of 101 study participants were interviewed, of which 85 (84.16%) were females. Respondents were categorized into three main groups by occupation: Nurses (76.24%), Medical Doctors (19.80%) and Physician Assistants (PA) (3.96%). Majority (54.46%) of the respondents were married. A total 52.48% (53) had been practicing their profession for less than 5 years (SD = 9.22 ± 10.52 years). At both facilities, 75.25% (76) of the respondents had been working in the facility for less than 5 years (SD = 4.04 ± 4.07 years). Table 2 shows the participants knowledge and awareness of EVD. Of the 101 HCWs interviewed, 83.17% (n = 84) correctly identified the cause of EVD, 13.86% (n = 14) did not know the cause, while 2.97% (n = 3) incorrectly labeled the cause to be a bacterium. Even though one (0.99%) Doctor and 16 (15.84%) Nurses were unable to correctly identify the cause; no group was significantly likely to incorrectly label the cause of EVD (χ 2 = 5.41, p = 0.247).
A total of 72 (71.29%) HCWs indicated media especially radio as the main source of information when asked where they first heard of EVD. This was significantly more than other sources (χ 2 = 45.44, p < 0.05). When asked which biosafety level laboratory (BSL) is required to test sample from suspected patient with EVD, a total 19 (18.81%) indicated BSL-3 of which 11 (10.89%) were Medical Doctors, while 8 (7.92) and 1 (0.99%) were Nurses and Physician Assistants, respectively. A further 76 (75.25%), of which 9 (8.91%) were doctors, 62 (61.39%) Nurses When asked which disinfectant to use after attending to and caring for a suspected patient with EVD, only 8.91% (n = 9) could correctly identify bleach (0.5% Sodium Hypochlorite) which disinfectant to use (χ 2 = 28.52, p = 0.001).
Preparedness for an EVD outbreak by HCW category Table 3 shows the levels of preparedness of HCWs to handle and manage EVD outbreak. When HCWs were asked if they considered their facilities sufficiently equipped to handle and manage EVD patients, 25.74% (n = 26) responded in the affirmative, while 54.46% (55) indicated otherwise. Of this, 14 (13.86%) were Medical Doctors, 39 (38.61%) Nurses and 2 (1.98%) were PA (χ 2 = 2.66, p = 0.62). If they became accidentally infected with EVD after attending to a patient with EVD, 98 (97.03%) of those surveyed indicated they would accept to be isolated (χ 2 = 4.69, p = 0.321). Meanwhile, 44.55% (n = 45) of HCWs would willingly attend to an EVD suspected patient (χ 2 = 8.03, p = 0.09).
A total of 92 (91.09%) HCWs surveyed indicated they were not adequately trained to handle an EVD suspected case. When asked to rate their competence in handling an EVD suspected patient, 18.81% (n = 19) indicated they had little confidence and competence, while 6.93% (n = 7) indicated they were extremely confident to handle a suspected case of EVD (χ 2 = 13.09, p = 0.11).
Beyond EVD, we asked our survey population to name other epidemic prone diseases. Of the total number of HCWs interviewed, 56.43% (57/101) mentioned epidemic diseases of bacteria origin such as tuberculosis and cholera. A further 33.70% (34/101) named diseases of viral origin such as SARS, Flu, HIV, Lassa fever and dengue, while 9.90% (10) mentioned others referring to malaria. When asked the form of compensation HCWs thought would be appropriate when taking care of an Ebola suspected patient, responses given included personal insurance (32/101), family compensation in case of death (31/101), money (30/101) and awards (8/101) while others also suggested job promotion (7/101), and others (18/101).
Our survey population recommended the provision of logistics and training as two key issues in the way forward in adequately preparing HCWs towards any epidemic prone diseases.
Many issues surrounding the preparedness of HCWs have been extensively discussed globally especially in the aftermath of the Severe Acute Respiratory Syndrome Coronavirus (SARS-CoV) and the Middle East Respiratory Syndrome (MERS)-CoV epidemic. While it is on record that the recent EVD outbreak recorded very high mortality among HCWs, to the best of our knowledge, only few studies have addressed these issues in anticipation of an EVD outbreak particularly in countries not hit by the EVD epidemic and especially in sub Saharan Africa, such a study is almost non-existent. Our study therefore assessed how prepared HCWs are in the face of a possible EVD epidemic.
The results of this survey showed that more than half (54.46%) HCWs indicated that their facilities were not ready to handle EVD cases. Nearly 92% indicated they were not adequately trained to handle an EVD suspected case and it is not surprising that less than 50% indicated they would willingly attend to a suspected patient. Moreover, nearly a third of HCWs would also want insurance for themselves and their families in case they were infected with EVD.
These results are clearly indicative of how ill-prepared the HCWs surveyed are in the face a potentially life threatening epidemic prone diseases, such as EVD in Ghana. In this study, only 25.7% of HCWs said their facility was sufficiently equipped to handle an EVD outbreak. Such low ratings of the hospitals by majority of HCWs is a mark of lack of confidence in their facilities preparedness and this may actually indicate a real lack of preparedness and readiness of the hospitals to handle not only EVD cases but potentially other epidemic prone diseases. Alternatively, it could also mean that HCWs were probably unaware of preparatory work and retooling of their facilities to handle EVD outbreak situation.
Willingness to work during outbreaks and emergencies is deemed a sense of duty even in the face of risk. In this study, less than 50% of HCWs indicated their willingness to work in the event of an EVD outbreak. Additionally, over one third indicated various forms of compensation for themselves or families in case of death or while taking care of an EVD case. This implies that if HCWs are assured or guaranteed that they and or their families would be taken care of in case of death or while taking care of an EVD case, they will willingly work in the face of any emergency scenario. The assumption is that HCWs would willingly work in the face of an infectious diseases emergency and respond appropriately; however, there are evidences of HCWs avoiding this "sacred duty" in caring for patients and would leave patients vulnerable in times of crisis [11] . In order to prevent HCWs from being infected while obliged to work even in the face of personal risk as required by their codes of ethics and professionalism, it is imperative to ensure that appropriate conventional standards, guarantees and effective public health practices are met to enable HCWs respond to such outbreaks so that they are not infected and or affected despite the risks they might face and continue to face [12] . Thus, appropriate training of HCWs as indicated by those surveyed during the study, coupled with retooling of some health facilities preparation is very critical in ensuring that they are equipped with the needed knowledge and tools needed to work with in the face of any epidemic.
General knowledge of EVD is crucial to adequately respond to and care for patients. Nearly 17% of our study population could not identify that EVD as caused by a virus. Arguably, infection control measures would be difficult and problematic for such HCWs. Less than 10% could correctly identify 0.5% Sodium Hypochlorite as the best disinfectant out of the many options provided. This strongly contradicts a similar study in Conakry conducted during the peak of the epidemic where 68% of HCWs knew the correct concentration of disinfectant [5] . While not trying to compare these two scenarios, this information may be vital in the realization that knowledge of HCWs in infection prevention and control measures is critical in their line of duty.
This study showed that most HCWs first heard of EVD through the media especially radio. This establishes the crucial role media plays in informing the general populace in such disease outbreaks. In Ghana, there are over 350 media outlets (radio and television put together) and majority of households either own a radio, television or have access to internet. Notwithstanding the media pluralism, it is still incumbent upon health institutions and facilities to organize special training on any emerging infectious disease that occurs globally to update the knowledge of HCWs.
Isolation is a key public health measure to prevent the spread of infectious diseases. In this study, over 97% of HCW indicated their willingness to comply and accept to be isolated in case they became infected after attending to suspected EVD patient. However, a small proportion of HCWs surveyed stated that they would be very unhappy, and this could ultimately affect compliance. Isolation is one of the oldest methods of controlling communicable disease outbreaks for patients [13] . However, it is worthy of note that less that 50% said they would be willing to attend to an EVD suspected patient and we suspected that this could be related to fear of personal safety [14] . Emergency response from an epidemic prone disease from an exotic virulent virus or pathogen will naturally spark some level of fear and skepticism among any group of individuals especially when their knowledge about the dynamics of the disease outbreak is low. There are stories of HCWs who have avoided the responsibility of treating patients [15] and this was apparent in the HIV/AIDS and Severe Acute Respiratory Syndrome-Coronavirus (SARS-CoV) during the 1980s and 2003, respectively where the fear of contact with suspected and infected patients gripped some HCWs [16, 17] . In the long run, this fear would likely affect their confidence and commitment to professionalism.
The results of this study point to the fact that knowledge and the provision of tools such as personnel protective equipment (PPE) and other logistics alone is not good enough strategy. There might be the need to as well address issues related to myth, and culture as well as assurances of upkeep should one be infected. The general outlook one's country's devotion to their health staff might be a contributory factor in all of this and cannot be ignored. However, getting HCWs inspired and feel safe in caring for such highly infectious disease outbreaks is critical. During our study, HCWs indicated various forms of compensation to be paid to them should they be affected in the case of EVD attack.
This study had some inherent limitations. This was an exploratory study and our sample size was limited. Therefore, while not trying to generalize the results, we are of the opinion that this may be a reflection of HCWs in general. Additionally, since our study focused mainly on two health facilities, we are again careful in extrapolating these to other to reflect other facilities. Moreover, since this has not been a real experience, and a questionnaire-based survey, responses may not accurately reflect real-life experiences in the event of an EVD epidemic. Despite these limitations, the need for training was strong among HCWs. The results further demonstrate the ill-preparedness of health facilities, and the large proportion of HCWs unwillingness to attend to a suspected case of EVD. This thus calls for concerted efforts of health institutions and facilities to fully equip and prepare HCWs with the requisite tools and knowledge and ensuring competency to handle any epidemic prone disease. | What does the study suggest would make healthcare workers more willing to care for patients during an Ebola virus outbreak? | 4,254 | if HCWs are assured or guaranteed that they and or their families would be taken care of in case of death or while taking care of an EVD case, | 17,089 |
1,691 | RNA sequencing-based analysis of the spleen transcriptome following infectious bronchitis virus infection of chickens selected for different mannose-binding lectin serum concentrations
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4729133/
SHA: f5f1cd43740b5b6eca8b3cf2714fc0854a746519
Authors: Hamzić, Edin; Kjærup, Rikke Brødsgaard; Mach, Núria; Minozzi, Guilietta; Strozzi, Francesco; Gualdi, Valentina; Williams, John L.; Chen, Jun; Wattrang, Eva; Buitenhuis, Bart; Juul-Madsen, Helle Risdahl; Dalgaard, Tina Sørensen
Date: 2016-01-27
DOI: 10.1186/s12864-016-2403-1
License: cc-by
Abstract: BACKGROUND: Avian infectious bronchitis is a highly contagious disease of the upper-respiratory tract caused by infectious bronchitis virus (IBV). Understanding the molecular mechanisms involved in the interaction between innate and adaptive immune responses to IBV infection is a crucial element for further improvements in strategies to control IB. To this end, two chicken lines, selected for high (L10H line) and low (L10L line) serum concentration of mannose-binding lectin (MBL) were studied. In total, 32 birds from each line were used. Sixteen birds from each line were infected with IBV and sixteen were left uninfected. Eight uninfected and infected birds from each line were euthanized at 1 and 3 weeks post infection. RNA sequencing was performed on spleen samples from all 64 birds and differential gene expression analysis was performed for four comparisons: L10L line versus L10H line for uninfected birds at weeks 1 and 3, respectively, and in the same way for infected birds. Functional analysis was performed using Gene Ontology (GO) Immune System Process terms specific for Gallus gallus. RESULTS: Comparing uninfected L10H and L10L birds, we identified 1698 and 1424 differentially expressed (DE) genes at weeks 1 and 3, respectively. For the IBV-infected birds, 1934 and 866 DE genes were identified between the two lines at weeks 1 and 3, respectively. The two most enriched GO terms emerging from the comparison of uninfected birds between the two lines were “Lymphocyte activation involved in immune response” and “Somatic recombination of immunoglobulin genes involved in immune response” at weeks 1 and 3, respectively. When comparing IBV-infected birds between the two lines, the most enriched GO terms were “Alpha-beta T cell activation” and “Positive regulation of leukocyte activation” at weeks 1 and 3, respectively. CONCLUSIONS: Healthy birds from the two lines showed significant differences in expression profiles for subsets of adaptive and innate immunity-related genes, whereas comparison of the IBV-infected birds from the two lines showed differences in expression of immunity-related genes involved in T cell activation and proliferation. The observed transcriptome differences between the two lines indicate that selection for MBL had influenced innate as well as adaptive immunity. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (doi:10.1186/s12864-016-2403-1) contains supplementary material, which is available to authorized users.
Text: Conclusions: Healthy birds from the two lines showed significant differences in expression profiles for subsets of adaptive and innate immunity-related genes, whereas comparison of the IBV-infected birds from the two lines showed differences in expression of immunity-related genes involved in T cell activation and proliferation. The observed transcriptome differences between the two lines indicate that selection for MBL had influenced innate as well as adaptive immunity.
Keywords: IBV, Coronavirus, Infectious bronchitis, Chicken, RNA sequencing, Transcriptome, Spleen, Mannose-binding lectin, Immune response Background Avian infectious bronchitis (IB) is an acute and highly contagious disease of the upper-respiratory tract caused by the infectious bronchitis virus (IBV). The virus is a member of the Coronaviridae family and has numerous serotypes and strains. Rapid replication combined with high mutation rate and recombination are the main causes of the observed high diversity [1] . The respiratory tract is the primary target organ and entry point for the virus, before further spread to kidneys and gonads. The most common symptoms of IB are related to the respiratory tract and include gasping, coughing, sneezing, tracheal rales, and nasal discharge [2] . Feed conversion and average daily gain are affected in broilers, and infection is often followed by secondary bacterial infections. In layers, IBV causes a reduction in egg production and egg quality. Today, IB is one of the most economically important diseases in the poultry industry [2] . Infection outbreaks are controlled by a combination of strict management practices and vaccination. The strict management practices, which include the maintenance of the housing temperature and ventilation, are essential, because IBV is highly contagious and spreads very fast. Live attenuated and inactivated vaccines are widely used for control and prevention of IBV infection [3, 4] . As there is little or no cross-protection between different serotypes/variants of the virus, hence vaccines should contain serotypes present in a particular area in order to induce adequate protection [1] . New multi-strain vaccines with the optimal antigen combination and optimal adjuvants are therefore required for future IBV control. Understanding the molecular mechanisms involved in the interaction between innate and adaptive immune responses to IBV infection is a crucial element for further improvements of the vaccines.
IBV infection induces a wide range of immune responses in chickens. An innate immune response is activated during the initial stages of infection in the mucosal lining of the trachea following binding of IBV virions to receptors on epithelial cells [5] . Activation of this innate immune response may be initiated by Toll-like receptor (TLR) signaling upon IBV recognition [6, 7] . In addition, rapid activation of natural killer (NK) cells has been observed one day after IBV infection [8] as well as increased macrophage numbers in lungs and trachea after primary IBV infection [9] . In the case of the adaptive immune responses, T lymphocyte subpopulations are actively involved in the early stages of IBV clearance [7, 10] exhibiting rapid activation upon IBV infection [6] . Furthermore, studies have shown that cytotoxic T lymphocytes (CTL) play an important role in responding to primary infections with IBV [10, 11] . In addition to T cell responses, IBV specific antibodies, of all three antibody classes present in chickens, have been reported [12] [13] [14] . A specific local antibody response in avian infectious bronchitis is characteristic for the response to a secondary infection [15] . The innate and adaptive immune systems are strongly interconnected, which is also seen in the response to IBV infection, and the connection possibly involves the serum collectin, mannose-binding lectin (MBL) as a key player [16] .
Two chicken lines which were selected for high and low MBL serum concentrations (designated L10H and L10L, respectively), were used in the present study. Selective breeding has been performed for 14 generations using the combination of two strains (67.5 % UM-B19 chickens and 33.5 % White Cornish) as a starting population, as described by Juul-Madsen et al. [17] . The final result was two divergent lines, with mean MBL serum concentrations of 33.4 μg/ml for the L10H line and 7.6 μg/ml for the L10L line, respectively [18, 19] . The mean MBL serum concentration for 14 different chicken lines representing both broilers and layers is around 6 μg/ml, but varies from 0.4 to 37.8 μg/ml in normal healthy chickens with protein produced in the liver as the main source of circulating MBL [17] . In chickens, a positive correlation between MBL serum concentrations and the severity of several infections, such as infections caused by IBV [19] , Escherichia coli [20] and Pasteurella multocida [21] , has been observed. Chicken MBL binds to IBV [16, 22] , therefore it is possible that MBL facilitates innate responses such as opsono-phagocytosis, complement activation or virus neutralization, in the early stages of IBV infection. In mammals MBL has also been shown to influence induction of adaptive immunity [23] . In support of the role of MBL in response to IBV, Kjaerup et al. [18] observed considerable differences in cellular adaptive immune parameters in response to an IBV infection between lines L10L and L10H. Furthermore, birds from L10H line exhibited lower viral loads and less severe damage of tracheal cilia following the IBV infection in comparison to birds from the L10L line.
The aim of this study was to characterize the spleen transcriptome of healthy birds from the two lines selected for serum MBL, and to investigate differences in molecular mechanisms behind the development of systemic adaptive immunity between the L10L and L10H lines infected with IBV.
The experimental timeline and sampling time points are as illustrated in Fig. 1 and a full description of the experimental infection is reported by Kjaerup et al. [18] . The birds were infected at 3 weeks of age and from day 2 post-infection (p.i.), showed clinical signs characteristic of IBV infection, including sneezing and labored breathing. Viral loads in tracheal swabs were assessed for all birds as reported in the previous paper published on the experimental infection study [18] . No virus was detected in the uninfected birds at any time point throughout the experiment. Viral genomes were detected in swabs from infected birds from day 1 to 8 p.i. Notably, significantly lower viral loads (p < 0.03) were observed in birds from line L10H in comparison to infected birds from line L10L [18] .
Detection and quantification of splenic gene expression RNA sequencing data were produced from eight infected and eight uninfected birds from each of the two lines at two sampling occasions, as described in the materials and methods section. All samples passed quality control measures for raw and trimmed sequenced reads except for individual no. 46, which was removed due to a very low number of sequenced reads. For the remaining birds, an average of over 37 million reads were obtained per sample for the 63 samples analyzed, with 81 % of the reads mapping to the chicken genome reference sequence, as described in the materials and methods section (See summary statistics with the number of mapped and total reads is presented in Additional file 1: Table S1 ). In total, 17,113 expressed genes were identified. After filtering genes with fewer than one read per million in eight samples [24] (genes which would not achieve statistical significance for differential expression), the final list contained 11,292 expressed genes. Before performing the differential gene expression analysis, further multivariate analysis was carried out on the raw and normalized gene count data to identify any discrepancies.
Multi-dimensional scaling (MDS) plot on expressed genes between the two lines, L10H and L10L, showed that they differ considerably in their transcriptome profiles for both uninfected and IBV-infected birds [See Additional file 2: Figure S1 ]. Moreover, inter-individual variation in gene expression at week 1 was considerably higher than that observed at week 3 for both uninfected and IBV-infected birds [See Additional file 2: Figure S1 ].
Birds 22 and 47 were separated from the rest on the MDS plot [See Additional file 2: Figure S1 ]. However, inspection of raw sequence data and mapping parameters did not identify any technical problems which would explain the observed out-grouping of these birds. In addition an interclass principal component analysis (PCA) was performed using raw and normalized gene counts. The interclass PCA revealed that the birds 22 and 47 were placed outside the 95 % confidence intervals of their respective treatments [See Additional file 3: Figure S2 ]. However, the PCA did not identify any gene having extreme count profiles which may have contributed to the transcriptome dispersion of birds 22 and 47 with respect to their treatment groups. Although there was no clear technical or biological explanation for their out-grouping, these samples were removed from further analysis.
Differential gene expression analysis was performed to compare the two chicken lines (L10L and L10H) at two time points for uninfected (C1 and C2, see Fig. 1 ) and IBV-infected birds (C3 and C4, see Fig. 1 ). A large number of genes were differentially expressed (DE) between L10L and L10H lines at weeks 1 and 3, for both uninfected and IBV-infected birds (see Table 1 , see Fig. 1 ).
We identified 1,698 and 1,424 DE genes for the uninfected birds between lines L10L and L10H at weeks 1 and 3, respectively (see Table 1 ). In total 692 genes had higher expression in L10H line and 1,006 had higher expression in line L10L for the uninfected birds at week 1 [See Additional file 4: Table S2 ] and 774 genes had higher expression in L10H line and 650 genes had higher expression in L10L line for uninfected birds at week 3 [See Additional file 5: Table S3 ].
Comparing IBV-infected L10H and L10L birds, we identified 1,934 and 866 DE genes at weeks 1 and 3, respectively (see Table 1 ). In total 931 genes had higher expression in line L10H and 1,003 had higher expression in line L10L at week 1 and at week 3, 508 had higher expression in line L10H and 358 had higher expression in line L10L (Table 1 , Additional file 6: Table S4 and Additional file 7: Table S5 ).
There were also status-related changes in gene expression as shown in the Venn diagram ( Fig. 2) . At week 1, the total number of DE genes in uninfected birds Fig. 2 ). Out of 3,011 (1077 + 621 + 1313) DE genes for both uninfected and infected birds between the two lines only 621 (~20 %) were common for two comparisons (Fig. 2 ). At week 3, the total number of DE genes in uninfected birds between the two lines was 1424 (883 + 541) ( Table 1 , Fig. 2 ) which was higher comparing to 866 (541 + 325) in infected birds between the two lines ( Table 1 , Fig. 2 ). When comparing the uninfected and infected birds between the two lines, 541 (~30 %) genes were common out of total of 1749 (883 + 541+ 325) DE genes for both comparisons (Fig. 2) .
Moreover, we also performed differential gene expression analysis to compare two time points (week 1 and week 3) in the two chicken lines (L10L and L10H) for uninfected (C5 and C6, see Fig. 1 ) and IBV-infected birds (C7 and C8, see Fig. 1 ). Finally, differential gene expression analysis was also conducted to compare the two infection states (uninfected and IBV-infected) at two time points for the L10L chicken line (C9 and C10, see Table S6 , Additional file 9: Table S7 , Additional file 10: Table S8 , Additional file 11: Table S9 , Additional file 12: Table S10 , Additional file 13: Table S11 , Additional file 14: Table S12 and Additional file 15: Table S13 ].
An enrichment gene set analysis was carried out to identify over-represented Gene Ontology (GO) "Immune System Process" terms using the lists of DE genes from comparisons between uninfected and infected birds from the two lines at 1 and 3 weeks p.i. The most enriched GO Immune System terms between the two lines when comparing uninfected birds from the two lines and then infected from the two lines are shown in Fig. 3 .
GO Immune System terms associated with genes that were differentially expressed between the two lines for uninfected birds at week 1 were "Lymphocyte activation involved in immune response" (GO:0002285), "Activation of innate immune response" (GO:0002218), "Lymphocyte mediated immunity" (GO:0002449), and "Leukocyte Comparison between the two lines, L10H and L10L, uninfected and infected birds at two time points (weeks 1 and 3). Comparisons C1 -C4 correspond to differential gene expression comparisons presented in Fig. 1 For uninfected birds at week 3, the most enriched GO Immune System terms were "Somatic recombination of immunoglobulin genes involved in immune response" (GO:0002204) and the "Adaptive immune response based on somatic recombination of immune receptors built from immunoglobulin superfamily" (GO:0002460) [See Fig. 3 , See Additional file 17: Figure S4 ]. In total, 47 DE genes mapped to GO Immune System terms in this comparison (Fig. 4) . Among the DE genes that had a higher expression in the line L10H at week 3, in the uninfected group, were IL7, FKBP1B, FAS and PTPN22, which were also seen differentially expressed between lines at week 1 [See Fig. 4 , Additional file 17: Figure S4 ].
Comparing infected birds from the two lines, at week 1, "Alpha-beta T cell activation" (GO:0046631), "Activation of innate immune response" (GO:0002218) and "Leukocyte differentiation" (GO:0002521) functions were the three most enriched GO Immune System terms (Fig. 3) . CXCR4 (Chemokine receptor 4), PTPN22 and FAS were among the most highly expressed genes in L10H [See Fig. 4 , Additional file 18: Figure S5 ].
The major GO Immune System term that was strongly enriched for in the infected birds at week 3 was "Positive regulation of leukocyte activation" (GO:0002696) [See Figure S6 ].
The present study used two lines, L10L and L10H, which have been divergently selected for high and low MBL serum concentration for 14 generations, Fig. 3 Functional map of differentially expressed genes enriched for GO Immune System terms. The top categories of the GO Immune System terms associated with differentially expressed (DE) genes. All categories were statistically significant (adjusted p-value < 0.001). The chart fragments represent the number of genes associated with the terms as a proportion of the total number of genes within the respective GO term. Terms which have not been grouped are shown in grey respectively. These two lines have earlier been extensively used for immunological studies and exhibit differences in immunological parameters after being challenged with several pathogens [18, 22, 25] .
The spleen is a secondary lymphoid organ where innate and adaptive immune responses can be efficiently mounted. In addition, the avian spleen is considered to play a very important immunological role because avian lymphatic vessels and lymph nodes are poorly developed [26] . The transcriptome differences in the spleen between the two lines, L10L and L10H, for uninfected (healthy) and IBV-infected birds were investigated, focusing on the differential expression of immune-related genes within significantly enriched immune-related GO terms. Large differences in transcriptome profiles were observed between birds from the two lines, both uninfected (healthy) and following the experimental IBV challenge [See Additional file 2: Figure S1 ]. This suggests Fig. 4 Differentially expressed genes associated with the GO Immune System term. Heatmap representation of the differentially expressed (DE) genes associated with the GO Immune System terms for the four comparisons between the two lines, L10H and L10L, uninfected and infected groups at two time points, weeks 1 and 3. The heat map is constructed using the average values of counts per millions for each group that selection for MBL serum levels in the two lines had a much wider effect which goes beyond the expression of the MBL gene [27] . The observed transcriptome differences can probably be attributed to correlated response to selection [28] or random genetic drift [29] . Correlated selection occurs when a trait is affected by selection on a another trait and is dependent on a genetic correlation between the two traits, which is well known in animal breeding [30] . Alternatively, random genetic drift could contribute to the observed differences, considering that the founder population of the two lines was small [17] .
Focusing on the expression of immune-related genes: at week 1, the uninfected birds showed differences in the expression of genes involved in both adaptive immunity and innate immunity-related pathways [See Additional file 16: Figure S3 ]. In addition, the line L10H had a lower expression for the subset of the innate immune genes, TYRO3, TRAF3 and TLR7 at week 1 [See Additional file 16: Figure S3 and Fig. 4 ]. TYRO3 encodes tyrosineprotein kinase receptor 3 (TYRO3) which is involved in inhibition of TLR signaling pathways and TLR-induced cytokine signaling pathways. These two pathways influence immune-related processes, including cell proliferation/survival, cell adhesion and migration and inhibits the innate inflammatory response to pathogens [31] . TRAF3 encodes a cytoplasmic signaling protein, which plays a critical role in the regulation of antiviral response and viral evasion [32] [33] [34] . TLR7 was also among the DE innate immunity-related genes with higher expression in the line L10H. Chicken TLR7 has been shown to play a part in the response to IBV infections [9, 35] .
In addition to the differences in the expression profiles for a subset of innate immune genes, the two lines also differed in their expression of adaptive immune genes. Uninfected birds from L10H had a higher gene expression compared to the line L10L, for TGFB3, IL7, FKBP1B, FAS and PTPN22. These genes are known to be involved in a wide range of adaptive immune processes. In humans, reducing the TGF-β signaling on T cells has been shown to increase the function of CD8 T cells in an indirect way which results in the rapid elimination of viruses, enabling the creation of an effective memory response [36] . Similarly, human IL-7 plays a key role in the survival of both naïve [37] and memory [38, 39] CD4 and CD8 T cells. Moreover, an in vitro study showed that inhibition of FKBP1B and other cyclophilins blocked the replication of different Coronaviruses, including IBV [40] . In analogy, uninfected birds from the L10H line had a higher expression of IL7, FKBP1B, FAS and PTPN22.
Generally the uninfected (healthy) birds from the two lines exhibit different expression profiles for this subset of innate and adaptive immune genes probably resulting from the divergent selection for the MBL serum concentration. Selection in animal breeding have been shown to have an extensive effect on a variety of traits including immunological [30] . Moreover, correlated response to selection has been observed in the case where selection was performed on less complex traits such as testosterone levels [41] . Finally, different expression profiles for the subset of innate and adaptive immune in the uninfected birds from the two lines might be due to the balance in the effect of MBL serum levels. High levels of human MBL have been mostly reported as beneficial while in case of intracellular parasitic disease the effect of MBL serum level might be opposite [42] . The results indicate that selection for MBL serum levels might lead to favoring specific modes of immune responses depending on the MBL function.
Large differences in the expression patterns were seen between the two lines following infection with IBV and these differences involved adaptive immunity-related pathways which are associated with "Alpha-beta T cell activation" (GO:0046631), and "Activation of innate immune response" (GO:0002218) [See Additional file 18: Figure S5 ]. The observed enrichment for GO terms related to T cell activation is in accordance with a previous study of these lines that showed that the IBV-specific T cells are present in large numbers in the spleen after IBV infection [43] . At week 1 post infection CXCR4, FAS and PTPN22 showed higher expression in line L10H. The CXC chemokine receptors are expressed on both effector and memory T cells and play a key role in the homeostasis of memory T cells [44] . FAS has been shown to be upregulated in the kidney of chickens challenged with IBV [45] . The Fas/FasL pathway is an important pathway of killing for cytotoxic T cells [46] . Similarly, PTPN22 has been shown to be differentially expressed in chickens following pathogen challenge, and in particular following infection with Escherichia coli [47] .
Additionally VCAM1 and JMJD6 were among the adaptive immunity-related genes DE between lines at week 3 [See Fig. 4 , Additional file 19: Figure S6 ]. The VCAM1 gene is known to be involved in the activation of T cells [48] . Furthermore, a recent study demonstrated that JMJD6 regulates proliferation of memory T cells during a viral infection [49] which is of great interest considering that had higher expression in the IBVinfected birds from L10H line at week 3.
The results show that the two lines differ greatly in the expression of adaptive immunity-related genes following infection, which may imply the presence of different modes of gene regulation. The sampling times were chosen to access responses both in the effector phase (week 1) and memory phase (week 3) of the adaptive immune response to IBV. In accordance, at 1 week, post infection subsets of genes actively involved in T cell proliferation show differences between the lines. Also, at week 3 immune-related gene expression profiles in response to IBV infection that differ between the lines are more related to maintenance of T cell memory. MBL is known to be involved in regulation of dendritic cell maturation as well as cytokine production [50] . Dendritic cells, which are the main antigen presenting cells and are actively involved in regulation of adaptive immune responses, possess the receptors for MBL in mammals [23] . Therefore, the two lines selected for different MBL serum concentration may display differences in adaptive immune responses and development of adaptive immunity as a result of differences in response to cytokine signaling from dendritic cells. Other studies of the two lines, L10L and L10H, have shown that they differ in disease response parameters after being challenged with different pathogens. The L10L line has been associated with increased viral replication in the airway after an infectious bronchitis virus (IBV) infection [18] , reduced growth rate after an Escherichia coli infection [20] and greater intestinal colonization after Salmonella Infantis infection [51] . In the present experiment, significantly lower viral loads (p < 0.03) were observed in birds from line L10H in comparison to infected birds from line L10L [18] . Furthermore, L10H birds in the present study exhibited a less severe damage of tracheal cilia following the IBV infection in comparison to the L10L line (unpublished data). In the current experiment phenotypic differences in additional traits connected to adaptive immunity were observed, including numbers of circulating B cells and cytotoxic T cells [18] . Based on these observations it seems that selection for high MBL serum concentration allows birds to cope better after being infected with a range of pathogens. Therefore, the observed differences in the expression profiles for the adaptive and innate immune-related genes are a reflection of differences in disease resistance and immune responses between the lines L10L and L10H.
In conclusion, large differences in the spleen transcriptome between the two chicken lines, L10L and L10H, were observed in both uninfected (healthy) and IBVinfected birds. The uninfected birds from the two lines showed differences in expression profiles for a subset of both adaptive and innate immunity-related genes, which may represent differences in preparedness to respond to an infection. Following infection with IBV, the two lines showed large differences in expression of genes involved in the adaptive cellular immune response such as T cell activation and proliferation pathways and hence their ability to respond to the infection, which is reflected in the difference in pathogen load seen between the two lines.
This study is a follow-up of the experiment performed by Kjaerup et al. [18] which characterized the cellular and humoral immune response of the two chicken lines, L10H and L10L, divergently selected for MBL serum concentrations following IBV infection. In total, 96 birds were used in the experimental study originating from the two Aarhus University inbred lines, L10H and L10L [19] . All 96 birds were reared together in a biosecure IBV-free environment until they were 3 weeks of age and then allocated to two different groups with 24 birds from each line in each group (uninfected and infected). The birds were transferred to a biosafety level 2 facility and placed in isolators. Two isolators contained uninfected chickens and two isolators contained infected chickens. Each isolator having an equal mix of the two lines as described by Kjaerup et al. [18] .
The virulent IBV-M41 strain was used for the infection (a kind gift from Dr. med. vet. Hans C. Philipp at the Lohmann Animal Health GmbH, Cuxhaven, Germany). The virus had been passaged twice in specific pathogen-free embryonated eggs. The IBV inocula were prepared in phosphate-buffered saline (PBS) immediately before use and contained 2 × 10 5.2 EID 50 /200 μl of IBV-M41 virus. The first and the second group (the uninfected groups) were mock-infected with 200 μl PBS per bird. The third and fourth groups (the infected groups) received 200 μL of IBV-M41. The inocula were given half nasally and half orally to mimic the natural infection routes of IBV in the chicken. Chickens were fed diets that met or exceeded the National Research Council requirements. Feed and water were provided ad libitum. The birds were monitored daily for clinical signs of disease and disease parameters were measured as reported by Kjaerup et al. [18] . None of the individuals received antibiotic therapy during the experimental period. The study was carried out under strict ethical approval and monitoring (see the statement at the end of the Materials and Methods section).
For this study 64 spleen samples were harvested and used for RNA sequencing. The birds were sacrificed 1 and 3 weeks post infection by cervical dislocation and spleen samples were collected. At both time points, eight samples from the two lines, L10H and line L10L, from each group (uninfected and infected) were collected as illustrated in Fig. 1 . After collection, spleens were sectioned (triangular cross-sectional slice from upper part) and identical samples from each chicken were immediately placed in RNAlater® Stabilization Solution (Ambion Inc., Austin, Texas) that were incubated at 4°C overnight and then transferred to -20°C the following day.
Tissue samples were homogenized on a TissueLyzer LT (Qiagen, Hilden, Germany). Total RNA was extracted with the Qiagen RNAeasy Kit (Catalog ID 74104, Qiagen, Venlo, Netherlands) according to the manufacturer's instructions. The quality of the 64 total RNA samples was verified using a 2200 TapeStation RNA Screen Tape device (Agilent, Santa Clara, CA, USA) and the concentration ascertained using an ND-1000 spectrophotometer (Nano-Drop, Wilmington, DE).
Libraries were prepared with the Illumina TruseqRNA sample prep kit (Catalog ID FC-122-1001, Illumina, San Diego, USA) following the manufacturer's protocol and evaluated with the Agilent Tape Station 2200. Libraries were quantified by Picogreen and then normalized to 10 nM as recommended by Illumina for cluster generation on the Hiseq2000. Equimolar amounts of each library were mixed before NaOH denaturation.
The Illumina Truseq PE cluster kit v3 (Catalog ID PE-401-3001) was used to generate clusters on the grafted Illumina Flowcell and the hybridized libraries were sequenced on six lines of a Flowcell on the Hiseq2000 with 100 cycles of a paired-end sequencing module using the Truseq SBS kit v3 (Catalog ID FC-401-3001).
Quality control, mapping of RNA sequencing reads and counting mapped reads Initial control quality was assessed by the FastQC software version 0.11.3 [52] . Raw reads were than trimmed for low quality bases using the Trimmomatic tool version 0.32 [53] applying minimum Phred quality score >10 averaged across the sliding window of five bases. Furthermore, all reads with the length below 40 bp were removed.
The trimmed reads were mapped to the Gallus gallus reference genome (Gallus_gallus-4.0, release 80 [54] ) using a spliced aligner TopHat2 version 2.014 [55] . The Gallus gallus gene annotation used for mapping was retrieved from Ensembl database version 80 (www.ensembl.org). The mapping quality was assessed using a set of Python scripts within the RSeQC toolkit [56] . The quality control assessment included inspection of the read coverage over the full gene body in order to assess if reads coverage was uniform and if there was any 5' or 3' bias as well as how the mapped reads were distributed over genome features.
Gene count estimation was performed using the HTSeq-count tool in 'union' mode. The HTSeq-count is a Python script within the HTSeq framework, version 0.7.1, which is an open source toolkit that allows the input of raw counts from aligned reads to be annotated with gene names based on genomic features [57] .
The read counts obtained were used to estimate gene expression and identify differentially expressed (DE) genes. This was achieved using Bioconductor package edgeR version 3.10.0 [58] and limma version 3.24.5 [59] following a previously described protocol [24] . Before performing statistical analysis, genes with low levels of expression were filtered out using a threshold of least one read per million in n of the samples, where n is the size of the smallest group of replicates, which in this case was eight.
In order to account for technical and biological effects reads counts were normalized using the "calc-NormFactors" function implemented in the edgeR package. This function normalizes the data by finding a set of scaling factors for the library sizes that minimizes the log-fold changes between the samples. The scale factors were computed using the trimmed mean of M-values (TMM) between samples [58] . Common and tag-wise dispersion estimates were calculated with the Cox-Reid profile adjusted likelihood method in order to correct for the technical and biological variation when fitting the multivariate negative binomial model [60] .
Multidimensional scaling (MDS) was implemented in the edgeR package, to assess similarity of the samples visually. The MDS plot was created in order to visualize the relationship between samples and identify possible outliers [58] . MDS is based on comparing the relationship between all pairs of samples by applying a countspecific pairwise distance measure [58] . Possible outliers were further investigated using principal component analysis (PCA) to remove samples which fell outside a 95 % confidence ellipse.
A design matrix was created in order to specify the factors that were expected to affect the expression level. The matrix was constructed to fit the saturated model where each treatment combination was considered separately. Eight treatment combinations were considered as illustrated in Fig. 1 : uninfected birds from the line L10H at week 1, uninfected birds from the line L10L at week 1, infected birds from the line L10H at week 1, infected birds from the line L10L at week 1, uninfected birds from the line L10H at week 3, uninfected birds from the line L10L at week 3, infected birds from the line L10H at week 3, infected birds from the line L10L at week 3.
A generalized linear model likelihood ratio test, specifying the difference of interest, was used to test for differential expression between these treatment combinations. The differential expression analysis was performed comparing the log-fold differences in gene counts between two lines (L10H and L10L) at different time points (weeks 1 and 3) and for different status (uninfected and infected) separately (Fig. 1 ). Benjamini Hochberg false discovery rates (FDR) for a transcriptome-wide experiment were calculated to correct for multiple testing [61] . All genes with an FDR-adjusted p-value <0.05 were considered individual genes of interest and were retained for further analysis.
Functional analysis of the DE genes was performed using the Cytoscape version 3.2.1 [62, 63] with the ClueGo version 2.1.7 plug-in [64] to enrich the annotation and enrichment of the differentially expressed (DE) genes for four comparisons (C1-C4, see Fig. 1 ). ClueGO determines the distribution of the target genes across the GO (Gene Ontology) terms and pathways: this study focused on. The p-value was calculated using right-sided hypergeometric tests and Benjamini-Hochberg adjustment was used for multiple test correction. An adjusted pvalue of 0.001 indicated a statistically significant deviation from the expected distribution, and that the corresponding GO terms and pathways were enriched for the target genes. The association strength between the terms was calculated using a corrected kappa statistic of 0.4. The network created represented the terms as nodes which were linked based on a 0.4 kappa score level. The size of the nodes reflected the enrichment significance of the terms. The network was automatically laid out using the Organic layout algorithm supported by Cytoscape. The functional groups were created by iterative merging of initially defined groups based on the predefined kappa score threshold. Only functional groups represented by their most significant term were visualized in the network providing an insightful view of their interrelations [64] .
The experimental procedures were conducted under the protocols approved by the Danish Animal Experiments Inspectorate and complied with the Danish Ministry of Justice Law no. 382 (June 10, 1987) and Acts 739 (December 6, 1988) and 333 (May 19, 1990) concerning animal experimentation and care of experimental animals. The license to conduct the animal experiment was obtained by Helle R. Juul-Madsen. comparison between uninfected birds at week 1. The GO Immune System terms were identified as nodes and linked based on their kappa score level (> = 0.4) and p-value < 0.001. Functionally related groups partially overlapped. The GO terms are labelled in colors according to hierarchical clustering of GO terms. Terms which have not been grouped are shown in grey. The colour pie charts of the GO Immune system nodes show the gene proportion associated with the respective term. (PDF 60 kb)
Additional file 17: Figure S4 . Network representation of enriched GO Immune System terms of differentially expressed (DE) genes for comparison between uninfected birds at week 3. The GO Immune System terms were identified as nodes and linked based on their kappa score level (> = 0.4) and p-value < 0.001. Functionally related groups partially overlapped. The GO terms are labelled in colours according to hierarchical clustering of GO terms. Terms which have not been grouped are shown in grey. The color pie charts of the GO Immune system nodes show the gene proportion associated with the respective term. (PDF 55 kb) Additional file 18: Figure S5 . Network representation of enriched GO Immune System terms of differentially expressed (DE) genes for comparison between infected birds at week 1. The GO Immune System terms were identified as nodes and linked based on their kappa score level (> = 0.4) and p-value < 0.001. Functionally related groups partially overlapped. The GO terms are labelled in colors according to hierarchical clustering of GO terms. Terms which have not been grouped are shown in grey. The color pie charts of the GO Immune system nodes show the gene proportion associated with the respective term. (PDF 70 kb) Additional file 19: Figure S6 . Network representation of enriched GO Immune System terms of differentially expressed (DE) genes for comparison between infected birds at week 3. The GO Immune System terms were identified as nodes and linked based on their kappa score level (> = 0.4) and p-value < 0.001. Functionally related groups partially overlapped. The GO terms are labelled in colors according to hierarchical clustering of GO terms. Terms which have not been grouped are shown in grey. The color pie charts of the GO Immune system nodes show the gene proportion associated with the respective term. (PDF 30 kb) | What causes avian infectious bronchitis? | 5,160 | infectious bronchitis virus (IBV) | 710 |
1,691 | RNA sequencing-based analysis of the spleen transcriptome following infectious bronchitis virus infection of chickens selected for different mannose-binding lectin serum concentrations
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4729133/
SHA: f5f1cd43740b5b6eca8b3cf2714fc0854a746519
Authors: Hamzić, Edin; Kjærup, Rikke Brødsgaard; Mach, Núria; Minozzi, Guilietta; Strozzi, Francesco; Gualdi, Valentina; Williams, John L.; Chen, Jun; Wattrang, Eva; Buitenhuis, Bart; Juul-Madsen, Helle Risdahl; Dalgaard, Tina Sørensen
Date: 2016-01-27
DOI: 10.1186/s12864-016-2403-1
License: cc-by
Abstract: BACKGROUND: Avian infectious bronchitis is a highly contagious disease of the upper-respiratory tract caused by infectious bronchitis virus (IBV). Understanding the molecular mechanisms involved in the interaction between innate and adaptive immune responses to IBV infection is a crucial element for further improvements in strategies to control IB. To this end, two chicken lines, selected for high (L10H line) and low (L10L line) serum concentration of mannose-binding lectin (MBL) were studied. In total, 32 birds from each line were used. Sixteen birds from each line were infected with IBV and sixteen were left uninfected. Eight uninfected and infected birds from each line were euthanized at 1 and 3 weeks post infection. RNA sequencing was performed on spleen samples from all 64 birds and differential gene expression analysis was performed for four comparisons: L10L line versus L10H line for uninfected birds at weeks 1 and 3, respectively, and in the same way for infected birds. Functional analysis was performed using Gene Ontology (GO) Immune System Process terms specific for Gallus gallus. RESULTS: Comparing uninfected L10H and L10L birds, we identified 1698 and 1424 differentially expressed (DE) genes at weeks 1 and 3, respectively. For the IBV-infected birds, 1934 and 866 DE genes were identified between the two lines at weeks 1 and 3, respectively. The two most enriched GO terms emerging from the comparison of uninfected birds between the two lines were “Lymphocyte activation involved in immune response” and “Somatic recombination of immunoglobulin genes involved in immune response” at weeks 1 and 3, respectively. When comparing IBV-infected birds between the two lines, the most enriched GO terms were “Alpha-beta T cell activation” and “Positive regulation of leukocyte activation” at weeks 1 and 3, respectively. CONCLUSIONS: Healthy birds from the two lines showed significant differences in expression profiles for subsets of adaptive and innate immunity-related genes, whereas comparison of the IBV-infected birds from the two lines showed differences in expression of immunity-related genes involved in T cell activation and proliferation. The observed transcriptome differences between the two lines indicate that selection for MBL had influenced innate as well as adaptive immunity. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (doi:10.1186/s12864-016-2403-1) contains supplementary material, which is available to authorized users.
Text: Conclusions: Healthy birds from the two lines showed significant differences in expression profiles for subsets of adaptive and innate immunity-related genes, whereas comparison of the IBV-infected birds from the two lines showed differences in expression of immunity-related genes involved in T cell activation and proliferation. The observed transcriptome differences between the two lines indicate that selection for MBL had influenced innate as well as adaptive immunity.
Keywords: IBV, Coronavirus, Infectious bronchitis, Chicken, RNA sequencing, Transcriptome, Spleen, Mannose-binding lectin, Immune response Background Avian infectious bronchitis (IB) is an acute and highly contagious disease of the upper-respiratory tract caused by the infectious bronchitis virus (IBV). The virus is a member of the Coronaviridae family and has numerous serotypes and strains. Rapid replication combined with high mutation rate and recombination are the main causes of the observed high diversity [1] . The respiratory tract is the primary target organ and entry point for the virus, before further spread to kidneys and gonads. The most common symptoms of IB are related to the respiratory tract and include gasping, coughing, sneezing, tracheal rales, and nasal discharge [2] . Feed conversion and average daily gain are affected in broilers, and infection is often followed by secondary bacterial infections. In layers, IBV causes a reduction in egg production and egg quality. Today, IB is one of the most economically important diseases in the poultry industry [2] . Infection outbreaks are controlled by a combination of strict management practices and vaccination. The strict management practices, which include the maintenance of the housing temperature and ventilation, are essential, because IBV is highly contagious and spreads very fast. Live attenuated and inactivated vaccines are widely used for control and prevention of IBV infection [3, 4] . As there is little or no cross-protection between different serotypes/variants of the virus, hence vaccines should contain serotypes present in a particular area in order to induce adequate protection [1] . New multi-strain vaccines with the optimal antigen combination and optimal adjuvants are therefore required for future IBV control. Understanding the molecular mechanisms involved in the interaction between innate and adaptive immune responses to IBV infection is a crucial element for further improvements of the vaccines.
IBV infection induces a wide range of immune responses in chickens. An innate immune response is activated during the initial stages of infection in the mucosal lining of the trachea following binding of IBV virions to receptors on epithelial cells [5] . Activation of this innate immune response may be initiated by Toll-like receptor (TLR) signaling upon IBV recognition [6, 7] . In addition, rapid activation of natural killer (NK) cells has been observed one day after IBV infection [8] as well as increased macrophage numbers in lungs and trachea after primary IBV infection [9] . In the case of the adaptive immune responses, T lymphocyte subpopulations are actively involved in the early stages of IBV clearance [7, 10] exhibiting rapid activation upon IBV infection [6] . Furthermore, studies have shown that cytotoxic T lymphocytes (CTL) play an important role in responding to primary infections with IBV [10, 11] . In addition to T cell responses, IBV specific antibodies, of all three antibody classes present in chickens, have been reported [12] [13] [14] . A specific local antibody response in avian infectious bronchitis is characteristic for the response to a secondary infection [15] . The innate and adaptive immune systems are strongly interconnected, which is also seen in the response to IBV infection, and the connection possibly involves the serum collectin, mannose-binding lectin (MBL) as a key player [16] .
Two chicken lines which were selected for high and low MBL serum concentrations (designated L10H and L10L, respectively), were used in the present study. Selective breeding has been performed for 14 generations using the combination of two strains (67.5 % UM-B19 chickens and 33.5 % White Cornish) as a starting population, as described by Juul-Madsen et al. [17] . The final result was two divergent lines, with mean MBL serum concentrations of 33.4 μg/ml for the L10H line and 7.6 μg/ml for the L10L line, respectively [18, 19] . The mean MBL serum concentration for 14 different chicken lines representing both broilers and layers is around 6 μg/ml, but varies from 0.4 to 37.8 μg/ml in normal healthy chickens with protein produced in the liver as the main source of circulating MBL [17] . In chickens, a positive correlation between MBL serum concentrations and the severity of several infections, such as infections caused by IBV [19] , Escherichia coli [20] and Pasteurella multocida [21] , has been observed. Chicken MBL binds to IBV [16, 22] , therefore it is possible that MBL facilitates innate responses such as opsono-phagocytosis, complement activation or virus neutralization, in the early stages of IBV infection. In mammals MBL has also been shown to influence induction of adaptive immunity [23] . In support of the role of MBL in response to IBV, Kjaerup et al. [18] observed considerable differences in cellular adaptive immune parameters in response to an IBV infection between lines L10L and L10H. Furthermore, birds from L10H line exhibited lower viral loads and less severe damage of tracheal cilia following the IBV infection in comparison to birds from the L10L line.
The aim of this study was to characterize the spleen transcriptome of healthy birds from the two lines selected for serum MBL, and to investigate differences in molecular mechanisms behind the development of systemic adaptive immunity between the L10L and L10H lines infected with IBV.
The experimental timeline and sampling time points are as illustrated in Fig. 1 and a full description of the experimental infection is reported by Kjaerup et al. [18] . The birds were infected at 3 weeks of age and from day 2 post-infection (p.i.), showed clinical signs characteristic of IBV infection, including sneezing and labored breathing. Viral loads in tracheal swabs were assessed for all birds as reported in the previous paper published on the experimental infection study [18] . No virus was detected in the uninfected birds at any time point throughout the experiment. Viral genomes were detected in swabs from infected birds from day 1 to 8 p.i. Notably, significantly lower viral loads (p < 0.03) were observed in birds from line L10H in comparison to infected birds from line L10L [18] .
Detection and quantification of splenic gene expression RNA sequencing data were produced from eight infected and eight uninfected birds from each of the two lines at two sampling occasions, as described in the materials and methods section. All samples passed quality control measures for raw and trimmed sequenced reads except for individual no. 46, which was removed due to a very low number of sequenced reads. For the remaining birds, an average of over 37 million reads were obtained per sample for the 63 samples analyzed, with 81 % of the reads mapping to the chicken genome reference sequence, as described in the materials and methods section (See summary statistics with the number of mapped and total reads is presented in Additional file 1: Table S1 ). In total, 17,113 expressed genes were identified. After filtering genes with fewer than one read per million in eight samples [24] (genes which would not achieve statistical significance for differential expression), the final list contained 11,292 expressed genes. Before performing the differential gene expression analysis, further multivariate analysis was carried out on the raw and normalized gene count data to identify any discrepancies.
Multi-dimensional scaling (MDS) plot on expressed genes between the two lines, L10H and L10L, showed that they differ considerably in their transcriptome profiles for both uninfected and IBV-infected birds [See Additional file 2: Figure S1 ]. Moreover, inter-individual variation in gene expression at week 1 was considerably higher than that observed at week 3 for both uninfected and IBV-infected birds [See Additional file 2: Figure S1 ].
Birds 22 and 47 were separated from the rest on the MDS plot [See Additional file 2: Figure S1 ]. However, inspection of raw sequence data and mapping parameters did not identify any technical problems which would explain the observed out-grouping of these birds. In addition an interclass principal component analysis (PCA) was performed using raw and normalized gene counts. The interclass PCA revealed that the birds 22 and 47 were placed outside the 95 % confidence intervals of their respective treatments [See Additional file 3: Figure S2 ]. However, the PCA did not identify any gene having extreme count profiles which may have contributed to the transcriptome dispersion of birds 22 and 47 with respect to their treatment groups. Although there was no clear technical or biological explanation for their out-grouping, these samples were removed from further analysis.
Differential gene expression analysis was performed to compare the two chicken lines (L10L and L10H) at two time points for uninfected (C1 and C2, see Fig. 1 ) and IBV-infected birds (C3 and C4, see Fig. 1 ). A large number of genes were differentially expressed (DE) between L10L and L10H lines at weeks 1 and 3, for both uninfected and IBV-infected birds (see Table 1 , see Fig. 1 ).
We identified 1,698 and 1,424 DE genes for the uninfected birds between lines L10L and L10H at weeks 1 and 3, respectively (see Table 1 ). In total 692 genes had higher expression in L10H line and 1,006 had higher expression in line L10L for the uninfected birds at week 1 [See Additional file 4: Table S2 ] and 774 genes had higher expression in L10H line and 650 genes had higher expression in L10L line for uninfected birds at week 3 [See Additional file 5: Table S3 ].
Comparing IBV-infected L10H and L10L birds, we identified 1,934 and 866 DE genes at weeks 1 and 3, respectively (see Table 1 ). In total 931 genes had higher expression in line L10H and 1,003 had higher expression in line L10L at week 1 and at week 3, 508 had higher expression in line L10H and 358 had higher expression in line L10L (Table 1 , Additional file 6: Table S4 and Additional file 7: Table S5 ).
There were also status-related changes in gene expression as shown in the Venn diagram ( Fig. 2) . At week 1, the total number of DE genes in uninfected birds Fig. 2 ). Out of 3,011 (1077 + 621 + 1313) DE genes for both uninfected and infected birds between the two lines only 621 (~20 %) were common for two comparisons (Fig. 2 ). At week 3, the total number of DE genes in uninfected birds between the two lines was 1424 (883 + 541) ( Table 1 , Fig. 2 ) which was higher comparing to 866 (541 + 325) in infected birds between the two lines ( Table 1 , Fig. 2 ). When comparing the uninfected and infected birds between the two lines, 541 (~30 %) genes were common out of total of 1749 (883 + 541+ 325) DE genes for both comparisons (Fig. 2) .
Moreover, we also performed differential gene expression analysis to compare two time points (week 1 and week 3) in the two chicken lines (L10L and L10H) for uninfected (C5 and C6, see Fig. 1 ) and IBV-infected birds (C7 and C8, see Fig. 1 ). Finally, differential gene expression analysis was also conducted to compare the two infection states (uninfected and IBV-infected) at two time points for the L10L chicken line (C9 and C10, see Table S6 , Additional file 9: Table S7 , Additional file 10: Table S8 , Additional file 11: Table S9 , Additional file 12: Table S10 , Additional file 13: Table S11 , Additional file 14: Table S12 and Additional file 15: Table S13 ].
An enrichment gene set analysis was carried out to identify over-represented Gene Ontology (GO) "Immune System Process" terms using the lists of DE genes from comparisons between uninfected and infected birds from the two lines at 1 and 3 weeks p.i. The most enriched GO Immune System terms between the two lines when comparing uninfected birds from the two lines and then infected from the two lines are shown in Fig. 3 .
GO Immune System terms associated with genes that were differentially expressed between the two lines for uninfected birds at week 1 were "Lymphocyte activation involved in immune response" (GO:0002285), "Activation of innate immune response" (GO:0002218), "Lymphocyte mediated immunity" (GO:0002449), and "Leukocyte Comparison between the two lines, L10H and L10L, uninfected and infected birds at two time points (weeks 1 and 3). Comparisons C1 -C4 correspond to differential gene expression comparisons presented in Fig. 1 For uninfected birds at week 3, the most enriched GO Immune System terms were "Somatic recombination of immunoglobulin genes involved in immune response" (GO:0002204) and the "Adaptive immune response based on somatic recombination of immune receptors built from immunoglobulin superfamily" (GO:0002460) [See Fig. 3 , See Additional file 17: Figure S4 ]. In total, 47 DE genes mapped to GO Immune System terms in this comparison (Fig. 4) . Among the DE genes that had a higher expression in the line L10H at week 3, in the uninfected group, were IL7, FKBP1B, FAS and PTPN22, which were also seen differentially expressed between lines at week 1 [See Fig. 4 , Additional file 17: Figure S4 ].
Comparing infected birds from the two lines, at week 1, "Alpha-beta T cell activation" (GO:0046631), "Activation of innate immune response" (GO:0002218) and "Leukocyte differentiation" (GO:0002521) functions were the three most enriched GO Immune System terms (Fig. 3) . CXCR4 (Chemokine receptor 4), PTPN22 and FAS were among the most highly expressed genes in L10H [See Fig. 4 , Additional file 18: Figure S5 ].
The major GO Immune System term that was strongly enriched for in the infected birds at week 3 was "Positive regulation of leukocyte activation" (GO:0002696) [See Figure S6 ].
The present study used two lines, L10L and L10H, which have been divergently selected for high and low MBL serum concentration for 14 generations, Fig. 3 Functional map of differentially expressed genes enriched for GO Immune System terms. The top categories of the GO Immune System terms associated with differentially expressed (DE) genes. All categories were statistically significant (adjusted p-value < 0.001). The chart fragments represent the number of genes associated with the terms as a proportion of the total number of genes within the respective GO term. Terms which have not been grouped are shown in grey respectively. These two lines have earlier been extensively used for immunological studies and exhibit differences in immunological parameters after being challenged with several pathogens [18, 22, 25] .
The spleen is a secondary lymphoid organ where innate and adaptive immune responses can be efficiently mounted. In addition, the avian spleen is considered to play a very important immunological role because avian lymphatic vessels and lymph nodes are poorly developed [26] . The transcriptome differences in the spleen between the two lines, L10L and L10H, for uninfected (healthy) and IBV-infected birds were investigated, focusing on the differential expression of immune-related genes within significantly enriched immune-related GO terms. Large differences in transcriptome profiles were observed between birds from the two lines, both uninfected (healthy) and following the experimental IBV challenge [See Additional file 2: Figure S1 ]. This suggests Fig. 4 Differentially expressed genes associated with the GO Immune System term. Heatmap representation of the differentially expressed (DE) genes associated with the GO Immune System terms for the four comparisons between the two lines, L10H and L10L, uninfected and infected groups at two time points, weeks 1 and 3. The heat map is constructed using the average values of counts per millions for each group that selection for MBL serum levels in the two lines had a much wider effect which goes beyond the expression of the MBL gene [27] . The observed transcriptome differences can probably be attributed to correlated response to selection [28] or random genetic drift [29] . Correlated selection occurs when a trait is affected by selection on a another trait and is dependent on a genetic correlation between the two traits, which is well known in animal breeding [30] . Alternatively, random genetic drift could contribute to the observed differences, considering that the founder population of the two lines was small [17] .
Focusing on the expression of immune-related genes: at week 1, the uninfected birds showed differences in the expression of genes involved in both adaptive immunity and innate immunity-related pathways [See Additional file 16: Figure S3 ]. In addition, the line L10H had a lower expression for the subset of the innate immune genes, TYRO3, TRAF3 and TLR7 at week 1 [See Additional file 16: Figure S3 and Fig. 4 ]. TYRO3 encodes tyrosineprotein kinase receptor 3 (TYRO3) which is involved in inhibition of TLR signaling pathways and TLR-induced cytokine signaling pathways. These two pathways influence immune-related processes, including cell proliferation/survival, cell adhesion and migration and inhibits the innate inflammatory response to pathogens [31] . TRAF3 encodes a cytoplasmic signaling protein, which plays a critical role in the regulation of antiviral response and viral evasion [32] [33] [34] . TLR7 was also among the DE innate immunity-related genes with higher expression in the line L10H. Chicken TLR7 has been shown to play a part in the response to IBV infections [9, 35] .
In addition to the differences in the expression profiles for a subset of innate immune genes, the two lines also differed in their expression of adaptive immune genes. Uninfected birds from L10H had a higher gene expression compared to the line L10L, for TGFB3, IL7, FKBP1B, FAS and PTPN22. These genes are known to be involved in a wide range of adaptive immune processes. In humans, reducing the TGF-β signaling on T cells has been shown to increase the function of CD8 T cells in an indirect way which results in the rapid elimination of viruses, enabling the creation of an effective memory response [36] . Similarly, human IL-7 plays a key role in the survival of both naïve [37] and memory [38, 39] CD4 and CD8 T cells. Moreover, an in vitro study showed that inhibition of FKBP1B and other cyclophilins blocked the replication of different Coronaviruses, including IBV [40] . In analogy, uninfected birds from the L10H line had a higher expression of IL7, FKBP1B, FAS and PTPN22.
Generally the uninfected (healthy) birds from the two lines exhibit different expression profiles for this subset of innate and adaptive immune genes probably resulting from the divergent selection for the MBL serum concentration. Selection in animal breeding have been shown to have an extensive effect on a variety of traits including immunological [30] . Moreover, correlated response to selection has been observed in the case where selection was performed on less complex traits such as testosterone levels [41] . Finally, different expression profiles for the subset of innate and adaptive immune in the uninfected birds from the two lines might be due to the balance in the effect of MBL serum levels. High levels of human MBL have been mostly reported as beneficial while in case of intracellular parasitic disease the effect of MBL serum level might be opposite [42] . The results indicate that selection for MBL serum levels might lead to favoring specific modes of immune responses depending on the MBL function.
Large differences in the expression patterns were seen between the two lines following infection with IBV and these differences involved adaptive immunity-related pathways which are associated with "Alpha-beta T cell activation" (GO:0046631), and "Activation of innate immune response" (GO:0002218) [See Additional file 18: Figure S5 ]. The observed enrichment for GO terms related to T cell activation is in accordance with a previous study of these lines that showed that the IBV-specific T cells are present in large numbers in the spleen after IBV infection [43] . At week 1 post infection CXCR4, FAS and PTPN22 showed higher expression in line L10H. The CXC chemokine receptors are expressed on both effector and memory T cells and play a key role in the homeostasis of memory T cells [44] . FAS has been shown to be upregulated in the kidney of chickens challenged with IBV [45] . The Fas/FasL pathway is an important pathway of killing for cytotoxic T cells [46] . Similarly, PTPN22 has been shown to be differentially expressed in chickens following pathogen challenge, and in particular following infection with Escherichia coli [47] .
Additionally VCAM1 and JMJD6 were among the adaptive immunity-related genes DE between lines at week 3 [See Fig. 4 , Additional file 19: Figure S6 ]. The VCAM1 gene is known to be involved in the activation of T cells [48] . Furthermore, a recent study demonstrated that JMJD6 regulates proliferation of memory T cells during a viral infection [49] which is of great interest considering that had higher expression in the IBVinfected birds from L10H line at week 3.
The results show that the two lines differ greatly in the expression of adaptive immunity-related genes following infection, which may imply the presence of different modes of gene regulation. The sampling times were chosen to access responses both in the effector phase (week 1) and memory phase (week 3) of the adaptive immune response to IBV. In accordance, at 1 week, post infection subsets of genes actively involved in T cell proliferation show differences between the lines. Also, at week 3 immune-related gene expression profiles in response to IBV infection that differ between the lines are more related to maintenance of T cell memory. MBL is known to be involved in regulation of dendritic cell maturation as well as cytokine production [50] . Dendritic cells, which are the main antigen presenting cells and are actively involved in regulation of adaptive immune responses, possess the receptors for MBL in mammals [23] . Therefore, the two lines selected for different MBL serum concentration may display differences in adaptive immune responses and development of adaptive immunity as a result of differences in response to cytokine signaling from dendritic cells. Other studies of the two lines, L10L and L10H, have shown that they differ in disease response parameters after being challenged with different pathogens. The L10L line has been associated with increased viral replication in the airway after an infectious bronchitis virus (IBV) infection [18] , reduced growth rate after an Escherichia coli infection [20] and greater intestinal colonization after Salmonella Infantis infection [51] . In the present experiment, significantly lower viral loads (p < 0.03) were observed in birds from line L10H in comparison to infected birds from line L10L [18] . Furthermore, L10H birds in the present study exhibited a less severe damage of tracheal cilia following the IBV infection in comparison to the L10L line (unpublished data). In the current experiment phenotypic differences in additional traits connected to adaptive immunity were observed, including numbers of circulating B cells and cytotoxic T cells [18] . Based on these observations it seems that selection for high MBL serum concentration allows birds to cope better after being infected with a range of pathogens. Therefore, the observed differences in the expression profiles for the adaptive and innate immune-related genes are a reflection of differences in disease resistance and immune responses between the lines L10L and L10H.
In conclusion, large differences in the spleen transcriptome between the two chicken lines, L10L and L10H, were observed in both uninfected (healthy) and IBVinfected birds. The uninfected birds from the two lines showed differences in expression profiles for a subset of both adaptive and innate immunity-related genes, which may represent differences in preparedness to respond to an infection. Following infection with IBV, the two lines showed large differences in expression of genes involved in the adaptive cellular immune response such as T cell activation and proliferation pathways and hence their ability to respond to the infection, which is reflected in the difference in pathogen load seen between the two lines.
This study is a follow-up of the experiment performed by Kjaerup et al. [18] which characterized the cellular and humoral immune response of the two chicken lines, L10H and L10L, divergently selected for MBL serum concentrations following IBV infection. In total, 96 birds were used in the experimental study originating from the two Aarhus University inbred lines, L10H and L10L [19] . All 96 birds were reared together in a biosecure IBV-free environment until they were 3 weeks of age and then allocated to two different groups with 24 birds from each line in each group (uninfected and infected). The birds were transferred to a biosafety level 2 facility and placed in isolators. Two isolators contained uninfected chickens and two isolators contained infected chickens. Each isolator having an equal mix of the two lines as described by Kjaerup et al. [18] .
The virulent IBV-M41 strain was used for the infection (a kind gift from Dr. med. vet. Hans C. Philipp at the Lohmann Animal Health GmbH, Cuxhaven, Germany). The virus had been passaged twice in specific pathogen-free embryonated eggs. The IBV inocula were prepared in phosphate-buffered saline (PBS) immediately before use and contained 2 × 10 5.2 EID 50 /200 μl of IBV-M41 virus. The first and the second group (the uninfected groups) were mock-infected with 200 μl PBS per bird. The third and fourth groups (the infected groups) received 200 μL of IBV-M41. The inocula were given half nasally and half orally to mimic the natural infection routes of IBV in the chicken. Chickens were fed diets that met or exceeded the National Research Council requirements. Feed and water were provided ad libitum. The birds were monitored daily for clinical signs of disease and disease parameters were measured as reported by Kjaerup et al. [18] . None of the individuals received antibiotic therapy during the experimental period. The study was carried out under strict ethical approval and monitoring (see the statement at the end of the Materials and Methods section).
For this study 64 spleen samples were harvested and used for RNA sequencing. The birds were sacrificed 1 and 3 weeks post infection by cervical dislocation and spleen samples were collected. At both time points, eight samples from the two lines, L10H and line L10L, from each group (uninfected and infected) were collected as illustrated in Fig. 1 . After collection, spleens were sectioned (triangular cross-sectional slice from upper part) and identical samples from each chicken were immediately placed in RNAlater® Stabilization Solution (Ambion Inc., Austin, Texas) that were incubated at 4°C overnight and then transferred to -20°C the following day.
Tissue samples were homogenized on a TissueLyzer LT (Qiagen, Hilden, Germany). Total RNA was extracted with the Qiagen RNAeasy Kit (Catalog ID 74104, Qiagen, Venlo, Netherlands) according to the manufacturer's instructions. The quality of the 64 total RNA samples was verified using a 2200 TapeStation RNA Screen Tape device (Agilent, Santa Clara, CA, USA) and the concentration ascertained using an ND-1000 spectrophotometer (Nano-Drop, Wilmington, DE).
Libraries were prepared with the Illumina TruseqRNA sample prep kit (Catalog ID FC-122-1001, Illumina, San Diego, USA) following the manufacturer's protocol and evaluated with the Agilent Tape Station 2200. Libraries were quantified by Picogreen and then normalized to 10 nM as recommended by Illumina for cluster generation on the Hiseq2000. Equimolar amounts of each library were mixed before NaOH denaturation.
The Illumina Truseq PE cluster kit v3 (Catalog ID PE-401-3001) was used to generate clusters on the grafted Illumina Flowcell and the hybridized libraries were sequenced on six lines of a Flowcell on the Hiseq2000 with 100 cycles of a paired-end sequencing module using the Truseq SBS kit v3 (Catalog ID FC-401-3001).
Quality control, mapping of RNA sequencing reads and counting mapped reads Initial control quality was assessed by the FastQC software version 0.11.3 [52] . Raw reads were than trimmed for low quality bases using the Trimmomatic tool version 0.32 [53] applying minimum Phred quality score >10 averaged across the sliding window of five bases. Furthermore, all reads with the length below 40 bp were removed.
The trimmed reads were mapped to the Gallus gallus reference genome (Gallus_gallus-4.0, release 80 [54] ) using a spliced aligner TopHat2 version 2.014 [55] . The Gallus gallus gene annotation used for mapping was retrieved from Ensembl database version 80 (www.ensembl.org). The mapping quality was assessed using a set of Python scripts within the RSeQC toolkit [56] . The quality control assessment included inspection of the read coverage over the full gene body in order to assess if reads coverage was uniform and if there was any 5' or 3' bias as well as how the mapped reads were distributed over genome features.
Gene count estimation was performed using the HTSeq-count tool in 'union' mode. The HTSeq-count is a Python script within the HTSeq framework, version 0.7.1, which is an open source toolkit that allows the input of raw counts from aligned reads to be annotated with gene names based on genomic features [57] .
The read counts obtained were used to estimate gene expression and identify differentially expressed (DE) genes. This was achieved using Bioconductor package edgeR version 3.10.0 [58] and limma version 3.24.5 [59] following a previously described protocol [24] . Before performing statistical analysis, genes with low levels of expression were filtered out using a threshold of least one read per million in n of the samples, where n is the size of the smallest group of replicates, which in this case was eight.
In order to account for technical and biological effects reads counts were normalized using the "calc-NormFactors" function implemented in the edgeR package. This function normalizes the data by finding a set of scaling factors for the library sizes that minimizes the log-fold changes between the samples. The scale factors were computed using the trimmed mean of M-values (TMM) between samples [58] . Common and tag-wise dispersion estimates were calculated with the Cox-Reid profile adjusted likelihood method in order to correct for the technical and biological variation when fitting the multivariate negative binomial model [60] .
Multidimensional scaling (MDS) was implemented in the edgeR package, to assess similarity of the samples visually. The MDS plot was created in order to visualize the relationship between samples and identify possible outliers [58] . MDS is based on comparing the relationship between all pairs of samples by applying a countspecific pairwise distance measure [58] . Possible outliers were further investigated using principal component analysis (PCA) to remove samples which fell outside a 95 % confidence ellipse.
A design matrix was created in order to specify the factors that were expected to affect the expression level. The matrix was constructed to fit the saturated model where each treatment combination was considered separately. Eight treatment combinations were considered as illustrated in Fig. 1 : uninfected birds from the line L10H at week 1, uninfected birds from the line L10L at week 1, infected birds from the line L10H at week 1, infected birds from the line L10L at week 1, uninfected birds from the line L10H at week 3, uninfected birds from the line L10L at week 3, infected birds from the line L10H at week 3, infected birds from the line L10L at week 3.
A generalized linear model likelihood ratio test, specifying the difference of interest, was used to test for differential expression between these treatment combinations. The differential expression analysis was performed comparing the log-fold differences in gene counts between two lines (L10H and L10L) at different time points (weeks 1 and 3) and for different status (uninfected and infected) separately (Fig. 1 ). Benjamini Hochberg false discovery rates (FDR) for a transcriptome-wide experiment were calculated to correct for multiple testing [61] . All genes with an FDR-adjusted p-value <0.05 were considered individual genes of interest and were retained for further analysis.
Functional analysis of the DE genes was performed using the Cytoscape version 3.2.1 [62, 63] with the ClueGo version 2.1.7 plug-in [64] to enrich the annotation and enrichment of the differentially expressed (DE) genes for four comparisons (C1-C4, see Fig. 1 ). ClueGO determines the distribution of the target genes across the GO (Gene Ontology) terms and pathways: this study focused on. The p-value was calculated using right-sided hypergeometric tests and Benjamini-Hochberg adjustment was used for multiple test correction. An adjusted pvalue of 0.001 indicated a statistically significant deviation from the expected distribution, and that the corresponding GO terms and pathways were enriched for the target genes. The association strength between the terms was calculated using a corrected kappa statistic of 0.4. The network created represented the terms as nodes which were linked based on a 0.4 kappa score level. The size of the nodes reflected the enrichment significance of the terms. The network was automatically laid out using the Organic layout algorithm supported by Cytoscape. The functional groups were created by iterative merging of initially defined groups based on the predefined kappa score threshold. Only functional groups represented by their most significant term were visualized in the network providing an insightful view of their interrelations [64] .
The experimental procedures were conducted under the protocols approved by the Danish Animal Experiments Inspectorate and complied with the Danish Ministry of Justice Law no. 382 (June 10, 1987) and Acts 739 (December 6, 1988) and 333 (May 19, 1990) concerning animal experimentation and care of experimental animals. The license to conduct the animal experiment was obtained by Helle R. Juul-Madsen. comparison between uninfected birds at week 1. The GO Immune System terms were identified as nodes and linked based on their kappa score level (> = 0.4) and p-value < 0.001. Functionally related groups partially overlapped. The GO terms are labelled in colors according to hierarchical clustering of GO terms. Terms which have not been grouped are shown in grey. The colour pie charts of the GO Immune system nodes show the gene proportion associated with the respective term. (PDF 60 kb)
Additional file 17: Figure S4 . Network representation of enriched GO Immune System terms of differentially expressed (DE) genes for comparison between uninfected birds at week 3. The GO Immune System terms were identified as nodes and linked based on their kappa score level (> = 0.4) and p-value < 0.001. Functionally related groups partially overlapped. The GO terms are labelled in colours according to hierarchical clustering of GO terms. Terms which have not been grouped are shown in grey. The color pie charts of the GO Immune system nodes show the gene proportion associated with the respective term. (PDF 55 kb) Additional file 18: Figure S5 . Network representation of enriched GO Immune System terms of differentially expressed (DE) genes for comparison between infected birds at week 1. The GO Immune System terms were identified as nodes and linked based on their kappa score level (> = 0.4) and p-value < 0.001. Functionally related groups partially overlapped. The GO terms are labelled in colors according to hierarchical clustering of GO terms. Terms which have not been grouped are shown in grey. The color pie charts of the GO Immune system nodes show the gene proportion associated with the respective term. (PDF 70 kb) Additional file 19: Figure S6 . Network representation of enriched GO Immune System terms of differentially expressed (DE) genes for comparison between infected birds at week 3. The GO Immune System terms were identified as nodes and linked based on their kappa score level (> = 0.4) and p-value < 0.001. Functionally related groups partially overlapped. The GO terms are labelled in colors according to hierarchical clustering of GO terms. Terms which have not been grouped are shown in grey. The color pie charts of the GO Immune system nodes show the gene proportion associated with the respective term. (PDF 30 kb) | What differentiated the two chicken lines used in this study? | 5,161 | serum concentration of mannose-binding lectin (MBL) | 1,031 |
1,691 | RNA sequencing-based analysis of the spleen transcriptome following infectious bronchitis virus infection of chickens selected for different mannose-binding lectin serum concentrations
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4729133/
SHA: f5f1cd43740b5b6eca8b3cf2714fc0854a746519
Authors: Hamzić, Edin; Kjærup, Rikke Brødsgaard; Mach, Núria; Minozzi, Guilietta; Strozzi, Francesco; Gualdi, Valentina; Williams, John L.; Chen, Jun; Wattrang, Eva; Buitenhuis, Bart; Juul-Madsen, Helle Risdahl; Dalgaard, Tina Sørensen
Date: 2016-01-27
DOI: 10.1186/s12864-016-2403-1
License: cc-by
Abstract: BACKGROUND: Avian infectious bronchitis is a highly contagious disease of the upper-respiratory tract caused by infectious bronchitis virus (IBV). Understanding the molecular mechanisms involved in the interaction between innate and adaptive immune responses to IBV infection is a crucial element for further improvements in strategies to control IB. To this end, two chicken lines, selected for high (L10H line) and low (L10L line) serum concentration of mannose-binding lectin (MBL) were studied. In total, 32 birds from each line were used. Sixteen birds from each line were infected with IBV and sixteen were left uninfected. Eight uninfected and infected birds from each line were euthanized at 1 and 3 weeks post infection. RNA sequencing was performed on spleen samples from all 64 birds and differential gene expression analysis was performed for four comparisons: L10L line versus L10H line for uninfected birds at weeks 1 and 3, respectively, and in the same way for infected birds. Functional analysis was performed using Gene Ontology (GO) Immune System Process terms specific for Gallus gallus. RESULTS: Comparing uninfected L10H and L10L birds, we identified 1698 and 1424 differentially expressed (DE) genes at weeks 1 and 3, respectively. For the IBV-infected birds, 1934 and 866 DE genes were identified between the two lines at weeks 1 and 3, respectively. The two most enriched GO terms emerging from the comparison of uninfected birds between the two lines were “Lymphocyte activation involved in immune response” and “Somatic recombination of immunoglobulin genes involved in immune response” at weeks 1 and 3, respectively. When comparing IBV-infected birds between the two lines, the most enriched GO terms were “Alpha-beta T cell activation” and “Positive regulation of leukocyte activation” at weeks 1 and 3, respectively. CONCLUSIONS: Healthy birds from the two lines showed significant differences in expression profiles for subsets of adaptive and innate immunity-related genes, whereas comparison of the IBV-infected birds from the two lines showed differences in expression of immunity-related genes involved in T cell activation and proliferation. The observed transcriptome differences between the two lines indicate that selection for MBL had influenced innate as well as adaptive immunity. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (doi:10.1186/s12864-016-2403-1) contains supplementary material, which is available to authorized users.
Text: Conclusions: Healthy birds from the two lines showed significant differences in expression profiles for subsets of adaptive and innate immunity-related genes, whereas comparison of the IBV-infected birds from the two lines showed differences in expression of immunity-related genes involved in T cell activation and proliferation. The observed transcriptome differences between the two lines indicate that selection for MBL had influenced innate as well as adaptive immunity.
Keywords: IBV, Coronavirus, Infectious bronchitis, Chicken, RNA sequencing, Transcriptome, Spleen, Mannose-binding lectin, Immune response Background Avian infectious bronchitis (IB) is an acute and highly contagious disease of the upper-respiratory tract caused by the infectious bronchitis virus (IBV). The virus is a member of the Coronaviridae family and has numerous serotypes and strains. Rapid replication combined with high mutation rate and recombination are the main causes of the observed high diversity [1] . The respiratory tract is the primary target organ and entry point for the virus, before further spread to kidneys and gonads. The most common symptoms of IB are related to the respiratory tract and include gasping, coughing, sneezing, tracheal rales, and nasal discharge [2] . Feed conversion and average daily gain are affected in broilers, and infection is often followed by secondary bacterial infections. In layers, IBV causes a reduction in egg production and egg quality. Today, IB is one of the most economically important diseases in the poultry industry [2] . Infection outbreaks are controlled by a combination of strict management practices and vaccination. The strict management practices, which include the maintenance of the housing temperature and ventilation, are essential, because IBV is highly contagious and spreads very fast. Live attenuated and inactivated vaccines are widely used for control and prevention of IBV infection [3, 4] . As there is little or no cross-protection between different serotypes/variants of the virus, hence vaccines should contain serotypes present in a particular area in order to induce adequate protection [1] . New multi-strain vaccines with the optimal antigen combination and optimal adjuvants are therefore required for future IBV control. Understanding the molecular mechanisms involved in the interaction between innate and adaptive immune responses to IBV infection is a crucial element for further improvements of the vaccines.
IBV infection induces a wide range of immune responses in chickens. An innate immune response is activated during the initial stages of infection in the mucosal lining of the trachea following binding of IBV virions to receptors on epithelial cells [5] . Activation of this innate immune response may be initiated by Toll-like receptor (TLR) signaling upon IBV recognition [6, 7] . In addition, rapid activation of natural killer (NK) cells has been observed one day after IBV infection [8] as well as increased macrophage numbers in lungs and trachea after primary IBV infection [9] . In the case of the adaptive immune responses, T lymphocyte subpopulations are actively involved in the early stages of IBV clearance [7, 10] exhibiting rapid activation upon IBV infection [6] . Furthermore, studies have shown that cytotoxic T lymphocytes (CTL) play an important role in responding to primary infections with IBV [10, 11] . In addition to T cell responses, IBV specific antibodies, of all three antibody classes present in chickens, have been reported [12] [13] [14] . A specific local antibody response in avian infectious bronchitis is characteristic for the response to a secondary infection [15] . The innate and adaptive immune systems are strongly interconnected, which is also seen in the response to IBV infection, and the connection possibly involves the serum collectin, mannose-binding lectin (MBL) as a key player [16] .
Two chicken lines which were selected for high and low MBL serum concentrations (designated L10H and L10L, respectively), were used in the present study. Selective breeding has been performed for 14 generations using the combination of two strains (67.5 % UM-B19 chickens and 33.5 % White Cornish) as a starting population, as described by Juul-Madsen et al. [17] . The final result was two divergent lines, with mean MBL serum concentrations of 33.4 μg/ml for the L10H line and 7.6 μg/ml for the L10L line, respectively [18, 19] . The mean MBL serum concentration for 14 different chicken lines representing both broilers and layers is around 6 μg/ml, but varies from 0.4 to 37.8 μg/ml in normal healthy chickens with protein produced in the liver as the main source of circulating MBL [17] . In chickens, a positive correlation between MBL serum concentrations and the severity of several infections, such as infections caused by IBV [19] , Escherichia coli [20] and Pasteurella multocida [21] , has been observed. Chicken MBL binds to IBV [16, 22] , therefore it is possible that MBL facilitates innate responses such as opsono-phagocytosis, complement activation or virus neutralization, in the early stages of IBV infection. In mammals MBL has also been shown to influence induction of adaptive immunity [23] . In support of the role of MBL in response to IBV, Kjaerup et al. [18] observed considerable differences in cellular adaptive immune parameters in response to an IBV infection between lines L10L and L10H. Furthermore, birds from L10H line exhibited lower viral loads and less severe damage of tracheal cilia following the IBV infection in comparison to birds from the L10L line.
The aim of this study was to characterize the spleen transcriptome of healthy birds from the two lines selected for serum MBL, and to investigate differences in molecular mechanisms behind the development of systemic adaptive immunity between the L10L and L10H lines infected with IBV.
The experimental timeline and sampling time points are as illustrated in Fig. 1 and a full description of the experimental infection is reported by Kjaerup et al. [18] . The birds were infected at 3 weeks of age and from day 2 post-infection (p.i.), showed clinical signs characteristic of IBV infection, including sneezing and labored breathing. Viral loads in tracheal swabs were assessed for all birds as reported in the previous paper published on the experimental infection study [18] . No virus was detected in the uninfected birds at any time point throughout the experiment. Viral genomes were detected in swabs from infected birds from day 1 to 8 p.i. Notably, significantly lower viral loads (p < 0.03) were observed in birds from line L10H in comparison to infected birds from line L10L [18] .
Detection and quantification of splenic gene expression RNA sequencing data were produced from eight infected and eight uninfected birds from each of the two lines at two sampling occasions, as described in the materials and methods section. All samples passed quality control measures for raw and trimmed sequenced reads except for individual no. 46, which was removed due to a very low number of sequenced reads. For the remaining birds, an average of over 37 million reads were obtained per sample for the 63 samples analyzed, with 81 % of the reads mapping to the chicken genome reference sequence, as described in the materials and methods section (See summary statistics with the number of mapped and total reads is presented in Additional file 1: Table S1 ). In total, 17,113 expressed genes were identified. After filtering genes with fewer than one read per million in eight samples [24] (genes which would not achieve statistical significance for differential expression), the final list contained 11,292 expressed genes. Before performing the differential gene expression analysis, further multivariate analysis was carried out on the raw and normalized gene count data to identify any discrepancies.
Multi-dimensional scaling (MDS) plot on expressed genes between the two lines, L10H and L10L, showed that they differ considerably in their transcriptome profiles for both uninfected and IBV-infected birds [See Additional file 2: Figure S1 ]. Moreover, inter-individual variation in gene expression at week 1 was considerably higher than that observed at week 3 for both uninfected and IBV-infected birds [See Additional file 2: Figure S1 ].
Birds 22 and 47 were separated from the rest on the MDS plot [See Additional file 2: Figure S1 ]. However, inspection of raw sequence data and mapping parameters did not identify any technical problems which would explain the observed out-grouping of these birds. In addition an interclass principal component analysis (PCA) was performed using raw and normalized gene counts. The interclass PCA revealed that the birds 22 and 47 were placed outside the 95 % confidence intervals of their respective treatments [See Additional file 3: Figure S2 ]. However, the PCA did not identify any gene having extreme count profiles which may have contributed to the transcriptome dispersion of birds 22 and 47 with respect to their treatment groups. Although there was no clear technical or biological explanation for their out-grouping, these samples were removed from further analysis.
Differential gene expression analysis was performed to compare the two chicken lines (L10L and L10H) at two time points for uninfected (C1 and C2, see Fig. 1 ) and IBV-infected birds (C3 and C4, see Fig. 1 ). A large number of genes were differentially expressed (DE) between L10L and L10H lines at weeks 1 and 3, for both uninfected and IBV-infected birds (see Table 1 , see Fig. 1 ).
We identified 1,698 and 1,424 DE genes for the uninfected birds between lines L10L and L10H at weeks 1 and 3, respectively (see Table 1 ). In total 692 genes had higher expression in L10H line and 1,006 had higher expression in line L10L for the uninfected birds at week 1 [See Additional file 4: Table S2 ] and 774 genes had higher expression in L10H line and 650 genes had higher expression in L10L line for uninfected birds at week 3 [See Additional file 5: Table S3 ].
Comparing IBV-infected L10H and L10L birds, we identified 1,934 and 866 DE genes at weeks 1 and 3, respectively (see Table 1 ). In total 931 genes had higher expression in line L10H and 1,003 had higher expression in line L10L at week 1 and at week 3, 508 had higher expression in line L10H and 358 had higher expression in line L10L (Table 1 , Additional file 6: Table S4 and Additional file 7: Table S5 ).
There were also status-related changes in gene expression as shown in the Venn diagram ( Fig. 2) . At week 1, the total number of DE genes in uninfected birds Fig. 2 ). Out of 3,011 (1077 + 621 + 1313) DE genes for both uninfected and infected birds between the two lines only 621 (~20 %) were common for two comparisons (Fig. 2 ). At week 3, the total number of DE genes in uninfected birds between the two lines was 1424 (883 + 541) ( Table 1 , Fig. 2 ) which was higher comparing to 866 (541 + 325) in infected birds between the two lines ( Table 1 , Fig. 2 ). When comparing the uninfected and infected birds between the two lines, 541 (~30 %) genes were common out of total of 1749 (883 + 541+ 325) DE genes for both comparisons (Fig. 2) .
Moreover, we also performed differential gene expression analysis to compare two time points (week 1 and week 3) in the two chicken lines (L10L and L10H) for uninfected (C5 and C6, see Fig. 1 ) and IBV-infected birds (C7 and C8, see Fig. 1 ). Finally, differential gene expression analysis was also conducted to compare the two infection states (uninfected and IBV-infected) at two time points for the L10L chicken line (C9 and C10, see Table S6 , Additional file 9: Table S7 , Additional file 10: Table S8 , Additional file 11: Table S9 , Additional file 12: Table S10 , Additional file 13: Table S11 , Additional file 14: Table S12 and Additional file 15: Table S13 ].
An enrichment gene set analysis was carried out to identify over-represented Gene Ontology (GO) "Immune System Process" terms using the lists of DE genes from comparisons between uninfected and infected birds from the two lines at 1 and 3 weeks p.i. The most enriched GO Immune System terms between the two lines when comparing uninfected birds from the two lines and then infected from the two lines are shown in Fig. 3 .
GO Immune System terms associated with genes that were differentially expressed between the two lines for uninfected birds at week 1 were "Lymphocyte activation involved in immune response" (GO:0002285), "Activation of innate immune response" (GO:0002218), "Lymphocyte mediated immunity" (GO:0002449), and "Leukocyte Comparison between the two lines, L10H and L10L, uninfected and infected birds at two time points (weeks 1 and 3). Comparisons C1 -C4 correspond to differential gene expression comparisons presented in Fig. 1 For uninfected birds at week 3, the most enriched GO Immune System terms were "Somatic recombination of immunoglobulin genes involved in immune response" (GO:0002204) and the "Adaptive immune response based on somatic recombination of immune receptors built from immunoglobulin superfamily" (GO:0002460) [See Fig. 3 , See Additional file 17: Figure S4 ]. In total, 47 DE genes mapped to GO Immune System terms in this comparison (Fig. 4) . Among the DE genes that had a higher expression in the line L10H at week 3, in the uninfected group, were IL7, FKBP1B, FAS and PTPN22, which were also seen differentially expressed between lines at week 1 [See Fig. 4 , Additional file 17: Figure S4 ].
Comparing infected birds from the two lines, at week 1, "Alpha-beta T cell activation" (GO:0046631), "Activation of innate immune response" (GO:0002218) and "Leukocyte differentiation" (GO:0002521) functions were the three most enriched GO Immune System terms (Fig. 3) . CXCR4 (Chemokine receptor 4), PTPN22 and FAS were among the most highly expressed genes in L10H [See Fig. 4 , Additional file 18: Figure S5 ].
The major GO Immune System term that was strongly enriched for in the infected birds at week 3 was "Positive regulation of leukocyte activation" (GO:0002696) [See Figure S6 ].
The present study used two lines, L10L and L10H, which have been divergently selected for high and low MBL serum concentration for 14 generations, Fig. 3 Functional map of differentially expressed genes enriched for GO Immune System terms. The top categories of the GO Immune System terms associated with differentially expressed (DE) genes. All categories were statistically significant (adjusted p-value < 0.001). The chart fragments represent the number of genes associated with the terms as a proportion of the total number of genes within the respective GO term. Terms which have not been grouped are shown in grey respectively. These two lines have earlier been extensively used for immunological studies and exhibit differences in immunological parameters after being challenged with several pathogens [18, 22, 25] .
The spleen is a secondary lymphoid organ where innate and adaptive immune responses can be efficiently mounted. In addition, the avian spleen is considered to play a very important immunological role because avian lymphatic vessels and lymph nodes are poorly developed [26] . The transcriptome differences in the spleen between the two lines, L10L and L10H, for uninfected (healthy) and IBV-infected birds were investigated, focusing on the differential expression of immune-related genes within significantly enriched immune-related GO terms. Large differences in transcriptome profiles were observed between birds from the two lines, both uninfected (healthy) and following the experimental IBV challenge [See Additional file 2: Figure S1 ]. This suggests Fig. 4 Differentially expressed genes associated with the GO Immune System term. Heatmap representation of the differentially expressed (DE) genes associated with the GO Immune System terms for the four comparisons between the two lines, L10H and L10L, uninfected and infected groups at two time points, weeks 1 and 3. The heat map is constructed using the average values of counts per millions for each group that selection for MBL serum levels in the two lines had a much wider effect which goes beyond the expression of the MBL gene [27] . The observed transcriptome differences can probably be attributed to correlated response to selection [28] or random genetic drift [29] . Correlated selection occurs when a trait is affected by selection on a another trait and is dependent on a genetic correlation between the two traits, which is well known in animal breeding [30] . Alternatively, random genetic drift could contribute to the observed differences, considering that the founder population of the two lines was small [17] .
Focusing on the expression of immune-related genes: at week 1, the uninfected birds showed differences in the expression of genes involved in both adaptive immunity and innate immunity-related pathways [See Additional file 16: Figure S3 ]. In addition, the line L10H had a lower expression for the subset of the innate immune genes, TYRO3, TRAF3 and TLR7 at week 1 [See Additional file 16: Figure S3 and Fig. 4 ]. TYRO3 encodes tyrosineprotein kinase receptor 3 (TYRO3) which is involved in inhibition of TLR signaling pathways and TLR-induced cytokine signaling pathways. These two pathways influence immune-related processes, including cell proliferation/survival, cell adhesion and migration and inhibits the innate inflammatory response to pathogens [31] . TRAF3 encodes a cytoplasmic signaling protein, which plays a critical role in the regulation of antiviral response and viral evasion [32] [33] [34] . TLR7 was also among the DE innate immunity-related genes with higher expression in the line L10H. Chicken TLR7 has been shown to play a part in the response to IBV infections [9, 35] .
In addition to the differences in the expression profiles for a subset of innate immune genes, the two lines also differed in their expression of adaptive immune genes. Uninfected birds from L10H had a higher gene expression compared to the line L10L, for TGFB3, IL7, FKBP1B, FAS and PTPN22. These genes are known to be involved in a wide range of adaptive immune processes. In humans, reducing the TGF-β signaling on T cells has been shown to increase the function of CD8 T cells in an indirect way which results in the rapid elimination of viruses, enabling the creation of an effective memory response [36] . Similarly, human IL-7 plays a key role in the survival of both naïve [37] and memory [38, 39] CD4 and CD8 T cells. Moreover, an in vitro study showed that inhibition of FKBP1B and other cyclophilins blocked the replication of different Coronaviruses, including IBV [40] . In analogy, uninfected birds from the L10H line had a higher expression of IL7, FKBP1B, FAS and PTPN22.
Generally the uninfected (healthy) birds from the two lines exhibit different expression profiles for this subset of innate and adaptive immune genes probably resulting from the divergent selection for the MBL serum concentration. Selection in animal breeding have been shown to have an extensive effect on a variety of traits including immunological [30] . Moreover, correlated response to selection has been observed in the case where selection was performed on less complex traits such as testosterone levels [41] . Finally, different expression profiles for the subset of innate and adaptive immune in the uninfected birds from the two lines might be due to the balance in the effect of MBL serum levels. High levels of human MBL have been mostly reported as beneficial while in case of intracellular parasitic disease the effect of MBL serum level might be opposite [42] . The results indicate that selection for MBL serum levels might lead to favoring specific modes of immune responses depending on the MBL function.
Large differences in the expression patterns were seen between the two lines following infection with IBV and these differences involved adaptive immunity-related pathways which are associated with "Alpha-beta T cell activation" (GO:0046631), and "Activation of innate immune response" (GO:0002218) [See Additional file 18: Figure S5 ]. The observed enrichment for GO terms related to T cell activation is in accordance with a previous study of these lines that showed that the IBV-specific T cells are present in large numbers in the spleen after IBV infection [43] . At week 1 post infection CXCR4, FAS and PTPN22 showed higher expression in line L10H. The CXC chemokine receptors are expressed on both effector and memory T cells and play a key role in the homeostasis of memory T cells [44] . FAS has been shown to be upregulated in the kidney of chickens challenged with IBV [45] . The Fas/FasL pathway is an important pathway of killing for cytotoxic T cells [46] . Similarly, PTPN22 has been shown to be differentially expressed in chickens following pathogen challenge, and in particular following infection with Escherichia coli [47] .
Additionally VCAM1 and JMJD6 were among the adaptive immunity-related genes DE between lines at week 3 [See Fig. 4 , Additional file 19: Figure S6 ]. The VCAM1 gene is known to be involved in the activation of T cells [48] . Furthermore, a recent study demonstrated that JMJD6 regulates proliferation of memory T cells during a viral infection [49] which is of great interest considering that had higher expression in the IBVinfected birds from L10H line at week 3.
The results show that the two lines differ greatly in the expression of adaptive immunity-related genes following infection, which may imply the presence of different modes of gene regulation. The sampling times were chosen to access responses both in the effector phase (week 1) and memory phase (week 3) of the adaptive immune response to IBV. In accordance, at 1 week, post infection subsets of genes actively involved in T cell proliferation show differences between the lines. Also, at week 3 immune-related gene expression profiles in response to IBV infection that differ between the lines are more related to maintenance of T cell memory. MBL is known to be involved in regulation of dendritic cell maturation as well as cytokine production [50] . Dendritic cells, which are the main antigen presenting cells and are actively involved in regulation of adaptive immune responses, possess the receptors for MBL in mammals [23] . Therefore, the two lines selected for different MBL serum concentration may display differences in adaptive immune responses and development of adaptive immunity as a result of differences in response to cytokine signaling from dendritic cells. Other studies of the two lines, L10L and L10H, have shown that they differ in disease response parameters after being challenged with different pathogens. The L10L line has been associated with increased viral replication in the airway after an infectious bronchitis virus (IBV) infection [18] , reduced growth rate after an Escherichia coli infection [20] and greater intestinal colonization after Salmonella Infantis infection [51] . In the present experiment, significantly lower viral loads (p < 0.03) were observed in birds from line L10H in comparison to infected birds from line L10L [18] . Furthermore, L10H birds in the present study exhibited a less severe damage of tracheal cilia following the IBV infection in comparison to the L10L line (unpublished data). In the current experiment phenotypic differences in additional traits connected to adaptive immunity were observed, including numbers of circulating B cells and cytotoxic T cells [18] . Based on these observations it seems that selection for high MBL serum concentration allows birds to cope better after being infected with a range of pathogens. Therefore, the observed differences in the expression profiles for the adaptive and innate immune-related genes are a reflection of differences in disease resistance and immune responses between the lines L10L and L10H.
In conclusion, large differences in the spleen transcriptome between the two chicken lines, L10L and L10H, were observed in both uninfected (healthy) and IBVinfected birds. The uninfected birds from the two lines showed differences in expression profiles for a subset of both adaptive and innate immunity-related genes, which may represent differences in preparedness to respond to an infection. Following infection with IBV, the two lines showed large differences in expression of genes involved in the adaptive cellular immune response such as T cell activation and proliferation pathways and hence their ability to respond to the infection, which is reflected in the difference in pathogen load seen between the two lines.
This study is a follow-up of the experiment performed by Kjaerup et al. [18] which characterized the cellular and humoral immune response of the two chicken lines, L10H and L10L, divergently selected for MBL serum concentrations following IBV infection. In total, 96 birds were used in the experimental study originating from the two Aarhus University inbred lines, L10H and L10L [19] . All 96 birds were reared together in a biosecure IBV-free environment until they were 3 weeks of age and then allocated to two different groups with 24 birds from each line in each group (uninfected and infected). The birds were transferred to a biosafety level 2 facility and placed in isolators. Two isolators contained uninfected chickens and two isolators contained infected chickens. Each isolator having an equal mix of the two lines as described by Kjaerup et al. [18] .
The virulent IBV-M41 strain was used for the infection (a kind gift from Dr. med. vet. Hans C. Philipp at the Lohmann Animal Health GmbH, Cuxhaven, Germany). The virus had been passaged twice in specific pathogen-free embryonated eggs. The IBV inocula were prepared in phosphate-buffered saline (PBS) immediately before use and contained 2 × 10 5.2 EID 50 /200 μl of IBV-M41 virus. The first and the second group (the uninfected groups) were mock-infected with 200 μl PBS per bird. The third and fourth groups (the infected groups) received 200 μL of IBV-M41. The inocula were given half nasally and half orally to mimic the natural infection routes of IBV in the chicken. Chickens were fed diets that met or exceeded the National Research Council requirements. Feed and water were provided ad libitum. The birds were monitored daily for clinical signs of disease and disease parameters were measured as reported by Kjaerup et al. [18] . None of the individuals received antibiotic therapy during the experimental period. The study was carried out under strict ethical approval and monitoring (see the statement at the end of the Materials and Methods section).
For this study 64 spleen samples were harvested and used for RNA sequencing. The birds were sacrificed 1 and 3 weeks post infection by cervical dislocation and spleen samples were collected. At both time points, eight samples from the two lines, L10H and line L10L, from each group (uninfected and infected) were collected as illustrated in Fig. 1 . After collection, spleens were sectioned (triangular cross-sectional slice from upper part) and identical samples from each chicken were immediately placed in RNAlater® Stabilization Solution (Ambion Inc., Austin, Texas) that were incubated at 4°C overnight and then transferred to -20°C the following day.
Tissue samples were homogenized on a TissueLyzer LT (Qiagen, Hilden, Germany). Total RNA was extracted with the Qiagen RNAeasy Kit (Catalog ID 74104, Qiagen, Venlo, Netherlands) according to the manufacturer's instructions. The quality of the 64 total RNA samples was verified using a 2200 TapeStation RNA Screen Tape device (Agilent, Santa Clara, CA, USA) and the concentration ascertained using an ND-1000 spectrophotometer (Nano-Drop, Wilmington, DE).
Libraries were prepared with the Illumina TruseqRNA sample prep kit (Catalog ID FC-122-1001, Illumina, San Diego, USA) following the manufacturer's protocol and evaluated with the Agilent Tape Station 2200. Libraries were quantified by Picogreen and then normalized to 10 nM as recommended by Illumina for cluster generation on the Hiseq2000. Equimolar amounts of each library were mixed before NaOH denaturation.
The Illumina Truseq PE cluster kit v3 (Catalog ID PE-401-3001) was used to generate clusters on the grafted Illumina Flowcell and the hybridized libraries were sequenced on six lines of a Flowcell on the Hiseq2000 with 100 cycles of a paired-end sequencing module using the Truseq SBS kit v3 (Catalog ID FC-401-3001).
Quality control, mapping of RNA sequencing reads and counting mapped reads Initial control quality was assessed by the FastQC software version 0.11.3 [52] . Raw reads were than trimmed for low quality bases using the Trimmomatic tool version 0.32 [53] applying minimum Phred quality score >10 averaged across the sliding window of five bases. Furthermore, all reads with the length below 40 bp were removed.
The trimmed reads were mapped to the Gallus gallus reference genome (Gallus_gallus-4.0, release 80 [54] ) using a spliced aligner TopHat2 version 2.014 [55] . The Gallus gallus gene annotation used for mapping was retrieved from Ensembl database version 80 (www.ensembl.org). The mapping quality was assessed using a set of Python scripts within the RSeQC toolkit [56] . The quality control assessment included inspection of the read coverage over the full gene body in order to assess if reads coverage was uniform and if there was any 5' or 3' bias as well as how the mapped reads were distributed over genome features.
Gene count estimation was performed using the HTSeq-count tool in 'union' mode. The HTSeq-count is a Python script within the HTSeq framework, version 0.7.1, which is an open source toolkit that allows the input of raw counts from aligned reads to be annotated with gene names based on genomic features [57] .
The read counts obtained were used to estimate gene expression and identify differentially expressed (DE) genes. This was achieved using Bioconductor package edgeR version 3.10.0 [58] and limma version 3.24.5 [59] following a previously described protocol [24] . Before performing statistical analysis, genes with low levels of expression were filtered out using a threshold of least one read per million in n of the samples, where n is the size of the smallest group of replicates, which in this case was eight.
In order to account for technical and biological effects reads counts were normalized using the "calc-NormFactors" function implemented in the edgeR package. This function normalizes the data by finding a set of scaling factors for the library sizes that minimizes the log-fold changes between the samples. The scale factors were computed using the trimmed mean of M-values (TMM) between samples [58] . Common and tag-wise dispersion estimates were calculated with the Cox-Reid profile adjusted likelihood method in order to correct for the technical and biological variation when fitting the multivariate negative binomial model [60] .
Multidimensional scaling (MDS) was implemented in the edgeR package, to assess similarity of the samples visually. The MDS plot was created in order to visualize the relationship between samples and identify possible outliers [58] . MDS is based on comparing the relationship between all pairs of samples by applying a countspecific pairwise distance measure [58] . Possible outliers were further investigated using principal component analysis (PCA) to remove samples which fell outside a 95 % confidence ellipse.
A design matrix was created in order to specify the factors that were expected to affect the expression level. The matrix was constructed to fit the saturated model where each treatment combination was considered separately. Eight treatment combinations were considered as illustrated in Fig. 1 : uninfected birds from the line L10H at week 1, uninfected birds from the line L10L at week 1, infected birds from the line L10H at week 1, infected birds from the line L10L at week 1, uninfected birds from the line L10H at week 3, uninfected birds from the line L10L at week 3, infected birds from the line L10H at week 3, infected birds from the line L10L at week 3.
A generalized linear model likelihood ratio test, specifying the difference of interest, was used to test for differential expression between these treatment combinations. The differential expression analysis was performed comparing the log-fold differences in gene counts between two lines (L10H and L10L) at different time points (weeks 1 and 3) and for different status (uninfected and infected) separately (Fig. 1 ). Benjamini Hochberg false discovery rates (FDR) for a transcriptome-wide experiment were calculated to correct for multiple testing [61] . All genes with an FDR-adjusted p-value <0.05 were considered individual genes of interest and were retained for further analysis.
Functional analysis of the DE genes was performed using the Cytoscape version 3.2.1 [62, 63] with the ClueGo version 2.1.7 plug-in [64] to enrich the annotation and enrichment of the differentially expressed (DE) genes for four comparisons (C1-C4, see Fig. 1 ). ClueGO determines the distribution of the target genes across the GO (Gene Ontology) terms and pathways: this study focused on. The p-value was calculated using right-sided hypergeometric tests and Benjamini-Hochberg adjustment was used for multiple test correction. An adjusted pvalue of 0.001 indicated a statistically significant deviation from the expected distribution, and that the corresponding GO terms and pathways were enriched for the target genes. The association strength between the terms was calculated using a corrected kappa statistic of 0.4. The network created represented the terms as nodes which were linked based on a 0.4 kappa score level. The size of the nodes reflected the enrichment significance of the terms. The network was automatically laid out using the Organic layout algorithm supported by Cytoscape. The functional groups were created by iterative merging of initially defined groups based on the predefined kappa score threshold. Only functional groups represented by their most significant term were visualized in the network providing an insightful view of their interrelations [64] .
The experimental procedures were conducted under the protocols approved by the Danish Animal Experiments Inspectorate and complied with the Danish Ministry of Justice Law no. 382 (June 10, 1987) and Acts 739 (December 6, 1988) and 333 (May 19, 1990) concerning animal experimentation and care of experimental animals. The license to conduct the animal experiment was obtained by Helle R. Juul-Madsen. comparison between uninfected birds at week 1. The GO Immune System terms were identified as nodes and linked based on their kappa score level (> = 0.4) and p-value < 0.001. Functionally related groups partially overlapped. The GO terms are labelled in colors according to hierarchical clustering of GO terms. Terms which have not been grouped are shown in grey. The colour pie charts of the GO Immune system nodes show the gene proportion associated with the respective term. (PDF 60 kb)
Additional file 17: Figure S4 . Network representation of enriched GO Immune System terms of differentially expressed (DE) genes for comparison between uninfected birds at week 3. The GO Immune System terms were identified as nodes and linked based on their kappa score level (> = 0.4) and p-value < 0.001. Functionally related groups partially overlapped. The GO terms are labelled in colours according to hierarchical clustering of GO terms. Terms which have not been grouped are shown in grey. The color pie charts of the GO Immune system nodes show the gene proportion associated with the respective term. (PDF 55 kb) Additional file 18: Figure S5 . Network representation of enriched GO Immune System terms of differentially expressed (DE) genes for comparison between infected birds at week 1. The GO Immune System terms were identified as nodes and linked based on their kappa score level (> = 0.4) and p-value < 0.001. Functionally related groups partially overlapped. The GO terms are labelled in colors according to hierarchical clustering of GO terms. Terms which have not been grouped are shown in grey. The color pie charts of the GO Immune system nodes show the gene proportion associated with the respective term. (PDF 70 kb) Additional file 19: Figure S6 . Network representation of enriched GO Immune System terms of differentially expressed (DE) genes for comparison between infected birds at week 3. The GO Immune System terms were identified as nodes and linked based on their kappa score level (> = 0.4) and p-value < 0.001. Functionally related groups partially overlapped. The GO terms are labelled in colors according to hierarchical clustering of GO terms. Terms which have not been grouped are shown in grey. The color pie charts of the GO Immune system nodes show the gene proportion associated with the respective term. (PDF 30 kb) | Which organ was used for the RNA sequencing samples? | 5,162 | spleen | 1,360 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | What is needed to elucidate zoonotic emergence? | 2,719 | A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics | 519 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | What is the conclusion of this report? | 2,720 | heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats. | 1,153 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | Why have bats received attention in recent years? | 2,721 | for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus | 1,588 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | What difference bats demonstrate compared to most non-Chiropteran mammals? | 2,722 | no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies | 2,159 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | What suite of species-specific mechanisms do bats have to limit viral load? | 2,723 | host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine | 2,697 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | How are mammalian cells typically rendered antiviral? | 2,724 | the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells | 2,916 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | In non-flying mammals, what what would be elicited by IFN expression upon viral infection? | 2,725 | widespread inflammation and concomitant immunopathology upon viral infection | 3,476 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | What do the bats do instead? | 2,726 | bats support unique adaptations to combat inflammation | 3,558 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | Why may the bats have this unique adaptation? | 2,727 | to mitigate metabolic damage induced during flight | 3,716 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | Why was the field of virus dynamics developed? | 2,728 | to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates | 4,505 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | How are bats connected to fatal viral diseases? | 2,729 | bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus. | 4,964 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | What is an example of anti-viral defense in bats? | 2,730 | some bats have an antiviral immune response called the interferon pathway perpetually switched on | 5,247 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | What would be caused by this hyper-vigilance in most other mammals? | 2,731 | harmful inflammation | 5,425 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | How are bats different? | 2,732 | Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. | 5,448 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | What bat species cells were compared? | 2,733 | -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. | 5,867 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | What was the conclusion of the study ? | 2,734 | In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not. | 6,368 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | What would be the benefit of learning more about bat's defenses and how they drive virus evolution? | 2,735 | help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. | 6,903 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | Which cells are IFN-defective and therefore limited in antiviral capacity? | 2,736 | demonstrate idiosyncratic induced interferon responses upon viral challenge | 9,543 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | What cells demonstrate idiosyncratic interferon response? | 2,737 | RoNi/7.1 (Rousettus aegyptiacus) cells | 9,498 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | Which cells express constitutive IFN-a? | 2,738 | PaKiT01 (Pteropus alecto) cell | 9,718 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | How were the spread of GFP-expressing virus-infected cells across tissue monolayers tracked via inverted fluorescence microscopy? | 2,739 | Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space | 11,105 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | How were the spread of GFP-expressing virus-infected cells tracked? | 2,740 | For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combinatio | 11,367 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | How was the modeling carried out? | 2,741 | an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. | 31,638 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | What was the finding in this study? | 2,742 | that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates | 31,802 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | What supports the results? | 2,743 | by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. | 32,062 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | What was additionally demonstrated? | 2,744 | the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. | 32,277 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | What do the studies suggest? | 2,745 | that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats. | 32,538 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | What was the methodology for this study? | 2,746 | we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. | 32,879 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | What was demonstrated in deriving the equation for R 0? | 2,747 | invasion threshold is elevated at high values of constitutive antiviral acquisition, | 33,549 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | What is a conclusion of the modeling? | 2,748 | Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. | 33,638 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | What is a conclusion of the study? | 2,749 | Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. | 33,956 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | What do fits to rVSV-MARV infections on PaKiT01 cells suggest? | 2,750 | that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. | 35,005 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | What do the findings indicate? | 2,751 | enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. | 35,958 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | What is presented in this study? | 2,752 | general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. | 37,732 |
1,698 | Accelerated viral dynamics in bat cell lines, with implications for zoonotic emergence
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7064339/
SHA: f2cc0d63ff2c4aaa127c4caae21d8f3a0067e3d5
Authors: Brook, Cara E; Boots, Mike; Chandran, Kartik; Dobson, Andrew P; Drosten, Christian; Graham, Andrea L; Grenfell, Bryan T; Müller, Marcel A; Ng, Melinda; Wang, Lin-Fa; van Leeuwen, Anieke
Date: 2020-02-03
DOI: 10.7554/elife.48401
License: cc-by
Abstract: Bats host virulent zoonotic viruses without experiencing disease. A mechanistic understanding of the impact of bats’ virus hosting capacities, including uniquely constitutive immune pathways, on cellular-scale viral dynamics is needed to elucidate zoonotic emergence. We carried out virus infectivity assays on bat cell lines expressing induced and constitutive immune phenotypes, then developed a theoretical model of our in vitro system, which we fit to empirical data. Best fit models recapitulated expected immune phenotypes for representative cell lines, supporting robust antiviral defenses in bat cells that correlated with higher estimates for within-host viral propagation rates. In general, heightened immune responses limit pathogen-induced cellular morbidity, which can facilitate the establishment of rapidly-propagating persistent infections within-host. Rapidly-transmitting viruses that have evolved with bat immune systems will likely cause enhanced virulence following emergence into secondary hosts with immune systems that diverge from those unique to bats.
Text: Bats have received much attention in recent years for their role as reservoir hosts for emerging viral zoonoses, including rabies and related lyssaviruses, Hendra and Nipah henipaviruses, Ebola and Marburg filoviruses, and SARS coronavirus (Calisher et al., 2006; Wang and Anderson, 2019) . In most non-Chiropteran mammals, henipaviruses, filoviruses, and coronaviruses induce substantial morbidity and mortality, display short durations of infection, and elicit robust, long-term immunity in hosts surviving infection (Nicholls et al., 2003; Hooper et al., 2001; Mahanty and Bray, 2004) . Bats, by contrast, demonstrate no obvious disease symptoms upon infection with pathogens that are highly virulent in non-volant mammals (Schountz et al., 2017) but may, instead, support viruses as longterm persistent infections, rather than transient, immunizing pathologies (Plowright et al., 2016) .
Recent research advances are beginning to shed light on the molecular mechanisms by which bats avoid pathology from these otherwise virulent pathogens (Brook and Dobson, 2015) . Bats leverage a suite of species-specific mechanisms to limit viral load, which include host receptor sequence incompatibilities for some bat-virus combinations (Ng et al., 2015; Takadate et al., 2020) and constitutive expression of the antiviral cytokine, IFN-a, for others (Zhou et al., 2016) . Typically, the presence of viral RNA or DNA in the cytoplasm of mammalian cells will induce secretion of type I interferon proteins (IFN-a and IFN-b), which promote expression and translation of interferon-stimulated genes (ISGs) in neighboring cells and render them effectively antiviral (Stetson and Medzhitov, 2006) . In some bat cells, the transcriptomic blueprints for this IFN response are expressed constitutively, even in the absence of stimulation by viral RNA or DNA (Zhou et al., 2016) . In non-flying mammals, constitutive IFN expression would likely elicit widespread inflammation and concomitant immunopathology upon viral infection, but bats support unique adaptations to combat inflammation (Zhang et al., 2013; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018) that may have evolved to mitigate metabolic damage induced during flight (Kacprzyk et al., 2017) . The extent to which constitutive IFN-a expression signifies constitutive antiviral defense in the form of functional IFN-a protein remains unresolved. In bat cells constitutively expressing IFN-a, some protein-stimulated, downstream ISGs appear to be also constitutively expressed, but additional ISG induction is nonetheless possible following viral challenge and stimulation of IFN-b (Zhou et al., 2016; Xie et al., 2018) . Despite recent advances in molecular understanding of bat viral tolerance, the consequences of this unique bat immunity on within-host virus dynamics-and its implications for understanding zoonotic emergence-have yet to be elucidated.
The field of 'virus dynamics' was first developed to describe the mechanistic underpinnings of long-term patterns of steady-state viral load exhibited by patients in chronic phase infections with HIV, who appeared to produce and clear virus at equivalent rates (Nowak and May, 2000; Ho et al., 1995) . Models of simple target cell depletion, in which viral load is dictated by a bottom-eLife digest Bats can carry viruses that are deadly to other mammals without themselves showing serious symptoms. In fact, bats are natural reservoirs for viruses that have some of the highest fatality rates of any viruses that people acquire from wild animals -including rabies, Ebola and the SARS coronavirus.
Bats have a suite of antiviral defenses that keep the amount of virus in check. For example, some bats have an antiviral immune response called the interferon pathway perpetually switched on. In most other mammals, having such a hyper-vigilant immune response would cause harmful inflammation. Bats, however, have adapted anti-inflammatory traits that protect them from such harm, include the loss of certain genes that normally promote inflammation. However, no one has previously explored how these unique antiviral defenses of bats impact the viruses themselves. Now, Brook et al. have studied this exact question using bat cells grown in the laboratory. The experiments made use of cells from one bat species -the black flying fox -in which the interferon pathway is always on, and another -the Egyptian fruit bat -in which this pathway is only activated during an infection. The bat cells were infected with three different viruses, and then Brook et al. observed how the interferon pathway helped keep the infections in check, before creating a computer model of this response. The experiments and model helped reveal that the bats' defenses may have a potential downside for other animals, including humans. In both bat species, the strongest antiviral responses were countered by the virus spreading more quickly from cell to cell. This suggests that bat immune defenses may drive the evolution of faster transmitting viruses, and while bats are well protected from the harmful effects of their own prolific viruses, other creatures like humans are not.
The findings may help to explain why bats are often the source for viruses that are deadly in humans. Learning more about bats' antiviral defenses and how they drive virus evolution may help scientists develop better ways to predict, prevent or limit the spread of viruses from bats to humans. More studies are needed in bats to help these efforts. In the meantime, the experiments highlight the importance of warning people to avoid direct contact with wild bats. up resource supply of infection-susceptible host cells, were first developed for HIV (Perelson, 2002) but have since been applied to other chronic infections, including hepatitis-C virus (Neumann et al., 1998) , hepatitis-B virus (Nowak et al., 1996) and cytomegalovirus (Emery et al., 1999) . Recent work has adopted similar techniques to model the within-host dynamics of acute infections, such as influenza A and measles, inspiring debate over the extent to which explicit modeling of top-down immune control can improve inference beyond the basic resource limitation assumptions of the target cell model (Baccam et al., 2006; Pawelek et al., 2012; Saenz et al., 2010; Morris et al., 2018) .
To investigate the impact of unique bat immune processes on in vitro viral kinetics, we first undertook a series of virus infection experiments on bat cell lines expressing divergent interferon phenotypes, then developed a theoretical model elucidating the dynamics of within-host viral spread. We evaluated our theoretical model analytically independent of the data, then fit the model to data recovered from in vitro experimental trials in order to estimate rates of within-host virus transmission and cellular progression to antiviral status under diverse assumptions of absent, induced, and constitutive immunity. Finally, we confirmed our findings in spatially-explicit stochastic simulations of fitted time series from our mean field model. We hypothesized that top-down immune processes would overrule classical resource-limitation in bat cell lines described as constitutively antiviral in the literature, offering a testable prediction for models fit to empirical data. We further predicted that the most robust antiviral responses would be associated with the most rapid within-host virus propagation rates but also protect cells against virus-induced mortality to support the longest enduring infections in tissue culture.
We first explored the influence of innate immune phenotype on within-host viral propagation in a series of infection experiments in cell culture. We conducted plaque assays on six-well plate monolayers of three immortalized mammalian kidney cell lines: [1] Vero (African green monkey) cells, which are IFN-defective and thus limited in antiviral capacity (Desmyter et al., 1968) ; [2] RoNi/7.1 (Rousettus aegyptiacus) cells which demonstrate idiosyncratic induced interferon responses upon viral challenge (Kuzmin et al., 2017; Arnold et al., 2018; Biesold et al., 2011; Pavlovich et al., 2018) ; and [3] PaKiT01 (Pteropus alecto) cells which constitutively express IFN-a (Zhou et al., 2016; Crameri et al., 2009) . To intensify cell line-specific differences in constitutive immunity, we carried out infectivity assays with GFP-tagged, replication-competent vesicular stomatitis Indiana viruses: rVSV-G, rVSV-EBOV, and rVSV-MARV, which have been previously described (Miller et al., 2012; Wong et al., 2010) . Two of these viruses, rVSV-EBOV and rVSV-MARV, are recombinants for which cell entry is mediated by the glycoprotein of the bat-evolved filoviruses, Ebola (EBOV) and Marburg (MARV), thus allowing us to modulate the extent of structural, as well as immunological, antiviral defense at play in each infection. Previous work in this lab has demonstrated incompatibilities in the NPC1 filovirus receptor which render PaKiT01 cells refractory to infection with rVSV-MARV (Ng and Chandrab, 2018, Unpublished results) , making them structurally antiviral, over and above their constitutive expression of IFN-a. All three cell lines were challenged with all three viruses at two multiplicities of infection (MOI): 0.001 and 0.0001. Between 18 and 39 trials were run at each cell-virus-MOI combination, excepting rVSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which only eight trials were run (see Materials and methods; Figure 1 -figure supplements 1-3, Supplementary file 1).
Because plaque assays restrict viral transmission neighbor-to-neighbor in two-dimensional cellular space (Howat et al., 2006) , we were able to track the spread of GFP-expressing virus-infected cells across tissue monolayers via inverted fluorescence microscopy. For each infection trial, we monitored and re-imaged plates for up to 200 hr of observations or until total monolayer destruction, processed resulting images, and generated a time series of the proportion of infectious-cell occupied plate space across the duration of each trial (see Materials and methods). We used generalized additive models to infer the time course of all cell culture replicates and construct the multi-trial dataset to which we eventually fit our mechanistic transmission model for each cell line-virus-specific combination ( Figure 1; Figure 1 -figure supplements 1-5).
All three recombinant vesicular stomatitis viruses (rVSV-G, rVSV-EBOV, and rVSV-MARV) infected Vero, RoNi/7.1, and PaKiT01 tissue cultures at both focal MOIs. Post-invasion, virus spread rapidly across most cell monolayers, resulting in virus-induced epidemic extinction. Epidemics were less severe in bat cell cultures, especially when infected with the recombinant filoviruses, rVSV-EBOV and rVSV-MARV. Monolayer destruction was avoided in the case of rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells: in the former, persistent viral infection was maintained throughout the 200 hr duration of each experiment, while, in the latter, infection was eliminated early in the time series, preserving a large proportion of live, uninfectious cells across the duration of the experiment. We assumed this pattern to be the result of immune-mediated epidemic extinction (Figure 1) . Patterns from MOI = 0.001 were largely recapitulated at MOI = 0.0001, though at somewhat reduced total proportions (Figure 1-figure supplement 5 ).
A theoretical model fit to in vitro data recapitulates expected immune phenotypes for bat cells We next developed a within-host model to fit to these data to elucidate the effects of induced and constitutive immunity on the dynamics of viral spread in host tissue ( Figure 1 ). The compartmental within-host system mimicked our two-dimensional cell culture monolayer, with cells occupying five distinct infection states: susceptible (S), antiviral (A), exposed (E), infectious (I), and dead (D). We modeled exposed cells as infected but not yet infectious, capturing the 'eclipse phase' of viral integration into a host cell which precedes viral replication. Antiviral cells were immune to viral infection, in accordance with the 'antiviral state' induced from interferon stimulation of ISGs in tissues adjacent to infection (Stetson and Medzhitov, 2006) . Because we aimed to translate available data into modeled processes, we did not explicitly model interferon dynamics but instead scaled the rate of cell progression from susceptible to antiviral (r) by the proportion of exposed cells (globally) in the system. In systems permitting constitutive immunity, a second rate of cellular acquisition of antiviral status (") additionally scaled with the global proportion of susceptible cells in the model. Compared with virus, IFN particles are small and highly diffusive, justifying this global signaling assumption at the limited spatial extent of a six-well plate and maintaining consistency with previous modeling approximations of IFN signaling in plaque assay (Howat et al., 2006) .
To best represent our empirical monolayer system, we expressed our state variables as proportions (P S , P A , P E , P I , and P D ), under assumptions of frequency-dependent transmission in a wellmixed population (Keeling and Rohani, 2008) , though note that the inclusion of P D (representing the proportion of dead space in the modeled tissue) had the functional effect of varying transmission with infectious cell density. This resulted in the following system of ordinary differential equations:
We defined 'induced immunity' as complete, modeling all cells as susceptible to viral invasion at disease-free equilibrium, with defenses induced subsequent to viral exposure through the term r. By contrast, we allowed the extent of constitutive immunity to vary across the parameter range of " > 0, defining a 'constitutive' system as one containing any antiviral cells at disease-free equilibrium. In fitting this model to tissue culture data, we independently estimated both r and "; as well as the cell-to-cell transmission rate, b, for each cell-virus combination. Since the extent to which constitutively-expressed IFN-a is constitutively translated into functional protein is not yet known for bat hosts (Zhou et al., 2016) , this approach permitted our tissue culture data to drive modeling inference: even in PaKiT01 cell lines known to constitutively express IFN-a, the true constitutive extent of the system (i.e. the quantity of antiviral cells present at disease-free equilibrium) was allowed to vary through estimation of ": For the purposes of model-fitting, we fixed the value of c, the return rate of antiviral cells to susceptible status, at 0. The small spatial scale and short time course (max 200 hours) of our experiments likely prohibited any return of antiviral cells to susceptible status in our empirical system; nonetheless, we retained the term c in analytical evaluations of our model because regression from antiviral to susceptible status is possible over long time periods in vitro and at the scale of a complete organism (Radke et al., 1974; Rasmussen and Farley, 1975; Samuel and Knutson, 1982) .
Before fitting to empirical time series, we undertook bifurcation analysis of our theoretical model and generated testable hypotheses on the basis of model outcomes. From our within-host model system (Equation 1-5), we derived the following expression for R 0 , the pathogen basic reproduction number (Supplementary file 2):
Pathogens can invade a host tissue culture when R 0 >1. Rapid rates of constitutive antiviral acquisition (") will drive R 0 <1: tissue cultures with highly constitutive antiviral immunity will be therefore resistant to virus invasion from the outset. Since, by definition, induced immunity is stimulated following initial virus invasion, the rate of induced antiviral acquisition (r) is not incorporated into the equation for R 0 ; while induced immune processes can control virus after initial invasion, they cannot prevent it from occurring to begin with. In cases of fully induced or absent immunity (" ¼ 0), the R 0 equation thus reduces to a form typical of the classic SEIR model:
At equilibrium, the theoretical, mean field model demonstrates one of three infection states: endemic equilibrium, stable limit cycles, or no infection ( Figure 2) . Respectively, these states approximate the persistent infection, virus-induced epidemic extinction, and immune-mediated epidemic extinction phenotypes previously witnessed in tissue culture experiments ( Figure 1 ). Theoretically, endemic equilibrium is maintained when new infections are generated at the same rate at which infections are lost, while limit cycles represent parameter space under which infectious and susceptible populations are locked in predictable oscillations. Endemic equilibria resulting from cellular regeneration (i.e. births) have been described in vivo for HIV (Coffin, 1995) and in vitro for herpesvirus plaque assays (Howat et al., 2006) , but, because they so closely approach zero, true limit cycles likely only occur theoretically, instead yielding stochastic extinctions in empirical time series.
Bifurcation analysis of our mean field model revealed that regions of no infection (pathogen extinction) were bounded at lower threshold (Branch point) values for b, below which the pathogen was unable to invade. We found no upper threshold to invasion for b under any circumstances (i.e. b high enough to drive pathogen-induced extinction), but high b values resulted in Hopf bifurcations, which delineate regions of parameter space characterized by limit cycles. Since limit cycles so closely approach zero, high bs recovered in this range would likely produce virus-induced epidemic extinctions under experimental conditions. Under more robust representations of immunity, with higher values for either or both induced (r) and constitutive (") rates of antiviral acquisition, Hopf bifurcations occurred at increasingly higher values for b, meaning that persistent infections could establish at higher viral transmission rates ( Figure 2 ). Consistent with our derivation for R 0 , we found that the Branch point threshold for viral invasion was independent of changes to the induced immune parameter (r) but saturated at high values of " that characterize highly constitutive immunity ( Figure 3) .
We next fit our theoretical model by least squares to each cell line-virus combination, under absent, induced, and constitutive assumptions of immunity. In general, best fit models recapitulated expected outcomes based on the immune phenotype of the cell line in question, as described in the general literature (Table 1 Ironically, the induced immune model offered a slightly better fit than the constitutive to rVSV-MARV infections on the PaKiT01 cell line (the one cell line-virus combination for which we know a constitutively antiviral cell-receptor incompatibility to be at play). Because constitutive immune assumptions can prohibit pathogen invasion (R 0 <1), model fits to this time series under constitutive assumptions were handicapped by overestimations of ", which prohibited pathogen invasion. Only by incorporating an exceedingly rapid rate of induced antiviral acquisition could the model guarantee that initial infection would be permitted and then rapidly controlled. In all panel (A) plots, the rate of induced immune antiviral acquisition (r) was fixed at 0.01. Panel (B) depicts dynamics under variably induced immunity, ranging from absent (left: r=0) to high (right: r=1). In all panel (B) plots, the rate of constitutive antiviral acquisition (") was fixed at 0.0001 Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycles, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3.
In fitting our theoretical model to in vitro data, we estimated the within-host virus transmission rate (b) and the rate(s) of cellular acquisition to antiviral status (r or r + ") ( Table 1 ; Supplementary file 4). Under absent immune assumptions, r and " were fixed at 0 while b was estimated; under induced immune assumptions, " was fixed at 0 while r and b were estimated; and under constitutive immune assumptions, all three parameters (r, ", and b) were simultaneously estimated for each cell-virus combination. Best fit parameter estimates for MOI=0.001 data are visualized in conjunction with br and b -" bifurcations in (r) and (B) the constitutive immunity rate of antiviral acquisition ("). Panels show variation in the extent of immunity, from absent (left) to high (right). Branch point curves are represented as solid lines and Hopf curves as dashed lines. White space indicates endemic equilibrium (persistence), gray space indicates limit cycling, and black space indicates no infection (extinction). Other parameter values for equilibrium analysis were fixed at: b = .025, m = .001, s = 1/6, a = 1/6, c = 0. Special points from bifurcations analyses are listed in Supplementary file 3. space corresponding to theoretical limit cycles, consistent with observed virus-induced epidemic extinctions in stochastic tissue cultures.
In contrast to Vero cells, the induced immunity model offered the best fit to all RoNi/7.1 data, consistent with reported patterns in the literature and our own validation by qPCR ( Table 1; Arnold et al., 2018; Kuzmin et al., 2017; Biesold et al., 2011; Pavlovich et al., 2018) . As in Vero cell trials, we estimated highest b values for rVSV-G infections on RoNi/7.1 cell lines but here recovered higher b estimates for rVSV-MARV than for rVSV-EBOV. This reversal was balanced by a higher estimated rate of acquisition to antiviral status (r) for rVSV-EBOV versus rVSV-MARV. In general, we observed that more rapid rates of antiviral acquisition (either induced, r, constitutive, ", or both) correlated with higher transmission rates (b). When offset by r, b values estimated for RoNi/7.1 infections maintained the same amplitude as those estimated for immune-absent Vero cell lines but caused gentler epidemics and reduced cellular mortality (Figure 1) . RoNi/7.1 parameter estimates localized in the region corresponding to endemic equilibrium for the deterministic, theoretical model (Figure 4) , yielding less acute epidemics which nonetheless went extinct in stochastic experiments.
Finally, rVSV-G and rVSV-EBOV trials on PaKiT01 cells were best fit by models assuming constitutive immunity, while rVSV-MARV infections on PaKiT01 were matched equivalently by models assuming either induced or constitutive immunity-with induced models favored over constitutive in AIC comparisons because one fewer parameter was estimated (Figure 1-figure supplements 4-5; Supplementary file 4). For all virus infections, PaKiT01 cell lines yielded b estimates a full order of magnitude higher than Vero or RoNi/7.1 cells, with each b balanced by an immune response (either r, or r combined with ") also an order of magnitude higher than that recovered for the other cell lines ( Figure 4 ; Table 1 ). As in RoNi/7.1 cells, PaKiT01 parameter fits localized in the region corresponding to endemic equilibrium for the deterministic theoretical model. Because constitutive immune processes can actually prohibit initial pathogen invasion, constitutive immune fits to rVSV-MARV infections on PaKiT01 cell lines consistently localized at or below the Branch point threshold for virus invasion (R 0 ¼ 1). During model fitting for optimization of ", any parameter tests of " values producing R 0 <1 resulted in no infection and, consequently, produced an exceedingly poor fit to infectious time series data. In all model fits assuming constitutive immunity, across all cell lines, antiviral contributions from " prohibited virus from invading at all. The induced immune model thus produced a more parsimonious recapitulation of these data because virus invasion was always permitted, then rapidly controlled.
In order to compare the relative contributions of each cell line's disparate immune processes to epidemic dynamics, we next used our mean field parameter estimates to calculate the initial 'antiviral rate'-the initial accumulation rate of antiviral cells upon virus invasion for each cell-virus-MOI combination-based on the following equation:
where P E was calculated from the initial infectious dose (MOI) of each infection experiment and P S was estimated at disease-free equilibrium:
Because and " both contribute to this initial antiviral rate, induced and constitutive immune assumptions are capable of yielding equally rapid rates, depending on parameter fits. Indeed, under fully induced immune assumptions, the induced antiviral acquisition rate (r) estimated for rVSV-MARV infection on PaKiT01 cells was so high that the initial antiviral rate exceeded even that estimated under constitutive assumptions for this cell-virus combination (Supplementary file 4) . In reality, we know that NPC1 receptor incompatibilities make PaKiT01 cell lines constitutively refractory to rVSV-MARV infection (Ng and Chandrab, 2018, Unpublished results) and that PaKiT01 cells also constitutively express the antiviral cytokine, IFN-a. Model fitting results suggest that this constitutive expression of IFN-a may act more as a rapidly inducible immune response following virus invasion than as a constitutive secretion of functional IFN-a protein. Nonetheless, as hypothesized, PaKiT01 cell lines were by far the most antiviral of any in our study-with initial antiviral rates estimated several orders of magnitude higher than any others in our study, under either induced or constitutive assumptions ( Table 1 ; Supplementary file 4). RoNi/7.1 cells displayed the second-most-pronounced signature of immunity, followed by Vero cells, for which the initial antiviral rate was essentially zero even under forced assumptions of induced or constitutive immunity ( Table 1 ; Supplementary file 4).
Using fitted parameters for b and ", we additionally calculated R 0 , the basic reproduction number for the virus, for each cell line-virus-MOI combination ( Table 1 ; Supplementary file 4). We found that R 0 was essentially unchanged across differing immune assumptions for RoNi/7.1 and Vero cells, for which the initial antiviral rate was low. In the case of PaKiT01 cells, a high initial antiviral rate under either induced or constitutive immunity resulted in a correspondingly high estimation of b (and, consequently, R 0 ) which still produced the same epidemic curve that resulted from the much lower estimates for b and R 0 paired with absent immunity. These findings suggest that antiviral immune responses protect host tissues against virus-induced cell mortality and may facilitate the establishment of more rapid within-host transmission rates.
Total monolayer destruction occurred in all cell-virus combinations excepting rVSV-EBOV infections on RoNi/7.1 cells and rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells. Monolayer destruction corresponded to susceptible cell depletion and epidemic turnover where R-effective (the product of R 0 and the proportion susceptible) was reduced below one ( Figure 5) . For rVSV-EBOV infections on RoNi/7.1, induced antiviral cells safeguarded remnant live cells, which birthed new susceptible cells late in the time series. In rVSV-EBOV and rVSV-MARV infections on PaKiT01 cells, this antiviral protection halted the epidemic ( Figure 5 ; R-effective <1) before susceptibles fully declined. In the case of rVSV-EBOV on PaKiT01, the birth of new susceptibles from remnant live cells protected by antiviral status maintained late-stage transmission to facilitate long-term epidemic persistence. Importantly, under fixed parameter values for the infection incubation rate (s) and infectioninduced mortality rate (a), models were unable to reproduce the longer-term infectious time series captured in data from rVSV-EBOV infections on PaKiT01 cell lines without incorporation of cell births, an assumption adopted in previous modeling representations of IFN-mediated viral dynamics in tissue culture (Howat et al., 2006) . In our experiments, we observed that cellular reproduction took place as plaque assays achieved confluency. Finally, because the protective effect of antiviral cells is more clearly observable spatially, we confirmed our results by simulating fitted time series in a spatially-explicit, stochastic reconstruction of our mean field model. In spatial simulations, rates of antiviral acquisition were fixed at fitted values for r and " derived from mean field estimates, while transmission rates (b) were fixed at values ten times greater than those estimated under mean field conditions, accounting for the intensification of parameter thresholds permitting pathogen invasion in local spatial interactions (see Materials and methods; Videos 1-3; Figure 5-figure supplement 3; Supplementary file 5; Webb et al., 2007) . In immune capable time series, spatial antiviral cells acted as 'refugia' which protected live cells from infection as each initial epidemic wave 'washed' across a cell monolayer. Eventual birth of new susceptibles from these living refugia allowed for sustained epidemic transmission in cases where some infectious cells persisted at later timepoints in simulation (Videos 1-3; Figure 5-figure supplement 3 ).
Bats are reservoirs for several important emerging zoonoses but appear not to experience disease from otherwise virulent viral pathogens. Though the molecular biological literature has made great progress in elucidating the mechanisms by which bats tolerate viral infections (Zhou et al., 2016; Ahn et al., 2019; Xie et al., 2018; Pavlovich et al., 2018; Zhang et al., 2013) , the impact of unique bat immunity on virus dynamics within-host has not been well-elucidated. We used an innovative combination of in vitro experimentation and within-host modeling to explore the impact of unique bat immunity on virus dynamics. Critically, we found that bat cell lines demonstrated a signature of enhanced interferon-mediated immune response, of either constitutive or induced form, which allowed for establishment of rapid within-host, cell-to-cell virus transmission rates (b). These results were supported by both data-independent bifurcation analysis of our mean field theoretical model, as well as fitting of this model to viral infection time series established in bat cell culture. Additionally, we demonstrated that the antiviral state induced by the interferon pathway protects live cells from mortality in tissue culture, resulting in in vitro epidemics of extended duration that enhance the probability of establishing a long-term persistent infection. Our findings suggest that viruses evolved in bat reservoirs possessing enhanced IFN capabilities could achieve more rapid within-host transmission rates without causing pathology to their hosts. Such rapidly-reproducing viruses would likely generate extreme virulence upon spillover to hosts lacking similar immune capacities to bats.
To achieve these results, we first developed a novel, within-host, theoretical model elucidating the effects of unique bat immunity, then undertook bifurcation analysis of the model's equilibrium properties under immune absent, induced, and constitutive assumptions. We considered a cell line to be constitutively immune if possessing any number of antiviral cells at disease-free equilibrium but allowed the extent of constitutive immunity to vary across the parameter range for ", the constitutive rate of antiviral acquisition. In deriving the equation for R 0 , the basic reproduction number, which defines threshold conditions for virus invasion of a tissue (R 0 >1), we demonstrated how the invasion threshold is elevated at high values of constitutive antiviral acquisition, ". Constitutive immune processes can thus prohibit pathogen invasion, while induced responses, by definition, can only control infections post-hoc. Once thresholds for pathogen invasion have been met, assumptions of constitutive immunity will limit the cellular mortality (virulence) incurred at high transmission rates. Regardless of mechanism (induced or constitutive), interferon-stimulated antiviral cells appear to play a key role in maintaining longer term or persistent infections by safeguarding susceptible cells from rapid infection and concomitant cell death. Fitting of our model to in vitro data supported expected immune phenotypes for different bat cell lines as described in the literature. Simple target cell models that ignore the effects of immunity best recapitulated infectious time series derived from IFN-deficient Vero cells, while models assuming induced immune processes most accurately reproduced trials derived from RoNi/7.1 (Rousettus aegyptiacus) cells, which possess a standard virusinduced IFN-response. In most cases, models assuming constitutive immune processes best recreated virus epidemics produced on PaKiT01 (Pteropus alecto) cells, which are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . Model support for induced immune assumptions in fits to rVSV-MARV infections on PaKiT01 cells suggests that the constitutive IFN-a expression characteristic of P. alecto cells may represent more of a constitutive immune priming process than a perpetual, functional, antiviral defense. Results from mean field model fitting were additionally confirmed in spatially explicit stochastic simulations of each time series.
As previously demonstrated in within-host models for HIV (Coffin, 1995; Perelson et al., 1996; Nowak et al., 1995; Bonhoeffer et al., 1997; Ho et al., 1995) , assumptions of simple target-cell depletion can often provide satisfactory approximations of viral dynamics, especially those reproduced in simple in vitro systems. Critically, our model fitting emphasizes the need for incorporation of top-down effects of immune control in order to accurately reproduce infectious time series derived from bat cell tissue cultures, especially those resulting from the robustly antiviral PaKiT01 P. alecto cell line. These findings indicate that enhanced IFN-mediated immune pathways in bat reservoirs may promote elevated within-host virus replication rates prior to cross-species emergence. We nonetheless acknowledge the limitations imposed by in vitro experiments in tissue culture, especially involving recombinant viruses and immortalized cell lines. Future work should extend these cell culture studies to include measurements of multiple state variables (i.e. antiviral cells) to enhance epidemiological inference.
The continued recurrence of Ebola epidemics across central Africa highlights the importance of understanding bats' roles as reservoirs for virulent zoonotic disease. The past decade has born witness to emerging consensus regarding the unique pathways by which bats resist and tolerate highly virulent infections (Brook and Dobson, 2015; Xie et al., 2018; Zhang et al., 2013; Ahn et al., 2019; Zhou et al., 2016; Ng et al., 2015; Pavlovich et al., 2018) . Nonetheless, an understanding of the mechanisms by which bats support endemic pathogens at the population level, or promote the evolution of virulent pathogens at the individual level, remains elusive. Endemic maintenance of infection is a defining characteristic of a pathogen reservoir (Haydon et al., 2002) , and bats appear to merit such a title, supporting long-term persistence of highly transmissible viral infections in isolated island populations well below expected critical community sizes (Peel et al., 2012) . Researchers debate the relative influence of population-level and within-host mechanisms which might explain these trends (Plowright et al., 2016) , but increasingly, field data are difficult to reconcile without acknowledgement of a role for persistent infections (Peel et al., 2018; Brook et al., 2019) . We present general methods to study cross-scale viral dynamics, which suggest that within-host persistence is supported by robust antiviral responses characteristic of bat immune processes. Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats.
All experiments were carried out on three immortalized mammalian kidney cell lines: Vero (African green monkey), RoNi/7.1 (Rousettus aegyptiacus) (Kühl et al., 2011; Biesold et al., 2011) and PaKiT01 (Pteropus alecto) (Crameri et al., 2009) . The species identifications of all bat cell lines was confirmed morphologically and genetically in the publications in which they were originally described (Kühl et al., 2011; Biesold et al., 2011; Crameri et al., 2009) . Vero cells were obtained from ATCC.
Monolayers of each cell line were grown to 90% confluency (~9Â10 5 cells) in 6-well plates. Cells were maintained in a humidified 37˚C, 5% CO 2 incubator and cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Grand Island, NY), supplemented with 2% fetal bovine serum (FBS) (Gemini Bio Products, West Sacramento, CA), and 1% penicillin-streptomycin (Life Technologies). Cells were tested monthly for mycoplasma contamination while experiments were taking place; all cells assayed negative for contamination at every testing.
Previous work has demonstrated that all cell lines used are capable of mounting a type I IFN response upon viral challenge, with the exception of Vero cells, which possess an IFN-b deficiency (Desmyter et al., 1968; Rhim et al., 1969; Emeny and Morgan, 1979) . RoNi/7.1 cells have been shown to mount idiosyncratic induced IFN defenses upon viral infection (Pavlovich et al., 2018; Kuzmin et al., 2017; Arnold et al., 2018; Kühl et al., 2011; Biesold et al., 2011) , while PaKiT01 cells are known to constitutively express the antiviral cytokine, IFN-a (Zhou et al., 2016) . This work is the first documentation of IFN signaling induced upon challenge with the particular recombinant VSVs outlined below. We verified known antiviral immune phenotypes via qPCR. Results were consistent with the literature, indicating a less pronounced role for interferon defense against viral infection in RoNi/7.1 versus PaKiT01 cells.
Replication-capable recombinant vesicular stomatitis Indiana viruses, expressing filovirus glycoproteins in place of wild type G (rVSV-G, rVSV-EBOV, and rVSV-MARV) have been previously described (Wong et al., 2010; Miller et al., 2012) . Viruses were selected to represent a broad range of anticipated antiviral responses from host cells, based on a range of past evolutionary histories between the virus glycoprotein mediating cell entry and the host cell's entry receptor. These interactions ranged from the total absence of evolutionary history in the case of rVSV-G infections on all cell lines to a known receptor-level cell entry incompatibility in the case of rVSV-MARV infections on PaKiT01 cell lines.
To measure infectivities of rVSVs on each of the cell lines outlined above, so as to calculate the correct viral dose for each MOI, NH 4 Cl (20 mM) was added to infected cell cultures at 1-2 hr postinfection to block viral spread, and individual eGFP-positive cells were manually counted at 12-14 hr post-infection.
Previously published work indicates that immortalized kidney cell lines of Rousettus aegyptiacus (RoNi/7.1) and Pteropus alecto (PaKiT01) exhibit different innate antiviral immune phenotypes through, respectively, induced (Biesold et al., 2011; Pavlovich et al., 2018; Kühl et al., 2011; Arnold et al., 2018) and constitutive (Zhou et al., 2016 ) expression of type I interferon genes. We verified these published phenotypes on our own cell lines infected with rVSV-G, rVSV-EBOV, and rVSV-MARV via qPCR of IFN-a and IFN-b genes across a longitudinal time series of infection.
Specifically, we carried out multiple time series of infection of each cell line with each of the viruses described above, under mock infection conditions and at MOIs of 0.0001 and 0.001-with the exception of rVSV-MARV on PaKiT01 cell lines, for which infection was only performed at MOI = 0.0001 due to limited viral stocks and the extremely low infectivity of this virus on this cell line (thus requiring high viral loads for initial infection). All experiments were run in duplicate on 6well plates, such that a typical plate for any of the three viruses had two control (mock) wells, two MOI = 0.0001 wells and two MOI = 0.001 wells, excepting PaKiT01 plates, which had two control and four MOI = 0.0001 wells at a given time. We justify this PaKiT01 exemption through the expectation that IFN-a expression is constitutive for these cells, and by the assumption that any expression exhibited at the lower MOI should also be present at the higher MOI.
For these gene expression time series, four 6-well plates for each cell line-virus combination were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with an agar plaque assay overlay to mimic conditions under which infection trials were run. Plates were then harvested sequentially at timepoints of roughly 5, 10, 15, and 20 hr post-infection (exact timing varied as multiple trials were running simultaneously). Upon harvest of each plate, agar overlay was removed, and virus was lysed and RNA extracted from cells using the Zymo Quick RNA Mini Prep kit, according to the manufacturer's instructions and including the step for cellular DNA digestion. Post-extraction, RNA quality was verified via nanodrop, and RNA was converted to cDNA using the Invitrogen Superscript III cDNA synthesis kit, according to the manufacturer's instructions. cDNA was then stored at 4˚C and as a frozen stock at À20˚C to await qPCR.
We undertook qPCR of cDNA to assess expression of the type I interferon genes, IFN-a and IFNb, and the housekeeping gene, b-Actin, using primers previously reported in the literature (Supplementary file 6) . For qPCR, 2 ml of each cDNA sample was incubated with 7 ml of deionized water, 1 ml of 5 UM forward/reverse primer mix and 10 ml of iTaq Universal SYBR Green, then cycled on a QuantStudio3 Real-Time PCR machine under the following conditions: initial denaturation at 94 C for 2 min followed by 40 cycles of: denaturation at 95˚C (5 s), annealing at 58˚C (15 s), and extension at 72˚C (10 s).
We report simple d-Ct values for each run, with raw Ct of the target gene of interest (IFN-a or IFN-b) subtracted from raw Ct of the b-Actin housekeeping gene in Figure 1 -figure supplement 6. Calculation of fold change upon viral infection in comparison to mock using the d-d-Ct method (Livak and Schmittgen, 2001) was inappropriate in this case, as we wished to demonstrate constitutive expression of IFN-a in PaKiT01 cells, whereby data from mock cells was identical to that produced from infected cells.
After being grown to~90% confluency, cells were incubated with pelleted rVSVs expressing eGFP (rVSV-G, rVSV-EBOV, rVSV-MARV). Cell lines were challenged with both a low (0.0001) and high (0.001) multiplicity of infection (MOI) for each virus. In a cell monolayer infected at a given MOI (m), the proportion of cells (P), infected by k viral particles can be described by the Poisson distribution: P k ð Þ ¼ e Àm m k k! , such that the number of initially infected cells in an experiment equals: 1 À e Àm . We assumed that a~90% confluent culture at each trial's origin was comprised of~9x10 5 cells and conducted all experiments at MOIs of 0.0001 and 0.001, meaning that we began each trial by introducing virus to, respectively,~81 or 810 cells, representing the state variable 'E' in our theoretical model. Low MOIs were selected to best approximate the dynamics of mean field infection and limit artifacts of spatial structuring, such as premature epidemic extinction when growing plaques collide with plate walls in cell culture.
Six-well plates were prepared with each infection in duplicate or triplicate, such that a control well (no virus) and 2-3 wells each at MOI 0.001 and 0.0001 were incubated simultaneously on the same plate. In total, we ran between 18 and 39 trials at each cell-virus-MOI combination, excepting r-VSV-MARV infections on PaKiT01 cells at MOI = 0.001, for which we ran only eight trials due to the low infectivity of this virus on this cell line, which required high viral loads for initial infection. Cells were incubated with virus for one hour at 37˚C. Following incubation, virus was aspirated off, and cell monolayers were washed in PBS, then covered with a molten viscous overlay (50% 2X MEM/Lglutamine; 5% FBS; 3% HEPES; 42% agarose), cooled for 20 min, and re-incubated in their original humidified 37˚C, 5% CO 2 environment.
After application of the overlay, plates were monitored periodically using an inverted fluorescence microscope until the first signs of GFP expression were witnessed (~6-9.5 hr post-infection, depending on the cell line and virus under investigation). From that time forward, a square subset of the center of each well (comprised of either 64-or 36-subframes and corresponding to roughly 60% and 40% of the entire well space) was imaged periodically, using a CellInsight CX5 High Content Screening (HCS) Platform with a 4X air objective (ThermoFisher, Inc, Waltham, MA). Microscope settings were held standard across all trials, with exposure time fixed at 0.0006 s for each image. One color channel was imaged, such that images produced show GFP-expressing cells in white and non-GFP-expressing cells in black (Figure 1-figure supplement 1) .
Wells were photographed in rotation, as frequently as possible, from the onset of GFP expression until the time that the majority of cells in the well were surmised to be dead, GFP expression could no longer be detected, or early termination was desired to permit Hoechst staining.
In the case of PaKiT01 cells infected with rVSV-EBOV, where an apparently persistent infection established, the assay was terminated after 200+ hours (8+ days) of continuous observation. Upon termination of all trials, cells were fixed in formaldehyde (4% for 15 min), incubated with Hoechst stain (0.0005% for 15 min) (ThermoFisher, Inc, Waltham, MA), then imaged at 4X on the CellInsight CX5 High Content Screening (HCS) Platform. The machine was allowed to find optimal focus for each Hoechst stain image. One color channel was permitted such that images produced showed live nuclei in white and dead cells in black.
Hoechst stain colors cellular DNA, and viral infection is thought to interfere with the clarity of the stain (Dembowski and DeLuca, 2015) . As such, infection termination, cell fixation, and Hoechst staining enables generation of a rough time series of uninfectious live cells (i.e. susceptible + antiviral cells) to complement the images which produced time series of proportions infectious. Due to uncertainty over the exact epidemic state of Hoechst-stained cells (i.e. exposed but not yet infectious cells may still stain), we elected to fit our models only to the infectious time series derived from GFPexpressing images and used Hoechst stain images as a post hoc visual check on our fit only ( Figure 5 ; Figure 5 -figure supplements 1-2).
Images recovered from the time series above were processed into binary ('infectious' vs. 'non-infectious' or, for Hoechst-stained images, 'live' vs. 'dead') form using the EBImage package (Pau et al., 2010) in R version 3.6 for MacIntosh, after methods further detailed in Supplementary file 7. Binary images were then further processed into time series of infectious or, for Hoechst-stained images, live cells using a series of cell counting scripts. Because of logistical constraints (i.e. many plates of simultaneously running infection trials and only one available imaging microscope), the time course of imaging across the duration of each trial was quite variable. As such, we fitted a series of statistical models to our processed image data to reconstruct reliable values of the infectious proportion of each well per hour for each distinct trial in all cell line-virus-MOI combinations (Figure 1
To derive the expression for R 0 , the basic pathogen reproductive number in vitro, we used Next Generation Matrix (NGM) techniques (Diekmann et al., 1990; Heffernan et al., 2005) , employing Wolfram Mathematica (version 11.2) as an analytical tool. R 0 describes the number of new infections generated by an existing infection in a completely susceptible host population; a pathogen will invade a population when R 0 >1 (Supplementary file 2). We then analyzed stability properties of the system, exploring dynamics across a range of parameter spaces, using MatCont (version 2.2) (Dhooge et al., 2008) for Matlab (version R2018a) (Supplementary file 3).
The birth rate, b, and natural mortality rate, m, balance to yield a population-level growth rate, such that it is impossible to estimate both b and m simultaneously from total population size data alone. As such, we fixed b at. 025 and estimated m by fitting an infection-absent version of our mean field model to the susceptible time series derived via Hoechst staining of control wells for each of the three cell lines (Figure 1-figure supplement 7) . This yielded a natural mortality rate, m, corresponding to a lifespan of approximately 121, 191, and 84 hours, respectively, for Vero, RoNi/7.1, and PaKiT01 cell lines (Figure 1-figure supplement 7) . We then fixed the virus incubation rate, s, as the inverse of the shortest observed duration of time from initial infection to the observation of the first infectious cells via fluorescent microscope for all nine cell line -virus combinations (ranging 6 to 9.5 hours). We fixed a, the infection-induced mortality rate, at 1/6, an accepted standard for general viral kinetics (Howat et al., 2006) , and held c, the rate of antiviral cell regression to susceptible status, at 0 for the timespan (<200 hours) of the experimental cell line infection trials.
We estimated cell line-virus-MOI-specific values for b, r, and " by fitting the deterministic output of infectious proportions in our mean field model to the full suite of statistical outputs of all trials for each infected cell culture time series (Figure 1-figure supplements 2-3) . Fitting was performed by minimizing the sum of squared differences between the deterministic model output and cell linevirus-MOI-specific infectious proportion of the data at each timestep. We optimized parameters for MOI = 0.001 and 0.0001 simultaneously to leverage statistical power across the two datasets, estimating a different transmission rate, b, for trials run at each infectious dose but, where applicable, estimating the same rates of r and " across the two time series. We used the differential equation solver lsoda() in the R package deSolve (Soetaert et al., 2010) to obtain numerical solutions for the mean field model and carried out minimization using the 'Nelder-Mead' algorithm of the optim() function in base R. All model fits were conducted using consistent starting guesses for the parameters, b (b = 3), and where applicable, r (r = 0.001) and " (" = 0.001). In the case of failed fits or indefinite hessians, we generated a series of random guesses around the starting conditions and continued estimation until successful fits were achieved.
All eighteen cell line-virus-MOI combinations of data were fit by an immune absent (" = r = 0) version of the theoretical model and, subsequently, an induced immunity (" = 0; r >0) and constitutive immunity (" >0; r >0) version of the model. Finally, we compared fits across each cell line-virus-MOI combination via AIC. In calculating AIC, the number of fitted parameters in each model (k) varied across the immune phenotypes, with one parameter (b) estimated for absent immune assumptions, two (b and r) for induced immune assumptions, and three (b, r, and ") for constitutive immune assumptions. The sample size (n) corresponded to the number of discrete time steps across all empirical infectious trials to which the model was fitted for each cell-line virus combination. All fitting and model comparison scripts are freely available for download at the following FigShare repository: DOI: 10.6084/m9.figshare.8312807.
Finally, we verified all mean field fits in a spatial context, in order to more thoroughly elucidate the role of antiviral cells in each time series. We constructed our spatial model in C++ implemented in R using the packages Rcpp and RcppArmadillo (Eddelbuettel and Francois, 2011; Eddelbuettel and Sanderson, 2017) . Following Nagai and Honda (2001) and Howat et al. (2006) , we modeled this system on a two-dimensional hexagonal lattice, using a ten-minute epidemic timestep for cell state transitions. At the initialization of each simulation, we randomly assigned a duration of natural lifespan, incubation period, infectivity period, and time from antiviral to susceptible status to all cells in a theoretical monolayer. Parameter durations were drawn from a normal distribution centered at the inverse of the respective fixed rates of m, s, a, and c, as reported with our mean field model. Transitions involving the induced (r) and constitutive (") rates of antiviral acquisition were governed probabilistically and adjusted dynamically at each timestep based on the global environment. As such, we fixed these parameters at the same values estimated in the mean field model, and multiplied both r and " by the global proportion of, respectively, exposed and susceptible cells at a given timestep.
In contrast to antiviral acquisition rates, transitions involving the birth rate (b) and the transmission rate (b) occurred probabilistically based on each cell's local environment. The birth rate, b, was multiplied by the proportion of susceptible cells within a six-neighbor circumference of a focal dead cell, while b was multiplied by the proportion of infectious cells within a thirty-six neighbor vicinity of a focal susceptible cell, thus allowing viral transmission to extend beyond the immediate nearestneighbor boundaries of an infectious cell. To compensate for higher thresholds to cellular persistence and virus invasion which occur under local spatial conditions (Webb et al., 2007) , we increased the birth rate, b, and the cell-to-cell transmission rate, b, respectively, to six and ten times the values used in the mean field model (Supplementary file 4) . We derived these increases based on the assumption that births took place exclusively based on pairwise nearest-neighbor interactions (the six immediately adjacent cells to a focal dead cell), while viral transmission was locally concentrated but included a small (7.5%) global contribution, representing the thirty-six cell surrounding vicinity of a focal susceptible. We justify these increases and derive their origins further in Supplementary file 5.
We simulated ten stochastic spatial time series for all cell-virus combinations under all three immune assumptions at a population size of 10,000 cells and compared model output with data in . Transparent reporting form Data availability All data generated or analysed during this study are included in the manuscript and supporting files. All images and code used in this study have been made available for download at the following Figshare | What is a conclusion of this study? | 2,753 | Viruses which evolve rapid replication rates under these robust antiviral defenses may pose the greatest hazard for cross-species pathogen emergence into spillover hosts with immune systems that differ from those unique to bats. | 37,910 |
185 | CDC Summary 21 MAR 2020,
https://www.cdc.gov/coronavirus/2019-ncov/cases-updates/summary.html
This is a rapidly evolving situation and CDC will provide updated information and guidance as it becomes available.
Updated March 21, 2020
CDC is responding to a pandemic of respiratory disease spreading from person-to-person caused by a novel (new) coronavirus. The disease has been named “coronavirus disease 2019” (abbreviated “COVID-19”). This situation poses a serious public health risk. The federal government is working closely with state, local, tribal, and territorial partners, as well as public health partners, to respond to this situation. COVID-19 can cause mild to severe illness; most severe illness occurs in older adults.
Situation in U.S.
Different parts of the country are seeing different levels of COVID-19 activity. The United States nationally is in the initiation phase of the pandemic. States in which community spread is occurring are in the acceleration phase. The duration and severity of each pandemic phase can vary depending on the characteristics of the virus and the public health response.
CDC and state and local public health laboratories are testing for the virus that causes COVID-19. View CDC’s Public Health Laboratory Testing map.
All 50 states have reported cases of COVID-19 to CDC.
U.S. COVID-19 cases include:
Imported cases in travelers
Cases among close contacts of a known case
Community-acquired cases where the source of the infection is unknown.
Twenty-seven U.S. states are reporting some community spread of COVID-19.
View latest case counts, deaths, and a map of states with reported cases.
CDC Recommends
Everyone can do their part to help us respond to this emerging public health threat:
On March 16, the White House announced a program called “15 Days to Slow the Spread,”pdf iconexternal icon which is a nationwide effort to slow the spread of COVID-19 through the implementation of social distancing at all levels of society.
Older people and people with severe chronic conditions should take special precautions because they are at higher risk of developing serious COVID-19 illness.
If you are a healthcare provider, use your judgment to determine if a patient has signs and symptoms compatible with COVID-19 and whether the patient should be tested. Factors to consider in addition to clinical symptoms may include:
Does the patient have recent travel from an affected area?
Has the patient been in close contact with someone with COVID-19 or with patients with pneumonia of unknown cause?
Does the patient reside in an area where there has been community spread of COVID-19?
If you are a healthcare provider or a public health responder caring for a COVID-19 patient, please take care of yourself and follow recommended infection control procedures.
People who get a fever or cough should consider whether they might have COVID-19, depending on where they live, their travel history or other exposures. More than half of the U.S. is seeing some level of community spread of COVID-19. Testing for COVID-19 may be accessed through medical providers or public health departments, but there is no treatment for this virus. Most people have mild illness and are able to recover at home without medical care.
For people who are ill with COVID-19, but are not sick enough to be hospitalized, please follow CDC guidance on how to reduce the risk of spreading your illness to others. People who are mildly ill with COVID-19 are able to isolate at home during their illness.
If you have been in China or another affected area or have been exposed to someone sick with COVID-19 in the last 14 days, you will face some limitations on your movement and activity. Please follow instructions during this time. Your cooperation is integral to the ongoing public health response to try to slow spread of this virus.
COVID-19 Emergence
COVID-19 is caused by a coronavirus. Coronaviruses are a large family of viruses that are common in people and many different species of animals, including camels, cattle, cats, and bats. Rarely, animal coronaviruses can infect people and then spread between people such as with MERS-CoV, SARS-CoV, and now with this new virus (named SARS-CoV-2).
The SARS-CoV-2 virus is a betacoronavirus, like MERS-CoV and SARS-CoV. All three of these viruses have their origins in bats. The sequences from U.S. patients are similar to the one that China initially posted, suggesting a likely single, recent emergence of this virus from an animal reservoir.
Early on, many of the patients at the epicenter of the outbreak in Wuhan, Hubei Province, China had some link to a large seafood and live animal market, suggesting animal-to-person spread. Later, a growing number of patients reportedly did not have exposure to animal markets, indicating person-to-person spread. Person-to-person spread was subsequently reported outside Hubei and in countries outside China, including in the United States. Some international destinations now have ongoing community spread with the virus that causes COVID-19, as do some parts of the United States. Community spread means some people have been infected and it is not known how or where they became exposed. Learn more about the spread of this newly emerged coronavirus.
Severity
The complete clinical picture with regard to COVID-19 is not fully known. Reported illnesses have ranged from very mild (including some with no reported symptoms) to severe, including illness resulting in death. While information so far suggests that most COVID-19 illness is mild, a reportexternal icon out of China suggests serious illness occurs in 16% of cases. Older people and people of all ages with severe chronic medical conditions — like heart disease, lung disease and diabetes, for example — seem to be at higher risk of developing serious COVID-19 illness. A CDC Morbidity & Mortality Weekly Report that looked at severity of disease among COVID-19 cases in the United States by age group found that 80% of deaths were among adults 65 years and older with the highest percentage of severe outcomes occurring in people 85 years and older.
Learn more about the symptoms associated with COVID-19.
COVID-19 Pandemic
A pandemic is a global outbreak of disease. Pandemics happen when a new virus emerges to infect people and can spread between people sustainably. Because there is little to no pre-existing immunity against the new virus, it spreads worldwide.
The virus that causes COVID-19 is infecting people and spreading easily from person-to-person. Cases have been detected in most countries worldwide and community spread is being detected in a growing number of countries. On March 11, the COVID-19 outbreak was characterized as a pandemic by the WHOexternal icon.
This is the first pandemic known to be caused by the emergence of a new coronavirus. In the past century, there have been four pandemics caused by the emergence of novel influenza viruses. As a result, most research and guidance around pandemics is specific to influenza, but the same premises can be applied to the current COVID-19 pandemic. Pandemics of respiratory disease follow a certain progression outlined in a “Pandemic Intervals Framework.” Pandemics begin with an investigation phase, followed by recognition, initiation, and acceleration phases. The peak of illnesses occurs at the end of the acceleration phase, which is followed by a deceleration phase, during which there is a decrease in illnesses. Different countries can be in different phases of the pandemic at any point in time and different parts of the same country can also be in different phases of a pandemic.
There are ongoing investigations to learn more. This is a rapidly evolving situation and information will be updated as it becomes available.
Risk Assessment
Risk depends on characteristics of the virus, including how well it spreads between people; the severity of resulting illness; and the medical or other measures available to control the impact of the virus (for example, vaccines or medications that can treat the illness) and the relative success of these. In the absence of vaccine or treatment medications, nonpharmaceutical interventions become the most important response strategy. These are community interventions that can reduce the impact of disease.
The risk from COVID-19 to Americans can be broken down into risk of exposure versus risk of serious illness and death.
Risk of exposure:
The immediate risk of being exposed to this virus is still low for most Americans, but as the outbreak expands, that risk will increase. Cases of COVID-19 and instances of community spread are being reported in a growing number of states.
People in places where ongoing community spread of the virus that causes COVID-19 has been reported are at elevated risk of exposure, with the level of risk dependent on the location.
Healthcare workers caring for patients with COVID-19 are at elevated risk of exposure.
Close contacts of persons with COVID-19 also are at elevated risk of exposure.
Travelers returning from affected international locations where community spread is occurring also are at elevated risk of exposure, with level of risk dependent on where they traveled.
Risk of Severe Illness:
Early information out of China, where COVID-19 first started, shows that some people are at higher risk of getting very sick from this illness. This includes:
Older adults, with risk increasing by age.
People who have serious chronic medical conditions like:
Heart disease
Diabetes
Lung disease
CDC has developed guidance to help in the risk assessment and management of people with potential exposures to COVID-19.
What May Happen
More cases of COVID-19 are likely to be identified in the United States in the coming days, including more instances of community spread. CDC expects that widespread transmission of COVID-19 in the United States will occur. In the coming months, most of the U.S. population will be exposed to this virus.
Widespread transmission of COVID-19 could translate into large numbers of people needing medical care at the same time. Schools, childcare centers, and workplaces, may experience more absenteeism. Mass gatherings may be sparsely attended or postponed. Public health and healthcare systems may become overloaded, with elevated rates of hospitalizations and deaths. Other critical infrastructure, such as law enforcement, emergency medical services, and sectors of the transportation industry may also be affected. Healthcare providers and hospitals may be overwhelmed. At this time, there is no vaccine to protect against COVID-19 and no medications approved to treat it. Nonpharmaceutical interventions will be the most important response strategy to try to delay the spread of the virus and reduce the impact of disease.
CDC Response
Global efforts at this time are focused concurrently on lessening the spread and impact of this virus. The federal government is working closely with state, local, tribal, and territorial partners, as well as public health partners, to respond to this public health threat.
Highlights of CDC’s Response
CDC established a COVID-19 Incident Management System on January 7, 2020. On January 21, CDC activated its Emergency Operations Center to better provide ongoing support to the COVID-19 response.
The U.S. government has taken unprecedented steps with respect to travel in response to the growing public health threat posed by this new coronavirus:
Foreign nationals who have been in China, Iran, the United Kingdom, Ireland and any one of the 26 European countries in the Schengen Area within the past 14 days cannot enter the United States.
U.S. citizens, residents, and their immediate family members who have been any one of those countries within in the past 14 days can enter the United States, but they are subject to health monitoring and possible quarantine for up to 14 days.
People at higher risk of serious COVID-19 illness avoid cruise travel and non-essential air travel.
CDC has issued additional specific travel guidance related to COVID-19.
CDC has issued clinical guidance, including:
Clinical Guidance for Management of Patients with Confirmed Coronavirus Disease (COVID-19).
Infection Prevention and Control Recommendations for Patients, including guidance on the use of personal protective equipment (PPE) during a shortage.
CDC also has issued guidance for other settings, including:
Preparing for COVID-19: Long-term Care Facilities, Nursing Homes
Discontinuation of Home Isolation for Persons with COVID-19
CDC has deployed multidisciplinary teams to support state health departments in case identification, contact tracing, clinical management, and public communications.
CDC has worked with federal partners to support the safe return of Americans overseas who have been affected by COVID-19.
An important part of CDC’s role during a public health emergency is to develop a test for the pathogen and equip state and local public health labs with testing capacity.
CDC developed an rRT-PCR test to diagnose COVID-19.
As of the evening of March 17, 89 state and local public health labs in 50 states, the District of Columbia, Guam, and Puerto Rico have successfully verified and are currently using CDC COVID-19 diagnostic tests.
Commercial manufacturers are now producing their own tests.
CDC has grown the COVID-19 virus in cell culture, which is necessary for further studies, including for additional genetic characterization. The cell-grown virus was sent to NIH’s BEI Resources Repositoryexternal icon for use by the broad scientific community.
CDC also is developing a serology test for COVID-19.
Other Available Resources
The following resources are available with information on COVID-19
World Health Organization, Coronavirusexternal icon | What age group has the highest rate of severe outcomes? | 236 | people 85 years and older | 6,117 |
185 | CDC Summary 21 MAR 2020,
https://www.cdc.gov/coronavirus/2019-ncov/cases-updates/summary.html
This is a rapidly evolving situation and CDC will provide updated information and guidance as it becomes available.
Updated March 21, 2020
CDC is responding to a pandemic of respiratory disease spreading from person-to-person caused by a novel (new) coronavirus. The disease has been named “coronavirus disease 2019” (abbreviated “COVID-19”). This situation poses a serious public health risk. The federal government is working closely with state, local, tribal, and territorial partners, as well as public health partners, to respond to this situation. COVID-19 can cause mild to severe illness; most severe illness occurs in older adults.
Situation in U.S.
Different parts of the country are seeing different levels of COVID-19 activity. The United States nationally is in the initiation phase of the pandemic. States in which community spread is occurring are in the acceleration phase. The duration and severity of each pandemic phase can vary depending on the characteristics of the virus and the public health response.
CDC and state and local public health laboratories are testing for the virus that causes COVID-19. View CDC’s Public Health Laboratory Testing map.
All 50 states have reported cases of COVID-19 to CDC.
U.S. COVID-19 cases include:
Imported cases in travelers
Cases among close contacts of a known case
Community-acquired cases where the source of the infection is unknown.
Twenty-seven U.S. states are reporting some community spread of COVID-19.
View latest case counts, deaths, and a map of states with reported cases.
CDC Recommends
Everyone can do their part to help us respond to this emerging public health threat:
On March 16, the White House announced a program called “15 Days to Slow the Spread,”pdf iconexternal icon which is a nationwide effort to slow the spread of COVID-19 through the implementation of social distancing at all levels of society.
Older people and people with severe chronic conditions should take special precautions because they are at higher risk of developing serious COVID-19 illness.
If you are a healthcare provider, use your judgment to determine if a patient has signs and symptoms compatible with COVID-19 and whether the patient should be tested. Factors to consider in addition to clinical symptoms may include:
Does the patient have recent travel from an affected area?
Has the patient been in close contact with someone with COVID-19 or with patients with pneumonia of unknown cause?
Does the patient reside in an area where there has been community spread of COVID-19?
If you are a healthcare provider or a public health responder caring for a COVID-19 patient, please take care of yourself and follow recommended infection control procedures.
People who get a fever or cough should consider whether they might have COVID-19, depending on where they live, their travel history or other exposures. More than half of the U.S. is seeing some level of community spread of COVID-19. Testing for COVID-19 may be accessed through medical providers or public health departments, but there is no treatment for this virus. Most people have mild illness and are able to recover at home without medical care.
For people who are ill with COVID-19, but are not sick enough to be hospitalized, please follow CDC guidance on how to reduce the risk of spreading your illness to others. People who are mildly ill with COVID-19 are able to isolate at home during their illness.
If you have been in China or another affected area or have been exposed to someone sick with COVID-19 in the last 14 days, you will face some limitations on your movement and activity. Please follow instructions during this time. Your cooperation is integral to the ongoing public health response to try to slow spread of this virus.
COVID-19 Emergence
COVID-19 is caused by a coronavirus. Coronaviruses are a large family of viruses that are common in people and many different species of animals, including camels, cattle, cats, and bats. Rarely, animal coronaviruses can infect people and then spread between people such as with MERS-CoV, SARS-CoV, and now with this new virus (named SARS-CoV-2).
The SARS-CoV-2 virus is a betacoronavirus, like MERS-CoV and SARS-CoV. All three of these viruses have their origins in bats. The sequences from U.S. patients are similar to the one that China initially posted, suggesting a likely single, recent emergence of this virus from an animal reservoir.
Early on, many of the patients at the epicenter of the outbreak in Wuhan, Hubei Province, China had some link to a large seafood and live animal market, suggesting animal-to-person spread. Later, a growing number of patients reportedly did not have exposure to animal markets, indicating person-to-person spread. Person-to-person spread was subsequently reported outside Hubei and in countries outside China, including in the United States. Some international destinations now have ongoing community spread with the virus that causes COVID-19, as do some parts of the United States. Community spread means some people have been infected and it is not known how or where they became exposed. Learn more about the spread of this newly emerged coronavirus.
Severity
The complete clinical picture with regard to COVID-19 is not fully known. Reported illnesses have ranged from very mild (including some with no reported symptoms) to severe, including illness resulting in death. While information so far suggests that most COVID-19 illness is mild, a reportexternal icon out of China suggests serious illness occurs in 16% of cases. Older people and people of all ages with severe chronic medical conditions — like heart disease, lung disease and diabetes, for example — seem to be at higher risk of developing serious COVID-19 illness. A CDC Morbidity & Mortality Weekly Report that looked at severity of disease among COVID-19 cases in the United States by age group found that 80% of deaths were among adults 65 years and older with the highest percentage of severe outcomes occurring in people 85 years and older.
Learn more about the symptoms associated with COVID-19.
COVID-19 Pandemic
A pandemic is a global outbreak of disease. Pandemics happen when a new virus emerges to infect people and can spread between people sustainably. Because there is little to no pre-existing immunity against the new virus, it spreads worldwide.
The virus that causes COVID-19 is infecting people and spreading easily from person-to-person. Cases have been detected in most countries worldwide and community spread is being detected in a growing number of countries. On March 11, the COVID-19 outbreak was characterized as a pandemic by the WHOexternal icon.
This is the first pandemic known to be caused by the emergence of a new coronavirus. In the past century, there have been four pandemics caused by the emergence of novel influenza viruses. As a result, most research and guidance around pandemics is specific to influenza, but the same premises can be applied to the current COVID-19 pandemic. Pandemics of respiratory disease follow a certain progression outlined in a “Pandemic Intervals Framework.” Pandemics begin with an investigation phase, followed by recognition, initiation, and acceleration phases. The peak of illnesses occurs at the end of the acceleration phase, which is followed by a deceleration phase, during which there is a decrease in illnesses. Different countries can be in different phases of the pandemic at any point in time and different parts of the same country can also be in different phases of a pandemic.
There are ongoing investigations to learn more. This is a rapidly evolving situation and information will be updated as it becomes available.
Risk Assessment
Risk depends on characteristics of the virus, including how well it spreads between people; the severity of resulting illness; and the medical or other measures available to control the impact of the virus (for example, vaccines or medications that can treat the illness) and the relative success of these. In the absence of vaccine or treatment medications, nonpharmaceutical interventions become the most important response strategy. These are community interventions that can reduce the impact of disease.
The risk from COVID-19 to Americans can be broken down into risk of exposure versus risk of serious illness and death.
Risk of exposure:
The immediate risk of being exposed to this virus is still low for most Americans, but as the outbreak expands, that risk will increase. Cases of COVID-19 and instances of community spread are being reported in a growing number of states.
People in places where ongoing community spread of the virus that causes COVID-19 has been reported are at elevated risk of exposure, with the level of risk dependent on the location.
Healthcare workers caring for patients with COVID-19 are at elevated risk of exposure.
Close contacts of persons with COVID-19 also are at elevated risk of exposure.
Travelers returning from affected international locations where community spread is occurring also are at elevated risk of exposure, with level of risk dependent on where they traveled.
Risk of Severe Illness:
Early information out of China, where COVID-19 first started, shows that some people are at higher risk of getting very sick from this illness. This includes:
Older adults, with risk increasing by age.
People who have serious chronic medical conditions like:
Heart disease
Diabetes
Lung disease
CDC has developed guidance to help in the risk assessment and management of people with potential exposures to COVID-19.
What May Happen
More cases of COVID-19 are likely to be identified in the United States in the coming days, including more instances of community spread. CDC expects that widespread transmission of COVID-19 in the United States will occur. In the coming months, most of the U.S. population will be exposed to this virus.
Widespread transmission of COVID-19 could translate into large numbers of people needing medical care at the same time. Schools, childcare centers, and workplaces, may experience more absenteeism. Mass gatherings may be sparsely attended or postponed. Public health and healthcare systems may become overloaded, with elevated rates of hospitalizations and deaths. Other critical infrastructure, such as law enforcement, emergency medical services, and sectors of the transportation industry may also be affected. Healthcare providers and hospitals may be overwhelmed. At this time, there is no vaccine to protect against COVID-19 and no medications approved to treat it. Nonpharmaceutical interventions will be the most important response strategy to try to delay the spread of the virus and reduce the impact of disease.
CDC Response
Global efforts at this time are focused concurrently on lessening the spread and impact of this virus. The federal government is working closely with state, local, tribal, and territorial partners, as well as public health partners, to respond to this public health threat.
Highlights of CDC’s Response
CDC established a COVID-19 Incident Management System on January 7, 2020. On January 21, CDC activated its Emergency Operations Center to better provide ongoing support to the COVID-19 response.
The U.S. government has taken unprecedented steps with respect to travel in response to the growing public health threat posed by this new coronavirus:
Foreign nationals who have been in China, Iran, the United Kingdom, Ireland and any one of the 26 European countries in the Schengen Area within the past 14 days cannot enter the United States.
U.S. citizens, residents, and their immediate family members who have been any one of those countries within in the past 14 days can enter the United States, but they are subject to health monitoring and possible quarantine for up to 14 days.
People at higher risk of serious COVID-19 illness avoid cruise travel and non-essential air travel.
CDC has issued additional specific travel guidance related to COVID-19.
CDC has issued clinical guidance, including:
Clinical Guidance for Management of Patients with Confirmed Coronavirus Disease (COVID-19).
Infection Prevention and Control Recommendations for Patients, including guidance on the use of personal protective equipment (PPE) during a shortage.
CDC also has issued guidance for other settings, including:
Preparing for COVID-19: Long-term Care Facilities, Nursing Homes
Discontinuation of Home Isolation for Persons with COVID-19
CDC has deployed multidisciplinary teams to support state health departments in case identification, contact tracing, clinical management, and public communications.
CDC has worked with federal partners to support the safe return of Americans overseas who have been affected by COVID-19.
An important part of CDC’s role during a public health emergency is to develop a test for the pathogen and equip state and local public health labs with testing capacity.
CDC developed an rRT-PCR test to diagnose COVID-19.
As of the evening of March 17, 89 state and local public health labs in 50 states, the District of Columbia, Guam, and Puerto Rico have successfully verified and are currently using CDC COVID-19 diagnostic tests.
Commercial manufacturers are now producing their own tests.
CDC has grown the COVID-19 virus in cell culture, which is necessary for further studies, including for additional genetic characterization. The cell-grown virus was sent to NIH’s BEI Resources Repositoryexternal icon for use by the broad scientific community.
CDC also is developing a serology test for COVID-19.
Other Available Resources
The following resources are available with information on COVID-19
World Health Organization, Coronavirusexternal icon | How is COVID-19 spread? | 225 | person-to-person | 306 |
185 | CDC Summary 21 MAR 2020,
https://www.cdc.gov/coronavirus/2019-ncov/cases-updates/summary.html
This is a rapidly evolving situation and CDC will provide updated information and guidance as it becomes available.
Updated March 21, 2020
CDC is responding to a pandemic of respiratory disease spreading from person-to-person caused by a novel (new) coronavirus. The disease has been named “coronavirus disease 2019” (abbreviated “COVID-19”). This situation poses a serious public health risk. The federal government is working closely with state, local, tribal, and territorial partners, as well as public health partners, to respond to this situation. COVID-19 can cause mild to severe illness; most severe illness occurs in older adults.
Situation in U.S.
Different parts of the country are seeing different levels of COVID-19 activity. The United States nationally is in the initiation phase of the pandemic. States in which community spread is occurring are in the acceleration phase. The duration and severity of each pandemic phase can vary depending on the characteristics of the virus and the public health response.
CDC and state and local public health laboratories are testing for the virus that causes COVID-19. View CDC’s Public Health Laboratory Testing map.
All 50 states have reported cases of COVID-19 to CDC.
U.S. COVID-19 cases include:
Imported cases in travelers
Cases among close contacts of a known case
Community-acquired cases where the source of the infection is unknown.
Twenty-seven U.S. states are reporting some community spread of COVID-19.
View latest case counts, deaths, and a map of states with reported cases.
CDC Recommends
Everyone can do their part to help us respond to this emerging public health threat:
On March 16, the White House announced a program called “15 Days to Slow the Spread,”pdf iconexternal icon which is a nationwide effort to slow the spread of COVID-19 through the implementation of social distancing at all levels of society.
Older people and people with severe chronic conditions should take special precautions because they are at higher risk of developing serious COVID-19 illness.
If you are a healthcare provider, use your judgment to determine if a patient has signs and symptoms compatible with COVID-19 and whether the patient should be tested. Factors to consider in addition to clinical symptoms may include:
Does the patient have recent travel from an affected area?
Has the patient been in close contact with someone with COVID-19 or with patients with pneumonia of unknown cause?
Does the patient reside in an area where there has been community spread of COVID-19?
If you are a healthcare provider or a public health responder caring for a COVID-19 patient, please take care of yourself and follow recommended infection control procedures.
People who get a fever or cough should consider whether they might have COVID-19, depending on where they live, their travel history or other exposures. More than half of the U.S. is seeing some level of community spread of COVID-19. Testing for COVID-19 may be accessed through medical providers or public health departments, but there is no treatment for this virus. Most people have mild illness and are able to recover at home without medical care.
For people who are ill with COVID-19, but are not sick enough to be hospitalized, please follow CDC guidance on how to reduce the risk of spreading your illness to others. People who are mildly ill with COVID-19 are able to isolate at home during their illness.
If you have been in China or another affected area or have been exposed to someone sick with COVID-19 in the last 14 days, you will face some limitations on your movement and activity. Please follow instructions during this time. Your cooperation is integral to the ongoing public health response to try to slow spread of this virus.
COVID-19 Emergence
COVID-19 is caused by a coronavirus. Coronaviruses are a large family of viruses that are common in people and many different species of animals, including camels, cattle, cats, and bats. Rarely, animal coronaviruses can infect people and then spread between people such as with MERS-CoV, SARS-CoV, and now with this new virus (named SARS-CoV-2).
The SARS-CoV-2 virus is a betacoronavirus, like MERS-CoV and SARS-CoV. All three of these viruses have their origins in bats. The sequences from U.S. patients are similar to the one that China initially posted, suggesting a likely single, recent emergence of this virus from an animal reservoir.
Early on, many of the patients at the epicenter of the outbreak in Wuhan, Hubei Province, China had some link to a large seafood and live animal market, suggesting animal-to-person spread. Later, a growing number of patients reportedly did not have exposure to animal markets, indicating person-to-person spread. Person-to-person spread was subsequently reported outside Hubei and in countries outside China, including in the United States. Some international destinations now have ongoing community spread with the virus that causes COVID-19, as do some parts of the United States. Community spread means some people have been infected and it is not known how or where they became exposed. Learn more about the spread of this newly emerged coronavirus.
Severity
The complete clinical picture with regard to COVID-19 is not fully known. Reported illnesses have ranged from very mild (including some with no reported symptoms) to severe, including illness resulting in death. While information so far suggests that most COVID-19 illness is mild, a reportexternal icon out of China suggests serious illness occurs in 16% of cases. Older people and people of all ages with severe chronic medical conditions — like heart disease, lung disease and diabetes, for example — seem to be at higher risk of developing serious COVID-19 illness. A CDC Morbidity & Mortality Weekly Report that looked at severity of disease among COVID-19 cases in the United States by age group found that 80% of deaths were among adults 65 years and older with the highest percentage of severe outcomes occurring in people 85 years and older.
Learn more about the symptoms associated with COVID-19.
COVID-19 Pandemic
A pandemic is a global outbreak of disease. Pandemics happen when a new virus emerges to infect people and can spread between people sustainably. Because there is little to no pre-existing immunity against the new virus, it spreads worldwide.
The virus that causes COVID-19 is infecting people and spreading easily from person-to-person. Cases have been detected in most countries worldwide and community spread is being detected in a growing number of countries. On March 11, the COVID-19 outbreak was characterized as a pandemic by the WHOexternal icon.
This is the first pandemic known to be caused by the emergence of a new coronavirus. In the past century, there have been four pandemics caused by the emergence of novel influenza viruses. As a result, most research and guidance around pandemics is specific to influenza, but the same premises can be applied to the current COVID-19 pandemic. Pandemics of respiratory disease follow a certain progression outlined in a “Pandemic Intervals Framework.” Pandemics begin with an investigation phase, followed by recognition, initiation, and acceleration phases. The peak of illnesses occurs at the end of the acceleration phase, which is followed by a deceleration phase, during which there is a decrease in illnesses. Different countries can be in different phases of the pandemic at any point in time and different parts of the same country can also be in different phases of a pandemic.
There are ongoing investigations to learn more. This is a rapidly evolving situation and information will be updated as it becomes available.
Risk Assessment
Risk depends on characteristics of the virus, including how well it spreads between people; the severity of resulting illness; and the medical or other measures available to control the impact of the virus (for example, vaccines or medications that can treat the illness) and the relative success of these. In the absence of vaccine or treatment medications, nonpharmaceutical interventions become the most important response strategy. These are community interventions that can reduce the impact of disease.
The risk from COVID-19 to Americans can be broken down into risk of exposure versus risk of serious illness and death.
Risk of exposure:
The immediate risk of being exposed to this virus is still low for most Americans, but as the outbreak expands, that risk will increase. Cases of COVID-19 and instances of community spread are being reported in a growing number of states.
People in places where ongoing community spread of the virus that causes COVID-19 has been reported are at elevated risk of exposure, with the level of risk dependent on the location.
Healthcare workers caring for patients with COVID-19 are at elevated risk of exposure.
Close contacts of persons with COVID-19 also are at elevated risk of exposure.
Travelers returning from affected international locations where community spread is occurring also are at elevated risk of exposure, with level of risk dependent on where they traveled.
Risk of Severe Illness:
Early information out of China, where COVID-19 first started, shows that some people are at higher risk of getting very sick from this illness. This includes:
Older adults, with risk increasing by age.
People who have serious chronic medical conditions like:
Heart disease
Diabetes
Lung disease
CDC has developed guidance to help in the risk assessment and management of people with potential exposures to COVID-19.
What May Happen
More cases of COVID-19 are likely to be identified in the United States in the coming days, including more instances of community spread. CDC expects that widespread transmission of COVID-19 in the United States will occur. In the coming months, most of the U.S. population will be exposed to this virus.
Widespread transmission of COVID-19 could translate into large numbers of people needing medical care at the same time. Schools, childcare centers, and workplaces, may experience more absenteeism. Mass gatherings may be sparsely attended or postponed. Public health and healthcare systems may become overloaded, with elevated rates of hospitalizations and deaths. Other critical infrastructure, such as law enforcement, emergency medical services, and sectors of the transportation industry may also be affected. Healthcare providers and hospitals may be overwhelmed. At this time, there is no vaccine to protect against COVID-19 and no medications approved to treat it. Nonpharmaceutical interventions will be the most important response strategy to try to delay the spread of the virus and reduce the impact of disease.
CDC Response
Global efforts at this time are focused concurrently on lessening the spread and impact of this virus. The federal government is working closely with state, local, tribal, and territorial partners, as well as public health partners, to respond to this public health threat.
Highlights of CDC’s Response
CDC established a COVID-19 Incident Management System on January 7, 2020. On January 21, CDC activated its Emergency Operations Center to better provide ongoing support to the COVID-19 response.
The U.S. government has taken unprecedented steps with respect to travel in response to the growing public health threat posed by this new coronavirus:
Foreign nationals who have been in China, Iran, the United Kingdom, Ireland and any one of the 26 European countries in the Schengen Area within the past 14 days cannot enter the United States.
U.S. citizens, residents, and their immediate family members who have been any one of those countries within in the past 14 days can enter the United States, but they are subject to health monitoring and possible quarantine for up to 14 days.
People at higher risk of serious COVID-19 illness avoid cruise travel and non-essential air travel.
CDC has issued additional specific travel guidance related to COVID-19.
CDC has issued clinical guidance, including:
Clinical Guidance for Management of Patients with Confirmed Coronavirus Disease (COVID-19).
Infection Prevention and Control Recommendations for Patients, including guidance on the use of personal protective equipment (PPE) during a shortage.
CDC also has issued guidance for other settings, including:
Preparing for COVID-19: Long-term Care Facilities, Nursing Homes
Discontinuation of Home Isolation for Persons with COVID-19
CDC has deployed multidisciplinary teams to support state health departments in case identification, contact tracing, clinical management, and public communications.
CDC has worked with federal partners to support the safe return of Americans overseas who have been affected by COVID-19.
An important part of CDC’s role during a public health emergency is to develop a test for the pathogen and equip state and local public health labs with testing capacity.
CDC developed an rRT-PCR test to diagnose COVID-19.
As of the evening of March 17, 89 state and local public health labs in 50 states, the District of Columbia, Guam, and Puerto Rico have successfully verified and are currently using CDC COVID-19 diagnostic tests.
Commercial manufacturers are now producing their own tests.
CDC has grown the COVID-19 virus in cell culture, which is necessary for further studies, including for additional genetic characterization. The cell-grown virus was sent to NIH’s BEI Resources Repositoryexternal icon for use by the broad scientific community.
CDC also is developing a serology test for COVID-19.
Other Available Resources
The following resources are available with information on COVID-19
World Health Organization, Coronavirusexternal icon | How many states in the U.S. have reported cases of COVID-19? | 226 | 50 | 1,277 |
185 | CDC Summary 21 MAR 2020,
https://www.cdc.gov/coronavirus/2019-ncov/cases-updates/summary.html
This is a rapidly evolving situation and CDC will provide updated information and guidance as it becomes available.
Updated March 21, 2020
CDC is responding to a pandemic of respiratory disease spreading from person-to-person caused by a novel (new) coronavirus. The disease has been named “coronavirus disease 2019” (abbreviated “COVID-19”). This situation poses a serious public health risk. The federal government is working closely with state, local, tribal, and territorial partners, as well as public health partners, to respond to this situation. COVID-19 can cause mild to severe illness; most severe illness occurs in older adults.
Situation in U.S.
Different parts of the country are seeing different levels of COVID-19 activity. The United States nationally is in the initiation phase of the pandemic. States in which community spread is occurring are in the acceleration phase. The duration and severity of each pandemic phase can vary depending on the characteristics of the virus and the public health response.
CDC and state and local public health laboratories are testing for the virus that causes COVID-19. View CDC’s Public Health Laboratory Testing map.
All 50 states have reported cases of COVID-19 to CDC.
U.S. COVID-19 cases include:
Imported cases in travelers
Cases among close contacts of a known case
Community-acquired cases where the source of the infection is unknown.
Twenty-seven U.S. states are reporting some community spread of COVID-19.
View latest case counts, deaths, and a map of states with reported cases.
CDC Recommends
Everyone can do their part to help us respond to this emerging public health threat:
On March 16, the White House announced a program called “15 Days to Slow the Spread,”pdf iconexternal icon which is a nationwide effort to slow the spread of COVID-19 through the implementation of social distancing at all levels of society.
Older people and people with severe chronic conditions should take special precautions because they are at higher risk of developing serious COVID-19 illness.
If you are a healthcare provider, use your judgment to determine if a patient has signs and symptoms compatible with COVID-19 and whether the patient should be tested. Factors to consider in addition to clinical symptoms may include:
Does the patient have recent travel from an affected area?
Has the patient been in close contact with someone with COVID-19 or with patients with pneumonia of unknown cause?
Does the patient reside in an area where there has been community spread of COVID-19?
If you are a healthcare provider or a public health responder caring for a COVID-19 patient, please take care of yourself and follow recommended infection control procedures.
People who get a fever or cough should consider whether they might have COVID-19, depending on where they live, their travel history or other exposures. More than half of the U.S. is seeing some level of community spread of COVID-19. Testing for COVID-19 may be accessed through medical providers or public health departments, but there is no treatment for this virus. Most people have mild illness and are able to recover at home without medical care.
For people who are ill with COVID-19, but are not sick enough to be hospitalized, please follow CDC guidance on how to reduce the risk of spreading your illness to others. People who are mildly ill with COVID-19 are able to isolate at home during their illness.
If you have been in China or another affected area or have been exposed to someone sick with COVID-19 in the last 14 days, you will face some limitations on your movement and activity. Please follow instructions during this time. Your cooperation is integral to the ongoing public health response to try to slow spread of this virus.
COVID-19 Emergence
COVID-19 is caused by a coronavirus. Coronaviruses are a large family of viruses that are common in people and many different species of animals, including camels, cattle, cats, and bats. Rarely, animal coronaviruses can infect people and then spread between people such as with MERS-CoV, SARS-CoV, and now with this new virus (named SARS-CoV-2).
The SARS-CoV-2 virus is a betacoronavirus, like MERS-CoV and SARS-CoV. All three of these viruses have their origins in bats. The sequences from U.S. patients are similar to the one that China initially posted, suggesting a likely single, recent emergence of this virus from an animal reservoir.
Early on, many of the patients at the epicenter of the outbreak in Wuhan, Hubei Province, China had some link to a large seafood and live animal market, suggesting animal-to-person spread. Later, a growing number of patients reportedly did not have exposure to animal markets, indicating person-to-person spread. Person-to-person spread was subsequently reported outside Hubei and in countries outside China, including in the United States. Some international destinations now have ongoing community spread with the virus that causes COVID-19, as do some parts of the United States. Community spread means some people have been infected and it is not known how or where they became exposed. Learn more about the spread of this newly emerged coronavirus.
Severity
The complete clinical picture with regard to COVID-19 is not fully known. Reported illnesses have ranged from very mild (including some with no reported symptoms) to severe, including illness resulting in death. While information so far suggests that most COVID-19 illness is mild, a reportexternal icon out of China suggests serious illness occurs in 16% of cases. Older people and people of all ages with severe chronic medical conditions — like heart disease, lung disease and diabetes, for example — seem to be at higher risk of developing serious COVID-19 illness. A CDC Morbidity & Mortality Weekly Report that looked at severity of disease among COVID-19 cases in the United States by age group found that 80% of deaths were among adults 65 years and older with the highest percentage of severe outcomes occurring in people 85 years and older.
Learn more about the symptoms associated with COVID-19.
COVID-19 Pandemic
A pandemic is a global outbreak of disease. Pandemics happen when a new virus emerges to infect people and can spread between people sustainably. Because there is little to no pre-existing immunity against the new virus, it spreads worldwide.
The virus that causes COVID-19 is infecting people and spreading easily from person-to-person. Cases have been detected in most countries worldwide and community spread is being detected in a growing number of countries. On March 11, the COVID-19 outbreak was characterized as a pandemic by the WHOexternal icon.
This is the first pandemic known to be caused by the emergence of a new coronavirus. In the past century, there have been four pandemics caused by the emergence of novel influenza viruses. As a result, most research and guidance around pandemics is specific to influenza, but the same premises can be applied to the current COVID-19 pandemic. Pandemics of respiratory disease follow a certain progression outlined in a “Pandemic Intervals Framework.” Pandemics begin with an investigation phase, followed by recognition, initiation, and acceleration phases. The peak of illnesses occurs at the end of the acceleration phase, which is followed by a deceleration phase, during which there is a decrease in illnesses. Different countries can be in different phases of the pandemic at any point in time and different parts of the same country can also be in different phases of a pandemic.
There are ongoing investigations to learn more. This is a rapidly evolving situation and information will be updated as it becomes available.
Risk Assessment
Risk depends on characteristics of the virus, including how well it spreads between people; the severity of resulting illness; and the medical or other measures available to control the impact of the virus (for example, vaccines or medications that can treat the illness) and the relative success of these. In the absence of vaccine or treatment medications, nonpharmaceutical interventions become the most important response strategy. These are community interventions that can reduce the impact of disease.
The risk from COVID-19 to Americans can be broken down into risk of exposure versus risk of serious illness and death.
Risk of exposure:
The immediate risk of being exposed to this virus is still low for most Americans, but as the outbreak expands, that risk will increase. Cases of COVID-19 and instances of community spread are being reported in a growing number of states.
People in places where ongoing community spread of the virus that causes COVID-19 has been reported are at elevated risk of exposure, with the level of risk dependent on the location.
Healthcare workers caring for patients with COVID-19 are at elevated risk of exposure.
Close contacts of persons with COVID-19 also are at elevated risk of exposure.
Travelers returning from affected international locations where community spread is occurring also are at elevated risk of exposure, with level of risk dependent on where they traveled.
Risk of Severe Illness:
Early information out of China, where COVID-19 first started, shows that some people are at higher risk of getting very sick from this illness. This includes:
Older adults, with risk increasing by age.
People who have serious chronic medical conditions like:
Heart disease
Diabetes
Lung disease
CDC has developed guidance to help in the risk assessment and management of people with potential exposures to COVID-19.
What May Happen
More cases of COVID-19 are likely to be identified in the United States in the coming days, including more instances of community spread. CDC expects that widespread transmission of COVID-19 in the United States will occur. In the coming months, most of the U.S. population will be exposed to this virus.
Widespread transmission of COVID-19 could translate into large numbers of people needing medical care at the same time. Schools, childcare centers, and workplaces, may experience more absenteeism. Mass gatherings may be sparsely attended or postponed. Public health and healthcare systems may become overloaded, with elevated rates of hospitalizations and deaths. Other critical infrastructure, such as law enforcement, emergency medical services, and sectors of the transportation industry may also be affected. Healthcare providers and hospitals may be overwhelmed. At this time, there is no vaccine to protect against COVID-19 and no medications approved to treat it. Nonpharmaceutical interventions will be the most important response strategy to try to delay the spread of the virus and reduce the impact of disease.
CDC Response
Global efforts at this time are focused concurrently on lessening the spread and impact of this virus. The federal government is working closely with state, local, tribal, and territorial partners, as well as public health partners, to respond to this public health threat.
Highlights of CDC’s Response
CDC established a COVID-19 Incident Management System on January 7, 2020. On January 21, CDC activated its Emergency Operations Center to better provide ongoing support to the COVID-19 response.
The U.S. government has taken unprecedented steps with respect to travel in response to the growing public health threat posed by this new coronavirus:
Foreign nationals who have been in China, Iran, the United Kingdom, Ireland and any one of the 26 European countries in the Schengen Area within the past 14 days cannot enter the United States.
U.S. citizens, residents, and their immediate family members who have been any one of those countries within in the past 14 days can enter the United States, but they are subject to health monitoring and possible quarantine for up to 14 days.
People at higher risk of serious COVID-19 illness avoid cruise travel and non-essential air travel.
CDC has issued additional specific travel guidance related to COVID-19.
CDC has issued clinical guidance, including:
Clinical Guidance for Management of Patients with Confirmed Coronavirus Disease (COVID-19).
Infection Prevention and Control Recommendations for Patients, including guidance on the use of personal protective equipment (PPE) during a shortage.
CDC also has issued guidance for other settings, including:
Preparing for COVID-19: Long-term Care Facilities, Nursing Homes
Discontinuation of Home Isolation for Persons with COVID-19
CDC has deployed multidisciplinary teams to support state health departments in case identification, contact tracing, clinical management, and public communications.
CDC has worked with federal partners to support the safe return of Americans overseas who have been affected by COVID-19.
An important part of CDC’s role during a public health emergency is to develop a test for the pathogen and equip state and local public health labs with testing capacity.
CDC developed an rRT-PCR test to diagnose COVID-19.
As of the evening of March 17, 89 state and local public health labs in 50 states, the District of Columbia, Guam, and Puerto Rico have successfully verified and are currently using CDC COVID-19 diagnostic tests.
Commercial manufacturers are now producing their own tests.
CDC has grown the COVID-19 virus in cell culture, which is necessary for further studies, including for additional genetic characterization. The cell-grown virus was sent to NIH’s BEI Resources Repositoryexternal icon for use by the broad scientific community.
CDC also is developing a serology test for COVID-19.
Other Available Resources
The following resources are available with information on COVID-19
World Health Organization, Coronavirusexternal icon | When did the White House launch the "15 Days to Slow the Spread" program? | 227 | March 16 | 1,750 |
185 | CDC Summary 21 MAR 2020,
https://www.cdc.gov/coronavirus/2019-ncov/cases-updates/summary.html
This is a rapidly evolving situation and CDC will provide updated information and guidance as it becomes available.
Updated March 21, 2020
CDC is responding to a pandemic of respiratory disease spreading from person-to-person caused by a novel (new) coronavirus. The disease has been named “coronavirus disease 2019” (abbreviated “COVID-19”). This situation poses a serious public health risk. The federal government is working closely with state, local, tribal, and territorial partners, as well as public health partners, to respond to this situation. COVID-19 can cause mild to severe illness; most severe illness occurs in older adults.
Situation in U.S.
Different parts of the country are seeing different levels of COVID-19 activity. The United States nationally is in the initiation phase of the pandemic. States in which community spread is occurring are in the acceleration phase. The duration and severity of each pandemic phase can vary depending on the characteristics of the virus and the public health response.
CDC and state and local public health laboratories are testing for the virus that causes COVID-19. View CDC’s Public Health Laboratory Testing map.
All 50 states have reported cases of COVID-19 to CDC.
U.S. COVID-19 cases include:
Imported cases in travelers
Cases among close contacts of a known case
Community-acquired cases where the source of the infection is unknown.
Twenty-seven U.S. states are reporting some community spread of COVID-19.
View latest case counts, deaths, and a map of states with reported cases.
CDC Recommends
Everyone can do their part to help us respond to this emerging public health threat:
On March 16, the White House announced a program called “15 Days to Slow the Spread,”pdf iconexternal icon which is a nationwide effort to slow the spread of COVID-19 through the implementation of social distancing at all levels of society.
Older people and people with severe chronic conditions should take special precautions because they are at higher risk of developing serious COVID-19 illness.
If you are a healthcare provider, use your judgment to determine if a patient has signs and symptoms compatible with COVID-19 and whether the patient should be tested. Factors to consider in addition to clinical symptoms may include:
Does the patient have recent travel from an affected area?
Has the patient been in close contact with someone with COVID-19 or with patients with pneumonia of unknown cause?
Does the patient reside in an area where there has been community spread of COVID-19?
If you are a healthcare provider or a public health responder caring for a COVID-19 patient, please take care of yourself and follow recommended infection control procedures.
People who get a fever or cough should consider whether they might have COVID-19, depending on where they live, their travel history or other exposures. More than half of the U.S. is seeing some level of community spread of COVID-19. Testing for COVID-19 may be accessed through medical providers or public health departments, but there is no treatment for this virus. Most people have mild illness and are able to recover at home without medical care.
For people who are ill with COVID-19, but are not sick enough to be hospitalized, please follow CDC guidance on how to reduce the risk of spreading your illness to others. People who are mildly ill with COVID-19 are able to isolate at home during their illness.
If you have been in China or another affected area or have been exposed to someone sick with COVID-19 in the last 14 days, you will face some limitations on your movement and activity. Please follow instructions during this time. Your cooperation is integral to the ongoing public health response to try to slow spread of this virus.
COVID-19 Emergence
COVID-19 is caused by a coronavirus. Coronaviruses are a large family of viruses that are common in people and many different species of animals, including camels, cattle, cats, and bats. Rarely, animal coronaviruses can infect people and then spread between people such as with MERS-CoV, SARS-CoV, and now with this new virus (named SARS-CoV-2).
The SARS-CoV-2 virus is a betacoronavirus, like MERS-CoV and SARS-CoV. All three of these viruses have their origins in bats. The sequences from U.S. patients are similar to the one that China initially posted, suggesting a likely single, recent emergence of this virus from an animal reservoir.
Early on, many of the patients at the epicenter of the outbreak in Wuhan, Hubei Province, China had some link to a large seafood and live animal market, suggesting animal-to-person spread. Later, a growing number of patients reportedly did not have exposure to animal markets, indicating person-to-person spread. Person-to-person spread was subsequently reported outside Hubei and in countries outside China, including in the United States. Some international destinations now have ongoing community spread with the virus that causes COVID-19, as do some parts of the United States. Community spread means some people have been infected and it is not known how or where they became exposed. Learn more about the spread of this newly emerged coronavirus.
Severity
The complete clinical picture with regard to COVID-19 is not fully known. Reported illnesses have ranged from very mild (including some with no reported symptoms) to severe, including illness resulting in death. While information so far suggests that most COVID-19 illness is mild, a reportexternal icon out of China suggests serious illness occurs in 16% of cases. Older people and people of all ages with severe chronic medical conditions — like heart disease, lung disease and diabetes, for example — seem to be at higher risk of developing serious COVID-19 illness. A CDC Morbidity & Mortality Weekly Report that looked at severity of disease among COVID-19 cases in the United States by age group found that 80% of deaths were among adults 65 years and older with the highest percentage of severe outcomes occurring in people 85 years and older.
Learn more about the symptoms associated with COVID-19.
COVID-19 Pandemic
A pandemic is a global outbreak of disease. Pandemics happen when a new virus emerges to infect people and can spread between people sustainably. Because there is little to no pre-existing immunity against the new virus, it spreads worldwide.
The virus that causes COVID-19 is infecting people and spreading easily from person-to-person. Cases have been detected in most countries worldwide and community spread is being detected in a growing number of countries. On March 11, the COVID-19 outbreak was characterized as a pandemic by the WHOexternal icon.
This is the first pandemic known to be caused by the emergence of a new coronavirus. In the past century, there have been four pandemics caused by the emergence of novel influenza viruses. As a result, most research and guidance around pandemics is specific to influenza, but the same premises can be applied to the current COVID-19 pandemic. Pandemics of respiratory disease follow a certain progression outlined in a “Pandemic Intervals Framework.” Pandemics begin with an investigation phase, followed by recognition, initiation, and acceleration phases. The peak of illnesses occurs at the end of the acceleration phase, which is followed by a deceleration phase, during which there is a decrease in illnesses. Different countries can be in different phases of the pandemic at any point in time and different parts of the same country can also be in different phases of a pandemic.
There are ongoing investigations to learn more. This is a rapidly evolving situation and information will be updated as it becomes available.
Risk Assessment
Risk depends on characteristics of the virus, including how well it spreads between people; the severity of resulting illness; and the medical or other measures available to control the impact of the virus (for example, vaccines or medications that can treat the illness) and the relative success of these. In the absence of vaccine or treatment medications, nonpharmaceutical interventions become the most important response strategy. These are community interventions that can reduce the impact of disease.
The risk from COVID-19 to Americans can be broken down into risk of exposure versus risk of serious illness and death.
Risk of exposure:
The immediate risk of being exposed to this virus is still low for most Americans, but as the outbreak expands, that risk will increase. Cases of COVID-19 and instances of community spread are being reported in a growing number of states.
People in places where ongoing community spread of the virus that causes COVID-19 has been reported are at elevated risk of exposure, with the level of risk dependent on the location.
Healthcare workers caring for patients with COVID-19 are at elevated risk of exposure.
Close contacts of persons with COVID-19 also are at elevated risk of exposure.
Travelers returning from affected international locations where community spread is occurring also are at elevated risk of exposure, with level of risk dependent on where they traveled.
Risk of Severe Illness:
Early information out of China, where COVID-19 first started, shows that some people are at higher risk of getting very sick from this illness. This includes:
Older adults, with risk increasing by age.
People who have serious chronic medical conditions like:
Heart disease
Diabetes
Lung disease
CDC has developed guidance to help in the risk assessment and management of people with potential exposures to COVID-19.
What May Happen
More cases of COVID-19 are likely to be identified in the United States in the coming days, including more instances of community spread. CDC expects that widespread transmission of COVID-19 in the United States will occur. In the coming months, most of the U.S. population will be exposed to this virus.
Widespread transmission of COVID-19 could translate into large numbers of people needing medical care at the same time. Schools, childcare centers, and workplaces, may experience more absenteeism. Mass gatherings may be sparsely attended or postponed. Public health and healthcare systems may become overloaded, with elevated rates of hospitalizations and deaths. Other critical infrastructure, such as law enforcement, emergency medical services, and sectors of the transportation industry may also be affected. Healthcare providers and hospitals may be overwhelmed. At this time, there is no vaccine to protect against COVID-19 and no medications approved to treat it. Nonpharmaceutical interventions will be the most important response strategy to try to delay the spread of the virus and reduce the impact of disease.
CDC Response
Global efforts at this time are focused concurrently on lessening the spread and impact of this virus. The federal government is working closely with state, local, tribal, and territorial partners, as well as public health partners, to respond to this public health threat.
Highlights of CDC’s Response
CDC established a COVID-19 Incident Management System on January 7, 2020. On January 21, CDC activated its Emergency Operations Center to better provide ongoing support to the COVID-19 response.
The U.S. government has taken unprecedented steps with respect to travel in response to the growing public health threat posed by this new coronavirus:
Foreign nationals who have been in China, Iran, the United Kingdom, Ireland and any one of the 26 European countries in the Schengen Area within the past 14 days cannot enter the United States.
U.S. citizens, residents, and their immediate family members who have been any one of those countries within in the past 14 days can enter the United States, but they are subject to health monitoring and possible quarantine for up to 14 days.
People at higher risk of serious COVID-19 illness avoid cruise travel and non-essential air travel.
CDC has issued additional specific travel guidance related to COVID-19.
CDC has issued clinical guidance, including:
Clinical Guidance for Management of Patients with Confirmed Coronavirus Disease (COVID-19).
Infection Prevention and Control Recommendations for Patients, including guidance on the use of personal protective equipment (PPE) during a shortage.
CDC also has issued guidance for other settings, including:
Preparing for COVID-19: Long-term Care Facilities, Nursing Homes
Discontinuation of Home Isolation for Persons with COVID-19
CDC has deployed multidisciplinary teams to support state health departments in case identification, contact tracing, clinical management, and public communications.
CDC has worked with federal partners to support the safe return of Americans overseas who have been affected by COVID-19.
An important part of CDC’s role during a public health emergency is to develop a test for the pathogen and equip state and local public health labs with testing capacity.
CDC developed an rRT-PCR test to diagnose COVID-19.
As of the evening of March 17, 89 state and local public health labs in 50 states, the District of Columbia, Guam, and Puerto Rico have successfully verified and are currently using CDC COVID-19 diagnostic tests.
Commercial manufacturers are now producing their own tests.
CDC has grown the COVID-19 virus in cell culture, which is necessary for further studies, including for additional genetic characterization. The cell-grown virus was sent to NIH’s BEI Resources Repositoryexternal icon for use by the broad scientific community.
CDC also is developing a serology test for COVID-19.
Other Available Resources
The following resources are available with information on COVID-19
World Health Organization, Coronavirusexternal icon | What should mildly-ill patients do? | 230 | isolate at home during their illness | 3,493 |
185 | CDC Summary 21 MAR 2020,
https://www.cdc.gov/coronavirus/2019-ncov/cases-updates/summary.html
This is a rapidly evolving situation and CDC will provide updated information and guidance as it becomes available.
Updated March 21, 2020
CDC is responding to a pandemic of respiratory disease spreading from person-to-person caused by a novel (new) coronavirus. The disease has been named “coronavirus disease 2019” (abbreviated “COVID-19”). This situation poses a serious public health risk. The federal government is working closely with state, local, tribal, and territorial partners, as well as public health partners, to respond to this situation. COVID-19 can cause mild to severe illness; most severe illness occurs in older adults.
Situation in U.S.
Different parts of the country are seeing different levels of COVID-19 activity. The United States nationally is in the initiation phase of the pandemic. States in which community spread is occurring are in the acceleration phase. The duration and severity of each pandemic phase can vary depending on the characteristics of the virus and the public health response.
CDC and state and local public health laboratories are testing for the virus that causes COVID-19. View CDC’s Public Health Laboratory Testing map.
All 50 states have reported cases of COVID-19 to CDC.
U.S. COVID-19 cases include:
Imported cases in travelers
Cases among close contacts of a known case
Community-acquired cases where the source of the infection is unknown.
Twenty-seven U.S. states are reporting some community spread of COVID-19.
View latest case counts, deaths, and a map of states with reported cases.
CDC Recommends
Everyone can do their part to help us respond to this emerging public health threat:
On March 16, the White House announced a program called “15 Days to Slow the Spread,”pdf iconexternal icon which is a nationwide effort to slow the spread of COVID-19 through the implementation of social distancing at all levels of society.
Older people and people with severe chronic conditions should take special precautions because they are at higher risk of developing serious COVID-19 illness.
If you are a healthcare provider, use your judgment to determine if a patient has signs and symptoms compatible with COVID-19 and whether the patient should be tested. Factors to consider in addition to clinical symptoms may include:
Does the patient have recent travel from an affected area?
Has the patient been in close contact with someone with COVID-19 or with patients with pneumonia of unknown cause?
Does the patient reside in an area where there has been community spread of COVID-19?
If you are a healthcare provider or a public health responder caring for a COVID-19 patient, please take care of yourself and follow recommended infection control procedures.
People who get a fever or cough should consider whether they might have COVID-19, depending on where they live, their travel history or other exposures. More than half of the U.S. is seeing some level of community spread of COVID-19. Testing for COVID-19 may be accessed through medical providers or public health departments, but there is no treatment for this virus. Most people have mild illness and are able to recover at home without medical care.
For people who are ill with COVID-19, but are not sick enough to be hospitalized, please follow CDC guidance on how to reduce the risk of spreading your illness to others. People who are mildly ill with COVID-19 are able to isolate at home during their illness.
If you have been in China or another affected area or have been exposed to someone sick with COVID-19 in the last 14 days, you will face some limitations on your movement and activity. Please follow instructions during this time. Your cooperation is integral to the ongoing public health response to try to slow spread of this virus.
COVID-19 Emergence
COVID-19 is caused by a coronavirus. Coronaviruses are a large family of viruses that are common in people and many different species of animals, including camels, cattle, cats, and bats. Rarely, animal coronaviruses can infect people and then spread between people such as with MERS-CoV, SARS-CoV, and now with this new virus (named SARS-CoV-2).
The SARS-CoV-2 virus is a betacoronavirus, like MERS-CoV and SARS-CoV. All three of these viruses have their origins in bats. The sequences from U.S. patients are similar to the one that China initially posted, suggesting a likely single, recent emergence of this virus from an animal reservoir.
Early on, many of the patients at the epicenter of the outbreak in Wuhan, Hubei Province, China had some link to a large seafood and live animal market, suggesting animal-to-person spread. Later, a growing number of patients reportedly did not have exposure to animal markets, indicating person-to-person spread. Person-to-person spread was subsequently reported outside Hubei and in countries outside China, including in the United States. Some international destinations now have ongoing community spread with the virus that causes COVID-19, as do some parts of the United States. Community spread means some people have been infected and it is not known how or where they became exposed. Learn more about the spread of this newly emerged coronavirus.
Severity
The complete clinical picture with regard to COVID-19 is not fully known. Reported illnesses have ranged from very mild (including some with no reported symptoms) to severe, including illness resulting in death. While information so far suggests that most COVID-19 illness is mild, a reportexternal icon out of China suggests serious illness occurs in 16% of cases. Older people and people of all ages with severe chronic medical conditions — like heart disease, lung disease and diabetes, for example — seem to be at higher risk of developing serious COVID-19 illness. A CDC Morbidity & Mortality Weekly Report that looked at severity of disease among COVID-19 cases in the United States by age group found that 80% of deaths were among adults 65 years and older with the highest percentage of severe outcomes occurring in people 85 years and older.
Learn more about the symptoms associated with COVID-19.
COVID-19 Pandemic
A pandemic is a global outbreak of disease. Pandemics happen when a new virus emerges to infect people and can spread between people sustainably. Because there is little to no pre-existing immunity against the new virus, it spreads worldwide.
The virus that causes COVID-19 is infecting people and spreading easily from person-to-person. Cases have been detected in most countries worldwide and community spread is being detected in a growing number of countries. On March 11, the COVID-19 outbreak was characterized as a pandemic by the WHOexternal icon.
This is the first pandemic known to be caused by the emergence of a new coronavirus. In the past century, there have been four pandemics caused by the emergence of novel influenza viruses. As a result, most research and guidance around pandemics is specific to influenza, but the same premises can be applied to the current COVID-19 pandemic. Pandemics of respiratory disease follow a certain progression outlined in a “Pandemic Intervals Framework.” Pandemics begin with an investigation phase, followed by recognition, initiation, and acceleration phases. The peak of illnesses occurs at the end of the acceleration phase, which is followed by a deceleration phase, during which there is a decrease in illnesses. Different countries can be in different phases of the pandemic at any point in time and different parts of the same country can also be in different phases of a pandemic.
There are ongoing investigations to learn more. This is a rapidly evolving situation and information will be updated as it becomes available.
Risk Assessment
Risk depends on characteristics of the virus, including how well it spreads between people; the severity of resulting illness; and the medical or other measures available to control the impact of the virus (for example, vaccines or medications that can treat the illness) and the relative success of these. In the absence of vaccine or treatment medications, nonpharmaceutical interventions become the most important response strategy. These are community interventions that can reduce the impact of disease.
The risk from COVID-19 to Americans can be broken down into risk of exposure versus risk of serious illness and death.
Risk of exposure:
The immediate risk of being exposed to this virus is still low for most Americans, but as the outbreak expands, that risk will increase. Cases of COVID-19 and instances of community spread are being reported in a growing number of states.
People in places where ongoing community spread of the virus that causes COVID-19 has been reported are at elevated risk of exposure, with the level of risk dependent on the location.
Healthcare workers caring for patients with COVID-19 are at elevated risk of exposure.
Close contacts of persons with COVID-19 also are at elevated risk of exposure.
Travelers returning from affected international locations where community spread is occurring also are at elevated risk of exposure, with level of risk dependent on where they traveled.
Risk of Severe Illness:
Early information out of China, where COVID-19 first started, shows that some people are at higher risk of getting very sick from this illness. This includes:
Older adults, with risk increasing by age.
People who have serious chronic medical conditions like:
Heart disease
Diabetes
Lung disease
CDC has developed guidance to help in the risk assessment and management of people with potential exposures to COVID-19.
What May Happen
More cases of COVID-19 are likely to be identified in the United States in the coming days, including more instances of community spread. CDC expects that widespread transmission of COVID-19 in the United States will occur. In the coming months, most of the U.S. population will be exposed to this virus.
Widespread transmission of COVID-19 could translate into large numbers of people needing medical care at the same time. Schools, childcare centers, and workplaces, may experience more absenteeism. Mass gatherings may be sparsely attended or postponed. Public health and healthcare systems may become overloaded, with elevated rates of hospitalizations and deaths. Other critical infrastructure, such as law enforcement, emergency medical services, and sectors of the transportation industry may also be affected. Healthcare providers and hospitals may be overwhelmed. At this time, there is no vaccine to protect against COVID-19 and no medications approved to treat it. Nonpharmaceutical interventions will be the most important response strategy to try to delay the spread of the virus and reduce the impact of disease.
CDC Response
Global efforts at this time are focused concurrently on lessening the spread and impact of this virus. The federal government is working closely with state, local, tribal, and territorial partners, as well as public health partners, to respond to this public health threat.
Highlights of CDC’s Response
CDC established a COVID-19 Incident Management System on January 7, 2020. On January 21, CDC activated its Emergency Operations Center to better provide ongoing support to the COVID-19 response.
The U.S. government has taken unprecedented steps with respect to travel in response to the growing public health threat posed by this new coronavirus:
Foreign nationals who have been in China, Iran, the United Kingdom, Ireland and any one of the 26 European countries in the Schengen Area within the past 14 days cannot enter the United States.
U.S. citizens, residents, and their immediate family members who have been any one of those countries within in the past 14 days can enter the United States, but they are subject to health monitoring and possible quarantine for up to 14 days.
People at higher risk of serious COVID-19 illness avoid cruise travel and non-essential air travel.
CDC has issued additional specific travel guidance related to COVID-19.
CDC has issued clinical guidance, including:
Clinical Guidance for Management of Patients with Confirmed Coronavirus Disease (COVID-19).
Infection Prevention and Control Recommendations for Patients, including guidance on the use of personal protective equipment (PPE) during a shortage.
CDC also has issued guidance for other settings, including:
Preparing for COVID-19: Long-term Care Facilities, Nursing Homes
Discontinuation of Home Isolation for Persons with COVID-19
CDC has deployed multidisciplinary teams to support state health departments in case identification, contact tracing, clinical management, and public communications.
CDC has worked with federal partners to support the safe return of Americans overseas who have been affected by COVID-19.
An important part of CDC’s role during a public health emergency is to develop a test for the pathogen and equip state and local public health labs with testing capacity.
CDC developed an rRT-PCR test to diagnose COVID-19.
As of the evening of March 17, 89 state and local public health labs in 50 states, the District of Columbia, Guam, and Puerto Rico have successfully verified and are currently using CDC COVID-19 diagnostic tests.
Commercial manufacturers are now producing their own tests.
CDC has grown the COVID-19 virus in cell culture, which is necessary for further studies, including for additional genetic characterization. The cell-grown virus was sent to NIH’s BEI Resources Repositoryexternal icon for use by the broad scientific community.
CDC also is developing a serology test for COVID-19.
Other Available Resources
The following resources are available with information on COVID-19
World Health Organization, Coronavirusexternal icon | What type of virus is SARS-CoV-2? | 231 | betacoronavirus | 4,258 |
Subsets and Splits